NAMIC Wiki - User contributions [en]

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:24:19Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3"/>.
|}

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

Interactions with Anneke and Giampaolo were important in order to understand how to improve the CNN topology.

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:23:43Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3"/>.
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

Interactions with Anneke and Giampaolo were important in order to understand how to improve the CNN topology.

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:22:18Z

GabrielePiantadosi: /* Preliminary Results */

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3">.
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

Interactions with Anneke and Giampaolo were important in order to understand how to improve the CNN topology.

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:20:22Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3">.
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:19:27Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3">.
|}

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:18:06Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3">.
|}

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:16:52Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in
<ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).</ref>
<ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.</ref>. In particular, "u-net" provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3">.
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:14:31Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms <ref name="r2">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.</ref> and (b) the classification of each segmented ROI according to its aggressiveness <ref name="r3">Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.</ref>. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 <ref name="r1">Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.</ref>)'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in <ref name="r4">Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).<ref/> <ref name="r5">Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.<ref/>. In particular, “u-net” provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation <ref name="r5"/>]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches <ref name="r1" />.
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed <ref name="r2" />.
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed <ref name="r3">.
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:05:02Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms [2] and (b) the classification of each segmented ROI according to its aggressiveness [3]. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 [1])'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in [4, 5]. In particular, “u-net” provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches [1].
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed [2].
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed [3].
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
1. Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.

2. Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.

3. Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.

4. Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).

5. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T12:04:10Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (gabriele.piantadosi@unina.it - University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of suspicious regions of interest (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms [1] and (b) the classification of each segmented ROI according to its aggressiveness. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Breast Cancer analysis could be approached as a classification problem via Computer Aided Detection (CAD) system.
[[File:Typical Computer Aided Detection System.png|thumb|center|upright=2.0|Typical Computer Aided Detection System]]

'''Breats-Mask Extraction'''

With the aim of reducing the computational cost of further steps and attenuate noise caused by extraneous voxel (VOlumetric piXEL) a binary mask representing only breast parenchyma and excluding background and other tissues is extracted. Segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements.

'''My Previous Proposal (ICPR2016 [1])'''

The segmentation of breast parenchyma has been addressed relying on fuzzy binary clustering, breast anatomical priors and morphological refinements. In breast segmentation, the most difficult issue to address is discriminating the breast parenchyma from the pectoral muscle since signal intensities, textures and anatomical structures of these tissues are very close each other.

[[File:A Breast-Mask extracted from DCE-MRI data.png|thumb|center|upright=2.0|A Breast-Mask extracted from DCE-MRI data]]

The proposed breast mask extraction approach overcomes the issue by mixing geometrical-based and pixel-based approaches. It relies on geometrical anatomical priors to take advance of anatomical knowledge of the breast key points and uses a pixel base segmentation to obtain the best threshold for each border.

[[File:Anatomical key-points.png|thumb|center|upright=2.0|Anatomical key-points]]

It uses a pixel-based Fuzzy C-Means (FCM) clustering to shift the breast mask extraction from a simple grey-level based segmentation to a membership probability one. Moreover, it exploits novel geometrical consideration to weight the classes membership probability according to the breast anatomy.

[[File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png|thumb|center|upright=2.0|Result of FCM segmentation (in red) overlapped to the Ground Truth (green)]]

The result is an automated procedure, able to extract an accurate breast mask without any prior information on the patient dataset (as in the case of atlases).

'''Deep Proposal'''

The novel proposed lesion detection module performs the breast segmentation relying on deep approaches such as those proposed in [4, 5]. In particular, “u-net” provides segmentation via deep Convolutional Neural Networks (dCNN).

[[File:U-Net architecture for image segmentation.png|thumb|center|upright=3.0|U-Net architecture for image segmentation]]

Segmentation has been performed in Axial projection

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Breast segmentation of 4D DCE-MRI volumes via deep approaches [1].
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
*Include the model in a Slicer Module.
*Relying on the Breast Mask Lesion detection via Deep approaches should be performed [2].
*Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed [3].
|}

==Preliminary Results==
{| class="wikitable"
! style="font-weight: bold;" | Approach
! style="font-weight: bold;" | Validation
! style="font-weight: bold;" | Dice
|-
| fuzzy c-means
| leave one patient out
| 83.25
|-
| fuzzy c-means + anatomical priors
| leave one patient out
| 88.85
|-
| fuzzy c-means + anatomical priors + postprocessing
| leave one patient out
| 91.36
|-
| "u-net" without batch-normalisation
| 30 epochs (hold-out)
| 76.58
|-
| "u-net" with batch-normalisation
| 30 epochs (hold-out)
| 82.19
|}

==Background and References==
1. Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2016, December). Breast Segmentation using Fuzzy C-Means and anatomical priors in DCE-MRI. In Pattern Recognition (ICPR), 2016 23rd International Conference on (pp. 1472-1477). IEEE.

2. Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg.

3. Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2015). LBP-TOP for volume lesion classification in breast DCE-MRI. In 18th International Conference on Image Analysis and Processing, ICIAP 2015. Springer Verlag.

4. Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440).

5. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. arXiv preprint arXiv:1505.04597.

File:U-Net architecture for image segmentation.png

2017-06-30T11:54:45Z

GabrielePiantadosi:

U-Net architecture for image segmentation

File:Result of FCM segmentation (in red) overlapped to the Ground Truth (green).png

2017-06-30T11:50:06Z

GabrielePiantadosi:

Result of FCM segmentation (in red) overlapped to the Ground Truth (green)

File:Anatomical key-points.png

2017-06-30T11:49:20Z

GabrielePiantadosi:

Anatomical key-points

File:A Breast-Mask extracted from DCE-MRI data.png

2017-06-30T11:47:57Z

GabrielePiantadosi:

A Breast-Mask extracted from DCE-MRI data

File:Typical Computer Aided Detection System.png

2017-06-30T11:43:30Z

GabrielePiantadosi:

Typical Computer Aided Detection System

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T11:39:34Z

GabrielePiantadosi:

Project Week 25/Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T10:13:01Z

GabrielePiantadosi: Blanked the page

Project Week 25

2017-06-30T10:12:15Z

GabrielePiantadosi:

__NOTOC__
Back to [[Events]]

A summary of all past [[Project_Events#Past_Project_Weeks|Project Events]].

[[image:PW25.png|300px]] [[image:IEL_logo.png|225px]]
<br>

=Welcome to the web page for the 25th Project Week!=
It is a pleasure to announce that the 25th Project week will be held in [https://goo.gl/maps/b9CpkFxNyWN2 Catanzaro Lido] (Calabria, Italy) on June 26-30, 2017. This is the first time in Italy for the Slicer Community, and the event is organized in cooperation with [http://www.imagenglab.com ImagEngLab]. Catanzaro Lido is a city on the Ionian Sea, in the middle of Squillace Gulf where, according to the ancient legend, Odysseus started his journey back to Ithaca. Of course bring your swimsuit...the conference room and the hotel are 20 meters far away from the beach!

This project week is an event [[Post-NCBC-2014|endorsed]] by the MICCAI society.

Please make sure that you are on the NA-MIC Project Week [http://public.kitware.com/mailman/listinfo/na-mic-project-week mailing list].

===Local Organizing Committee===
*[http://www.imagenglab.com/newsite/mf_spadea/ Maria Francesca Spadea, PhD]
*[http://www.imagenglab.com/newsite/paolo_zaffino/ Paolo Zaffino, PhD].

==Logistics==
*'''Dates:''' June 26-30, 2017. Details in the calendar below.
*'''Location:''' [http://www.hotelperladelporto.it/en/home-page.aspx Perla del Porto Hotel]. [mailto:prenotazioni@hotelperladelporto.it Booking]. Subject line: "Slicer Summer Project Week 2017". Special rates: Single room, full bed, 79 € per night (1 person)/ Single room, queen bed 89 € per night (1 person)/ Double room, queen bed 99 € per night (2 people)/Triple room, 110 € (3 people)
*'''Registration:''' To register please visit this [http://www.imagenglab.com/newsite/project-week page]
*'''Registration Fee:''' 320€ and it includes lunches, coffee breaks and airport connections
*'''Hotel:''' [http://www.hotelperladelporto.it/en/home-page.aspx Perla del Porto Hotel]. The closest airport is [http://www.lameziaairport.it/english/ Lamezia Terme Airport (IATA: SUF)].
*'''Transportation from Airport to Hotel:''' Your registration fee includes ground transportation [https://www.google.com/maps/dir/Lamezia+Terme+International+Airport,+Via+Aeroporto,+88046+Lamezia+Terme+CZ,+Italy/BEST+WESTERN+PLUS+Hotel+Perla+Del+Porto,+Via+Martiri+di+Cefalonia,+64,+88100+Catanzaro,+Italy/@38.868758,16.1564814,10z/data=!3m1!4b1!4m14!4m13!1m5!1m1!1s0x133fe15a3cbed47f:0x544ab120c3de78a6!2m2!1d16.2434017!2d38.9065845!1m5!1m1!1s0x134003d668252a13:0x2989caf676f45a72!2m2!1d16.6312407!2d38.827712!3e0 to/from the hotel and airport]. Please fill out this [https://goo.gl/forms/7vmhxZSHy8Z1A62z2 form] to request transportation
*'''Local points of interest (pubs, restaurants, bar):''' [https://www.google.com/maps/d/viewer?mid=1FU63ik9Do3zzP6K2kvLVTtM2at8&ll=38.86221979925013%2C16.44292274999998&z=12 map] (constantly updated)

==Calendar==
{{#widget:Google Calendar
|id=kitware.com_sb07i171olac9aavh46ir495c4@group.calendar.google.com
|timezone=America/New_York&dates=20170108%2F20170114
|title=NA-MIC Project Week (Timezone is Italy / GMT+02.)
|view=WEEK
|dates=20170626/20170701
}}

iCal (.ics) link: https://calendar.google.com/calendar/ical/kitware.com_sb07i171olac9aavh46ir495c4%40group.calendar.google.com/public/basic.ics

==Breakout Sessions==
*[[Project_Week_25/Neuro-navigation_breakout_session|Neuro-navigation Breakout Session]]
*[[Project_Week_25/Segment_editor_breakout_session|Segment Editor Breakout Session]]

=Projects=

<big>Please duplicate [https://na-mic.org/wiki/Project_Week_Template this template] to create a page for your project. </big>

Please put a brief preliminary title for your project here with some names in parenthesis for potential team members
==Deep Learning Applications==
#[[Project_Week_25/NeedleSegmentation |CNN for Needle Segmentation from MRI Images]]
#[[Project_Week_25/CNN_for_multi-plane_prostate_segmentation | CNN for Multi-plane Prostate Segmentation]]
#[[Project_Week_25/CNN for PseudoCT Generation from T1T2 MR|CNN for PseudoCT Generation from T1/T2 MRI]]
#[[Project_Week_25/Breast Segmentation in DCE-MRI via Deep Learning Approaches | Breast Segmentation in DCE-MRI via Deep Learning Approaches]]

==Augmented Reality and Virtual Reality==
#[[Project_Week_25/Improving Depth Perception in Interventional Augmented Reality Visualization/Sonification | Improving Depth Perception in Interventional Augmented Reality Visualization/Sonification]]
# [[Project_Week_25/Next_Generation_GPU_Volume_Rendering | Next Generation of Volume Rendering in VTK ]]
#[[Project_Week_25/Intra-operative deformable_registration_based_on_dense_point_cloud_reconstruction |Intra-operative Deformable Registration Based on Dense Point Cloud Reconstruction for Augmented Reality in Laproscopic Surgery]]


==Navigating Ultrasound==
# [[Project_Week_25/Tracked-Ultrasound-Standardization-IV | Tracked Ultrasound Standardization IV: Controlling US Acquisition]]
#[[Project_Week_25/Segmentation for improving image registration of preoperative MRI with intraoperative ultrasound images for neuro-navigation |Segmentation for Improving Image Registration of Preoperative MRI with Intraoperative Ultrasound Images for Neuro-navigation]]
==DICOM==
#[[Project_Week_25/DICOM_for_Quantitative_Imaging_and_Integration_with_Processing_Applications|DICOM for Quantitative Imaging and Integration with Processing Applications]]
#[[Project_Week_25/Conversion of DICOM Single Frame MR to Enhanced Multiframe | Conversion of DICOM Single Frame MR to Enhanced Multiframe]]
#[[Project_Week_25/DICOM Segmentation Support for Cornerstone and OHIF Viewer | DICOM Segmentation Support for Cornerstone/ OHIF Viewer]]

==Chest Image Processing==
#[[Project_Week_25/SlicerCIP_Bronchiectasis | SlicerCIP: Tool for Quantitative Analysis of Bronchiectasis]]
#[[Project_Week_25/SlicerCIP_ReportsTool | SlicerCIP: Tool for Creation of Reports for Quantitive Analysis]]
==Platform==
#[[Project_Week_25/Interactive_Manipulation_of_Plots_and_Graphs | Interactive Manipulation of Plots and Graphs]]
#[[Project_Week_25/Internationalizing Slicer Modules|Internationalizing Slicer Modules]]
#[[Project_Week_25/Interfacing Slicer to Mobile Phone-controlled Sensors|Interfacing Slicer to Mobile Phone-controlled Sensors]]
#[[Project_Week_25/Slicer and 3D Printing|Slicer and 3D Printing]]
#[[Project_Week_25/Multimodal: | Multimodal, Multiresolution, Multivolume Data]]
#[[Project_Week_25/Human-Computer_Interaction_under_sterile_conditions |Human-Computer Interaction under Sterile Conditions]]

==Applications Smorgasbord==
#[[Project_Week_25/Steerable Catheters Path Planner Extension for Brain Surgery Applications | Steerable Catheters Path Planner Extension for Brain Surgery Applications]]
#[[Project_Week_25/Slice-to-Volume Registration to Support MRI guided Interventions | Slice-to-Volume Registration to Support MRI Guided Interventions ]]
#[[Project_Week_25/Surgical_Planning_In_Stereotaxy | Surgical Planning In Stereotaxy ]]
#[[Project_Week_25/SALT_Spatiotemporal_Modeling: | Slicer SALT Validation: Spatiotemporal Modeling of Subcortical Structures ]]
#[[Project_Week_25/Wrist_Kinematics: | Kinematic Analysis of the Wrist from Dynamic MRI]]
#[[Project_Week_25/EBP | External Beam Planning]]

=Registrants=

Do not add your name to this list - it is maintained by the organizers based on your paid registration. To register, visit this [http://www.imagenglab.com/newsite/project-week/ registration site].

# Kikinis, Ron :: Brigham and Women's Hospital, Harvard Medical School, USA
# Pieper, Steve :: Isomics, Inc., USA
# Kapur, Tina :: Brigham and Women's Hospital, Harvard Medical School, USA
# Spadea, Maria Francesca :: Magna Graecia University, Italy
# Zaffino, Paolo :: Magna Graecia University, Italy
# Scaramuzzino, Salvatore :: Magna Graecia University/ASL Vercelli, Italy
# Pileggi, Giampaolo :: Magna Graecia University, Italy/German Cancer Research Center (DKFZ), Germany
# Rackerseder, Julia :: Technical University of Munich, Germany
# Pinter, Csaba :: Queen's University, Canada
# Kraß, Scheherazade :: University of Bremen, Germany
# Gerig, Guido :: NYU Tandon School of Engineering, USA
# Punzo, Davide :: Kapteyn Astronomical Institute, University of Groningen, The Netherlands
# Drouin, Simon :: NeuroImaging and Surgical Technologies (NIST) Lab, Canada
# Lasso, Andras :: School of Computing, Queen's University, Canada
# Favaro, Alberto :: Politecnico di Milano, Italy
# Leger, Etienne :: Concordia University, Canada
# Ziegler, Erik :: Ziegler Consult SAS
# Onken, Michael :: Open Connections GmbH, Germany
# Pinzi, Marlene :: Imperial College, UK
# Nitsch, Jennifer :: University of Bremen, Germany
# Moccia, Sara :: Politecnico di Milano, Italy
# Black, David :: Fraunhofer Institute for Medical Image Computing MEVIS, Bremen, Germany
# Penzkofer, Tobias :: Charité Universitätsmedizin, Berlin, Germany
# Hansen, Christian :: University of Magdeburg, Germany
# Vegard Solberg, Ole :: Norway
# Heinrich, Florian :: University of Magdeburg, Germany
# Mewes, André :: University of Magdeburg, Germany
# Hatscher, Benjamin :: University of Magdeburg, Germany
# Hettig, Julian :: University of Magdeburg, Germany
# Meyer, Anneke :: University of Magdeburg, Germany
# Gulamhussene, Gino :: University of Magdeburg, Germany
# Cassetta, Roberto :: Polytechnic University of Milan, Italy
# Fillion-Robin, Jean-Christophe :: Kitware, Inc., USA
# Metzger, Jasmin :: German Cancer Research Center (DKFZ), Germany
# Fishbaugh, James :: NYU Tandon School of Engineering, USA
# Nolden, Marco :: German Cancer Research Center (DKFZ), Germany
# Nehrkorn, Jorge Quintero :: Canary Islands, Spain
# Perez Garcia, Fernando :: ICM Brain & Spine Institute, Paris, France
# De Momi, Elena :: Polytechnic University of Milan, Italy
# Piantadosi, Gabriele :: DIETI, Federico II di Napoli, Italy
# Pernelle, Guillaume :: Imperial College, UK
# San Jose, Raul :: Brigham and Women's Hospital, Harvard Medical School, USA
# Nardelli, Pietro :: Brigham and Women's Hospital, Harvard Medical School, USA
# Fernandez Vidal, Sara :: ICM Brain & Spine Institute, Paris, France
# Sharp, Gregory :: Massachusetts General Hospital, Harvard Medical School, USA
# Moiraghi, Alessandro :: University of Milan, Italy
# Seco, Joao :: German Cancer Research Center (DKFZ), Germany
# Pumar Carreras, Nayra :: Canary Islands, Spain
# Afonso Suarez, Maria Dolores :: Canary Islands, Spain
# Alzola Ruiz, Juan :: Canary Islands, Spain
# Pujol, Sonia :: Brigham and Women's Hospital, Harvard Medical School, USA

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T10:12:03Z

GabrielePiantadosi: GabrielePiantadosi moved page Project Week 25/Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches to Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Cancer Analysis in DCE-MRI=

==Key Investigators==

*[https://www.researchgate.net/profile/Gabriele_Piantadosi Gabriele Piantadosi] (University Federico II di Napoli, Italy)

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of the suspicious region of interests (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms and (b) the classification of each segmented ROI according to its aggressiveness. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Segmentation of the breast parenchyma could be approached as a classification problem.

[[File:Lesion Segmentation in DCE-MRI.png|thumb|center|upright=2.0|Lesion Segmentation in DCE-MRI]]

In the image, my previous proposal<ref>Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-68). Springer Berlin Heidelberg</ref>: classical machine learning approaches using Support Vector Machine (SVM) trained with dynamic features, extracted from a suitable pre-selected area by using a pixel-based approach. A pre-selection mask strongly improved the final result.

The novel proposed lesion detection module performs the segmentation of lesions in Regions of Interest (ROIs) by means of classification at a pixel level or as dense regions relying on deep approaches such as those proposed in<ref>Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-40)</ref>.

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Lesion segmentation of 4D DCE-MRI volumes via deep approaches.
[[File:Segmented lesion in a 3D reconstructed Breast.png|frame|center|Segmented lesion in a 3D reconstructed Breast]]
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavor
#Try to do better by using a specifically designed network architecture.

|
Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed.
|}

==Background and References==
<references />

Project Week 25/Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T10:12:03Z

#REDIRECT [[Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches]]

Project Week 25/Breast Cancer Analysis in DCE-MRI

2017-06-30T10:11:33Z

GabrielePiantadosi: Blanked the page

Project Week 25/Breast Analysis in DCE-MRI

2017-06-30T10:11:23Z

GabrielePiantadosi: Blanked the page

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T10:10:33Z

GabrielePiantadosi: GabrielePiantadosi moved page Project Week 25/Breast Analysis in DCE-MRI to Project Week 25/Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches

Project Week 25/Breast Analysis in DCE-MRI

2017-06-30T10:10:33Z

GabrielePiantadosi: GabrielePiantadosi moved page Project Week 25/Breast Analysis in DCE-MRI to Project Week 25/Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches

#REDIRECT [[Project Week 25/Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches]]

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-30T10:09:30Z

GabrielePiantadosi: GabrielePiantadosi moved page Project Week 25/Breast Cancer Analysis in DCE-MRI to Project Week 25/Breast Analysis in DCE-MRI

Project Week 25/Breast Cancer Analysis in DCE-MRI

2017-06-30T10:09:30Z

GabrielePiantadosi: GabrielePiantadosi moved page Project Week 25/Breast Cancer Analysis in DCE-MRI to Project Week 25/Breast Analysis in DCE-MRI

#REDIRECT [[Project Week 25/Breast Analysis in DCE-MRI]]

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-26T08:53:38Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Cancer Analysis in DCE-MRI=

==Key Investigators==

*Gabriele Piantadosi (University Federico II di Napoli, Italy)\

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of the suspicious region of interests (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms and (b) the classification of each segmented ROI according to its aggressiveness. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Segmentation of the breast parenchyma could be approached as a classification problem.

[[File:Lesion Segmentation in DCE-MRI.png|thumb|center|upright=2.0|Lesion Segmentation in DCE-MRI]]

In the image, my previous proposal<ref>Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg</ref>: classical machine learning approaches using Support Vector Machine (SVM) trained with dynamic features, extracted from a suitable pre-selected area by using a pixel-based approach. A pre-selection mask strongly improved the final result.

The novel proposed lesion detection module performs the segmentation of lesions in Regions of Interest (ROIs) by means of classification at a pixel level or as dense regions relying on deep approaches such as those proposed in<ref>Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440)</ref>.

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Lesion segmentation of 4D DCE-MRI volumes via deep approaches.
[[File:Segmented lesion in a 3D reconstructed Breast.png|frame|center|Segmented lesion in a 3D reconstructed Breast]]
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the Caffe environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavour
#Try to do better by using a specifically designed network architecture.

|
Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed.
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-24T08:00:17Z

GabrielePiantadosi:

__NOTOC__
Back to [[Project_Week_25#Projects|Projects List]]

=Breast Cancer Analysis in DCE-MRI=

==Key Investigators==

*Gabriele Piantadosi (University Federico II di Napoli, Italy)\

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of the suspicious region of interests (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms and (b) the classification of each segmented ROI according to its aggressiveness. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Segmentation of the breast parenchyma could be approached as a classification problem.

[[File:Lesion Segmentation in DCE-MRI.png|thumb|center|upright=2.0|Lesion Segmentation in DCE-MRI]]

In the image, my previous proposal<ref>Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg</ref>: classical machine learning approaches using Support Vector Machine (SVM) trained with dynamic features, extracted from a suitable pre-selected area by using a pixel-based approach. A pre-selection mask strongly improved the final result.

The novel proposed lesion detection module performs the segmentation of lesions in Regions of Interest (ROIs) by means of classification at a pixel level or as dense regions relying on deep approaches such as those proposed in<ref>Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440)</ref>.

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Lesion segmentation of 4D DCE-MRI volumes via deep approaches.
[[File:Segmented lesion in a 3D reconstructed Breast.png|frame|center|Segmented lesion in a 3D reconstructed Breast]]
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the TensorFlow environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavour
#Try to do better by using a specifically designed network architecture.

|
Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed.
|}

==Background and References==
<references />

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-24T07:58:42Z

GabrielePiantadosi:

__NOTOC__

Back to [[Project_Week_25#Projects|Projects List]]

'''Breast Cancer Analysis in DCE-MRI'''

==Key Investigators==

*Gabriele Piantadosi (University Federico II di Napoli, Italy)\

==Background==
In recent years Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has gained popularity as an important complementary diagnostic methodology for early detection of breast cancer. It has demonstrated a great potential in the screening of high-risk women, in staging newly diagnosed breast cancer patients and in assessing therapy effects thanks to its minimal invasiveness and to the possibility to visualise 3D high resolution dynamic (functional) information not available with conventional RX imaging.
Among the major issues in developing CAD systems for breast DCE-MRI there are: (a) the detection of the suspicious region of interests (ROIs) as sensibly as possible, while simultaneously minimising the number of false alarms and (b) the classification of each segmented ROI according to its aggressiveness. This task is made harder by the peculiarity of DCE-MRI breast examinations: breast movements due to inspiration, a huge diversity of lesion types.

==Project Description==
Segmentation of the breast parenchyma could be approached as a classification problem.

[[File:Lesion Segmentation in DCE-MRI.png|thumb|center|upright=2.0|Lesion Segmentation in DCE-MRI]]

In the image, my previous proposal<ref>Marrone, S., Piantadosi, G., Fusco, R., Petrillo, A., Sansone, M., & Sansone, C. (2013, September). Automatic lesion detection in breast DCE-MRI. In International Conference on Image Analysis and Processing (pp. 359-368). Springer Berlin Heidelberg</ref>: classical machine learning approaches using Support Vector Machine (SVM) trained with dynamic features, extracted from a suitable pre-selected area by using a pixel-based approach. A pre-selection mask strongly improved the final result.

The novel proposed lesion detection module performs the segmentation of lesions in Regions of Interest (ROIs) by means of classification at a pixel level or as dense regions relying on deep approaches such as those proposed in<ref>Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 3431-3440)</ref>.

{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|
*Lesion segmentation of 4D DCE-MRI volumes via deep approaches.
[[File:Segmented lesion in a 3D reconstructed Breast.png|frame|center|Segmented lesion in a 3D reconstructed Breast]]
*Results evaluation with 42 patients provided of manually segmented ground-truth.

|
#Prepare the TensorFlow environment
#Prepare the Data
#Test the classical deep networks architectures
#Evaluate results in a k-fold flavour
#Try to do better by using a specifically designed network architecture.

|
Relying on the segmented regions of interest (ROIs) a lesion malignity assessment via deep approaches should be performed.
|}

==Background and References==
<references />

File:Segmented lesion in a 3D reconstructed Breast.png

2017-06-24T07:50:33Z

GabrielePiantadosi: GabrielePiantadosi uploaded a new version of File:Segmented lesion in a 3D reconstructed Breast.png

Segmented lesion in a 3D reconstructed Breast

File:Segmented lesion in a 3D reconstructed Breast.png

2017-06-24T07:44:26Z

GabrielePiantadosi:

Segmented lesion in a 3D reconstructed Breast

File:Lesion Segmentation in DCE-MRI.png

2017-06-24T07:34:31Z

GabrielePiantadosi:

Lesion Segmentation in DCE-MRI

Project Week 25/Breast Segmentation in DCE-MRI via Deep Learning Approaches

2017-06-24T07:21:29Z

GabrielePiantadosi: Created page with "__NOTOC__ Back to Projects List ==Key Investigators==  *Investigator 1 (Affiliation) *Investigator 2 (Af..."

__NOTOC__

Back to [[Project_Week_25#Projects|Projects List]]

==Key Investigators==

*Investigator 1 (Affiliation)
*Investigator 2 (Affiliation)
*Investigator 3 (Affiliation)

==Project Description==
{| class="wikitable"
! style="text-align: left; width:27%" | Objective
! style="text-align: left; width:27%" | Approach and Plan
! style="text-align: left; width:27%" | Progress and Next Steps
|- style="vertical-align:top;"
|

|

|

|}

==Illustrations==

https://www.slicer.org/img/Slicer4Announcement-HiRes.png

<embedvideo service="youtube">https://www.youtube.com/watch?v=MKLWzD0PiIc</embedvideo>

==Background and References==

Project Week 25

2017-06-24T07:21:15Z

GabrielePiantadosi: /* Projects */

__NOTOC__
Back to [[Events]]

A summary of all past [[Project_Events#Past_Project_Weeks|Project Events]].

[[image:PW25.png|300px]] [[image:IEL_logo.png|225px]]
<br>

=Welcome to the web page for the 25th Project Week!=
It is a pleasure to announce that the 25th Project week will be held in [https://goo.gl/maps/b9CpkFxNyWN2 Catanzaro Lido] (Calabria, Italy) on June 26-30, 2017. This is the first time in Italy for the Slicer Community, and the event is organized in cooperation with [http://www.imagenglab.com ImagEngLab]. Catanzaro Lido is a city on the Ionian Sea, in the middle of Squillace Gulf where, according to the ancient legend, Odysseus started his journey back to Ithaca. Of course bring your swimsuit...the conference room and the hotel are 20 meters far away from the beach!

This project week is an event [[Post-NCBC-2014|endorsed]] by the MICCAI society.

Please make sure that you are on the NA-MIC Project Week [http://public.kitware.com/mailman/listinfo/na-mic-project-week mailing list].

===Local Organizing Committee===
*[http://www.imagenglab.com/newsite/mf_spadea/ Maria Francesca Spadea, PhD]
*[http://www.imagenglab.com/newsite/paolo_zaffino/ Paolo Zaffino, PhD].

==Videoconferences for preparation==
Time: Wednesday 9am EDT (GMT -4), April 15 to June 21, 2017
URL: [https://zoom.us/j/725310144 click here] to join the videoconference (zoom is the tool as of May 31)

#(Tina Kapur) Hangout #1: April 5 ([[PW25 Hangouts Notes|Notes]])
#(Steve Pieper) Hangout #2: April 12: Web browser based computing: Dockerized Slicer with remote computing and GPU computing; Cornerstone/LesionTracker OHIF; XTK->AMI (threejs); ePad; vtk.js; QWebEngine;
#(Andras Lasso) Hangout #3: April 19: Connecting devices such as surgical navigation, ultrasound, 3D Slicer, PLUS, OpenIGTLink, Augmented reality;
#(Tina Kapur) Hangout #4: April 26: Deep Learning for Detection of Cancer and Instruments;
#(Simon Drouin) Hangout #5: May 3: Volume Rendering, Augmented Reality, and Virtual Reality. [https://docs.google.com/document/d/1UwdSzjnDm1yEeQ44OEhXWbH6V83Uo1Cd4KngxoyrRdI/edit Notes];
#(Tina Kapur) Hangout #6: May 10: For new participants: What is project week and how to get the most out of participating in it?
#(Andrey Fedorov) Hangout #7: May 17:DICOM for Quantitative Imaging and integration with processing applications. ([http://bit.ly/2017NSPW-DICOM notes])
#(Tina Kapur) Hangout #8: May 24:Discussion of Projects and teams that have been provided on the wiki page by participants, internationalization strategy for Slicer.
#(Francesca Spadea) Hangout #9: May 31: Review of local logistics -- all registered attendees should join
#(Tina Kapur) Hangout #10: June 7: Discussion of Projects, Project Pages
#(Francesca Spadea, Tina Kapur) Hangout #11: June 14: Review of Project Pages and Logistics
#(Francesca Spadea) Hangout #12: June 21: Review of local logistics -- all registered attendees with questions should join

==Logistics==
*'''Dates:''' June 26-30, 2017. Consider staying for one more day (leaving Sunday morning), as a day off in a gorgeous sea place is planned on July 1st. More details in the calendar below.
*'''Location:''' [http://www.hotelperladelporto.it/en/home-page.aspx Perla del Porto Hotel]
**[mailto:prenotazioni@hotelperladelporto.it Booking]. Subject line: "Slicer Summer Project Week 2017".
***Special rates are:
****Single room, full bed, 79 € per night (1 person)
****Single room, queen bed 89 € per night (1 person)
****Double room, queen bed 99 € per night (2 people)
****Triple room, 110 € (3 people)
*'''Registration:''' To register please visit this [http://www.imagenglab.com/newsite/project-week page]
*'''Registration Fee:''' 320€ and it includes lunches, coffee breaks and airport connections
*'''Hotel:''' [http://www.hotelperladelporto.it/en/home-page.aspx Perla del Porto Hotel]. The closest airport is [http://www.lameziaairport.it/english/ Lamezia Terme Airport (IATA: SUF)].
*'''Transportation from Airport to Hotel:''' Your registration fee includes ground transportation [https://www.google.com/maps/dir/Lamezia+Terme+International+Airport,+Via+Aeroporto,+88046+Lamezia+Terme+CZ,+Italy/BEST+WESTERN+PLUS+Hotel+Perla+Del+Porto,+Via+Martiri+di+Cefalonia,+64,+88100+Catanzaro,+Italy/@38.868758,16.1564814,10z/data=!3m1!4b1!4m14!4m13!1m5!1m1!1s0x133fe15a3cbed47f:0x544ab120c3de78a6!2m2!1d16.2434017!2d38.9065845!1m5!1m1!1s0x134003d668252a13:0x2989caf676f45a72!2m2!1d16.6312407!2d38.827712!3e0 to/from the hotel and airport].
** Please fill out this [https://goo.gl/forms/7vmhxZSHy8Z1A62z2 form] to request transportation
*'''Local points of interest (pubs, restaurants, bar):''' [https://www.google.com/maps/d/viewer?mid=1FU63ik9Do3zzP6K2kvLVTtM2at8&ll=38.86221979925013%2C16.44292274999998&z=12 map] (constantly updated)

==Calendar==
{{#widget:Google Calendar
|id=kitware.com_sb07i171olac9aavh46ir495c4@group.calendar.google.com
|timezone=America/New_York&dates=20170108%2F20170114
|title=NA-MIC Project Week (Timezone is Italy / GMT+02.)
|view=WEEK
|dates=20170626/20170701
}}

iCal (.ics) link: https://calendar.google.com/calendar/ical/kitware.com_sb07i171olac9aavh46ir495c4%40group.calendar.google.com/public/basic.ics

=Projects=

<big>Please duplicate [https://na-mic.org/wiki/Project_Week_Template this template] to create a page for your project. </big>

Please put a brief preliminary title for your project here with some names in parenthesis for potential team members

#[[Project_Week_25/NeedleSegmentation | Needle Segmentation]]
#[[Project_Week_25/Human-Computer_Interaction_under_sterile_conditions |Human-Computer Interaction under Sterile Conditions]]
# [[Project_Week_25/Next_Generation_GPU_Volume_Rendering | Next Generation of Volume Rendering in VTK ]]
# [[Project_Week_25/Tracked-Ultrasound-Standardization-IV | Tracked Ultrasound Standardization IV: Controlling US Acquisition]]
#[[Project_Week_25/Intra-operative deformable_registration_based_on_dense_point_cloud_reconstruction |Intra-operative Deformable Registration Based on Dense Point Cloud Reconstruction]]
#[[Project_Week_25/Segmentation for improving image registration of preoperative MRI with intraoperative ultrasound images for neuro-navigation |Segmentation for Improving Image Registration of Preoperative MRI with Intraoperative Ultrasound Images for Neuro-navigation]]
#[[Project_Week_25/CNN_for_multi-plane_prostate_segmentation | CNN for multi-plane prostate segmentation]]
#[[Project_Week_25/Conversion of DICOM Single Frame MR to Enhanced Multiframe | Conversion of DICOM Single Frame MR to Enhanced Multiframe]]
#[[Project_Week_25/Development_and_Evaluation_of_New_AR_Visualization_Techniques_to_Support_Radiological_Interventions | Development and Evaluation of New AR Visualization Techniques to Support Radiological Interventions]]
#[[Project_Week_25/Interactive_Manipulation_of_Plots_and_Graphs | Interactive Manipulation of Plots and Graphs]]
#[[Project_Week_25/Steerable Catheters Path Planner Extension for Brain Surgery Applications | Steerable Catheters Path Planner Extension for Brain Surgery Applications]]
#[[Project_Week_25/Improving Depth Perception in Interventional Augmented Reality Visualization/Sonification | Improving Depth Perception in Interventional Augmented Reality Visualization/Sonification]]
#[[Project_Week_25/DICOM_for_Quantitative_Imaging_and_Integration_with_Processing_Applications|DICOM for Quantitative Imaging and Integration with Processing Applications]]
#[[Project_Week_25/CNN for PseudoCT Generation from T1T2 MR|CNN for PseudoCT Generation from T1/T2 MRI]]
#[[Project_Week_25/Internationalizing Slicer Modules|Internationalizing Slicer Modules]]
#[[Project_Week_25/Interfacing Slicer to Mobile Phone-controlled Sensors|Interfacing Slicer to Mobile Phone-controlled Sensors]]
#[[Project_Week_25/Slicer and 3D Printing|Slicer and 3D Printing]]
#[[Project_Week_25/Slice-to-Volume Registration to Support MRI guided Interventions | Slice-to-Volume Registration to Support MRI guided Interventions ]]
#[[Project_Week_25/SALT_Spatiotemporal_Modeling: | Slicer SALT Validation: Spatiotemporal Modeling of Subcortical Structures ]]
#[[Project_Week_25/Wrist_Kinematics: | Kinematic Analysis of the Wrist from Dynamic MRI]]
#[[Project_Week_25/SlicerCIP_Bronchiectasis | SlicerCIP: Tool for Quantitative Analysis of Bronchiectasis]]
#[[Project_Week_25/SlicerCIP_ReportsTool | SlicerCIP: Tool for Creation of Reports for Quantitive Analysis]]
#[[Project_Week_25/Multimodal: | Multimodal, Multiresolution, Multivolume Data]]
#[[Project_Week_25/DICOM Segmentation Support for Cornerstone and OHIF Viewer | DICOM Segmentation Support for Cornerstone/ OHIF Viewer]]
#Slicer Export to VR (Juan Ruiz Alzola, Mike Halle)
#[[Project_Week_25/Breast Cancer Analysis in DCE-MRI | Breast Cancer Segmentation in DCE-MRI via Deep Learning Approaches (Gabriele Piantadosi)]]

=Registrants=

Do not add your name to this list - it is maintained by the organizers based on your paid registration. To register, visit this [http://www.imagenglab.com/newsite/project-week/ registration site].

# Kikinis, Ron :: Brigham and Women's Hospital, Harvard Medical School, USA
# Pieper, Steve :: Isomics, Inc., USA
# Kapur, Tina :: Brigham and Women's Hospital, Harvard Medical School, USA
# Spadea, Maria Francesca :: Magna Graecia University, Italy
# Zaffino, Paolo :: Magna Graecia University, Italy
# Scaramuzzino, Salvatore :: Magna Graecia University/ASL Vercelli, Italy
# Pileggi, Giampaolo :: Magna Graecia University, Italy/German Cancer Research Center (DKFZ), Germany
# Rackerseder, Julia :: Technical University of Munich, Germany
# Pinter, Csaba :: Queen's University, Canada
# Kraß, Scheherazade :: University of Bremen, Germany
# Gerig, Guido :: NYU Tandon School of Engineering, USA
# Punzo, Davide :: Kapteyn Astronomical Institute, University of Groningen, The Netherlands
# Drouin, Simon :: NeuroImaging and Surgical Technologies (NIST) Lab, Canada
# Lasso, Andras :: School of Computing, Queen's University, Canada
# Favaro, Alberto :: Politecnico di Milano, Italy
# Leger, Etienne :: Concordia University, Canada
# Ziegler, Erik :: Ziegler Consult SAS
# Onken, Michael :: Open Connections GmbH, Germany
# Pinzi, Marlene :: Imperial College, UK
# Nitsch, Jennifer :: University of Bremen, Germany
# Moccia, Sara :: Politecnico di Milano, Italy
# Black, David :: Fraunhofer Institute for Medical Image Computing MEVIS, Bremen, Germany
# Penzkofer, Tobias :: Charité Universitätsmedizin, Berlin, Germany
# Hansen, Christian :: University of Magdeburg, Germany
# Vegard Solberg, Ole :: Norway
# Heinrich, Florian :: University of Magdeburg, Germany
# Mewes, André :: University of Magdeburg, Germany
# Hatscher, Benjamin :: University of Magdeburg, Germany
# Hettig, Julian :: University of Magdeburg, Germany
# Meyer, Anneke :: University of Magdeburg, Germany
# Gulamhussene, Gino :: University of Magdeburg, Germany
# Cassetta, Roberto :: Politecnico di Milano, Italy
# Fillion-Robin, Jean-Christophe :: Kitware, Inc., USA
# Metzger, Jasmin :: German Cancer Research Center (DKFZ), Germany
# Fishbaugh, James :: NYU Tandon School of Engineering, USA
# Nolden, Marco :: German Cancer Research Center (DKFZ), Germany
# Nehrkorn, Jorge Quintero :: Canary Islands, Spain
# Perez Garcia, Fernando :: ICM Brain & Spine Institute, Paris, France
# De Momi, Elena :: Politecnico di Milano, Italy
# Piantadosi, Gabriele :: DIETI, Federico II di Napoli, Italy
# Pernelle, Guillaume :: Imperial College, UK
# San Jose, Raul :: Brigham and Women's Hospital, Harvard Medical School, USA
# Nardelli, Pietro :: Brigham and Women's Hospital, Harvard Medical School, USA
# Fernandez Vidal, Sara :: ICM Brain & Spine Institute, Paris, France
# Sharp, Gregory :: Massachusetts General Hospital, Harvard Medical School, USA
# Moiraghi, Alessandro :: Università degli Studi di Milano, Italy
# Seco, Joao :: German Cancer Research Center (DKFZ), Germany
# Pumar Carreras, Nayra :: Canary Islands, Spain
# Afonso Suarez, Maria Dolores :: Canary Islands, Spain
# Alzola Ruiz, Juan :: Canary Islands, Spain
# Pujol, Sonia :: Brigham and Women's Hospital, Harvard Medical School, USA

Project Week 25

2017-06-24T07:15:29Z

GabrielePiantadosi: add a project

__NOTOC__
Back to [[Events]]

A summary of all past [[Project_Events#Past_Project_Weeks|Project Events]].

[[image:PW25.png|300px]] [[image:IEL_logo.png|225px]]
<br>

=Welcome to the web page for the 25th Project Week!=
It is a pleasure to announce that the 25th Project week will be held in [https://goo.gl/maps/b9CpkFxNyWN2 Catanzaro Lido] (Calabria, Italy) on June 26-30, 2017. This is the first time in Italy for the Slicer Community, and the event is organized in cooperation with [http://www.imagenglab.com ImagEngLab]. Catanzaro Lido is a city on the Ionian Sea, in the middle of Squillace Gulf where, according to the ancient legend, Odysseus started his journey back to Ithaca. Of course bring your swimsuit...the conference room and the hotel are 20 meters far away from the beach!

This project week is an event [[Post-NCBC-2014|endorsed]] by the MICCAI society.

Please make sure that you are on the NA-MIC Project Week [http://public.kitware.com/mailman/listinfo/na-mic-project-week mailing list].

===Local Organizing Committee===
*[http://www.imagenglab.com/newsite/mf_spadea/ Maria Francesca Spadea, PhD]
*[http://www.imagenglab.com/newsite/paolo_zaffino/ Paolo Zaffino, PhD].

==Videoconferences for preparation==
Time: Wednesday 9am EDT (GMT -4), April 15 to June 21, 2017
URL: [https://zoom.us/j/725310144 click here] to join the videoconference (zoom is the tool as of May 31)

#(Tina Kapur) Hangout #1: April 5 ([[PW25 Hangouts Notes|Notes]])
#(Steve Pieper) Hangout #2: April 12: Web browser based computing: Dockerized Slicer with remote computing and GPU computing; Cornerstone/LesionTracker OHIF; XTK->AMI (threejs); ePad; vtk.js; QWebEngine;
#(Andras Lasso) Hangout #3: April 19: Connecting devices such as surgical navigation, ultrasound, 3D Slicer, PLUS, OpenIGTLink, Augmented reality;
#(Tina Kapur) Hangout #4: April 26: Deep Learning for Detection of Cancer and Instruments;
#(Simon Drouin) Hangout #5: May 3: Volume Rendering, Augmented Reality, and Virtual Reality. [https://docs.google.com/document/d/1UwdSzjnDm1yEeQ44OEhXWbH6V83Uo1Cd4KngxoyrRdI/edit Notes];
#(Tina Kapur) Hangout #6: May 10: For new participants: What is project week and how to get the most out of participating in it?
#(Andrey Fedorov) Hangout #7: May 17:DICOM for Quantitative Imaging and integration with processing applications. ([http://bit.ly/2017NSPW-DICOM notes])
#(Tina Kapur) Hangout #8: May 24:Discussion of Projects and teams that have been provided on the wiki page by participants, internationalization strategy for Slicer.
#(Francesca Spadea) Hangout #9: May 31: Review of local logistics -- all registered attendees should join
#(Tina Kapur) Hangout #10: June 7: Discussion of Projects, Project Pages
#(Francesca Spadea, Tina Kapur) Hangout #11: June 14: Review of Project Pages and Logistics
#(Francesca Spadea) Hangout #12: June 21: Review of local logistics -- all registered attendees with questions should join

==Logistics==
*'''Dates:''' June 26-30, 2017. Consider staying for one more day (leaving Sunday morning), as a day off in a gorgeous sea place is planned on July 1st. More details in the calendar below.
*'''Location:''' [http://www.hotelperladelporto.it/en/home-page.aspx Perla del Porto Hotel]
**[mailto:prenotazioni@hotelperladelporto.it Booking]. Subject line: "Slicer Summer Project Week 2017".
***Special rates are:
****Single room, full bed, 79 € per night (1 person)
****Single room, queen bed 89 € per night (1 person)
****Double room, queen bed 99 € per night (2 people)
****Triple room, 110 € (3 people)
*'''Registration:''' To register please visit this [http://www.imagenglab.com/newsite/project-week page]
*'''Registration Fee:''' 320€ and it includes lunches, coffee breaks and airport connections
*'''Hotel:''' [http://www.hotelperladelporto.it/en/home-page.aspx Perla del Porto Hotel]. The closest airport is [http://www.lameziaairport.it/english/ Lamezia Terme Airport (IATA: SUF)].
*'''Transportation from Airport to Hotel:''' Your registration fee includes ground transportation [https://www.google.com/maps/dir/Lamezia+Terme+International+Airport,+Via+Aeroporto,+88046+Lamezia+Terme+CZ,+Italy/BEST+WESTERN+PLUS+Hotel+Perla+Del+Porto,+Via+Martiri+di+Cefalonia,+64,+88100+Catanzaro,+Italy/@38.868758,16.1564814,10z/data=!3m1!4b1!4m14!4m13!1m5!1m1!1s0x133fe15a3cbed47f:0x544ab120c3de78a6!2m2!1d16.2434017!2d38.9065845!1m5!1m1!1s0x134003d668252a13:0x2989caf676f45a72!2m2!1d16.6312407!2d38.827712!3e0 to/from the hotel and airport].
** Please fill out this [https://goo.gl/forms/7vmhxZSHy8Z1A62z2 form] to request transportation
*'''Local points of interest (pubs, restaurants, bar):''' [https://www.google.com/maps/d/viewer?mid=1FU63ik9Do3zzP6K2kvLVTtM2at8&ll=38.86221979925013%2C16.44292274999998&z=12 map] (constantly updated)

==Calendar==
{{#widget:Google Calendar
|id=kitware.com_sb07i171olac9aavh46ir495c4@group.calendar.google.com
|timezone=America/New_York&dates=20170108%2F20170114
|title=NA-MIC Project Week (Timezone is Italy / GMT+02.)
|view=WEEK
|dates=20170626/20170701
}}

iCal (.ics) link: https://calendar.google.com/calendar/ical/kitware.com_sb07i171olac9aavh46ir495c4%40group.calendar.google.com/public/basic.ics

=Projects=

<big>Please duplicate [https://na-mic.org/wiki/Project_Week_Template this template] to create a page for your project. </big>

Please put a brief preliminary title for your project here with some names in parenthesis for potential team members

#[[Project_Week_25/NeedleSegmentation | Needle Segmentation]]
#[[Project_Week_25/Human-Computer_Interaction_under_sterile_conditions |Human-Computer Interaction under Sterile Conditions]]
# [[Project_Week_25/Next_Generation_GPU_Volume_Rendering | Next Generation of Volume Rendering in VTK ]]
# [[Project_Week_25/Tracked-Ultrasound-Standardization-IV | Tracked Ultrasound Standardization IV: Controlling US Acquisition]]
#[[Project_Week_25/Intra-operative deformable_registration_based_on_dense_point_cloud_reconstruction |Intra-operative Deformable Registration Based on Dense Point Cloud Reconstruction]]
#[[Project_Week_25/Segmentation for improving image registration of preoperative MRI with intraoperative ultrasound images for neuro-navigation |Segmentation for Improving Image Registration of Preoperative MRI with Intraoperative Ultrasound Images for Neuro-navigation]]
#[[Project_Week_25/CNN_for_multi-plane_prostate_segmentation | CNN for multi-plane prostate segmentation]]
#[[Project_Week_25/Conversion of DICOM Single Frame MR to Enhanced Multiframe | Conversion of DICOM Single Frame MR to Enhanced Multiframe]]
#[[Project_Week_25/Development_and_Evaluation_of_New_AR_Visualization_Techniques_to_Support_Radiological_Interventions | Development and Evaluation of New AR Visualization Techniques to Support Radiological Interventions]]
#[[Project_Week_25/Interactive_Manipulation_of_Plots_and_Graphs | Interactive Manipulation of Plots and Graphs]]
#[[Project_Week_25/Steerable Catheters Path Planner Extension for Brain Surgery Applications | Steerable Catheters Path Planner Extension for Brain Surgery Applications]]
#[[Project_Week_25/Improving Depth Perception in Interventional Augmented Reality Visualization/Sonification | Improving Depth Perception in Interventional Augmented Reality Visualization/Sonification]]
#[[Project_Week_25/DICOM_for_Quantitative_Imaging_and_Integration_with_Processing_Applications|DICOM for Quantitative Imaging and Integration with Processing Applications]]
#[[Project_Week_25/CNN for PseudoCT Generation from T1T2 MR|CNN for PseudoCT Generation from T1/T2 MRI]]
#[[Project_Week_25/Internationalizing Slicer Modules|Internationalizing Slicer Modules]]
#[[Project_Week_25/Interfacing Slicer to Mobile Phone-controlled Sensors|Interfacing Slicer to Mobile Phone-controlled Sensors]]
#[[Project_Week_25/Slicer and 3D Printing|Slicer and 3D Printing]]
#[[Project_Week_25/Slice-to-Volume Registration to Support MRI guided Interventions | Slice-to-Volume Registration to Support MRI guided Interventions ]]
#[[Project_Week_25/SALT_Spatiotemporal_Modeling: | Slicer SALT Validation: Spatiotemporal Modeling of Subcortical Structures ]]
#[[Project_Week_25/Wrist_Kinematics: | Kinematic Analysis of the Wrist from Dynamic MRI]]
#[[Project_Week_25/SlicerCIP_Bronchiectasis | SlicerCIP: Tool for Quantitative Analysis of Bronchiectasis]]
#[[Project_Week_25/SlicerCIP_ReportsTool | SlicerCIP: Tool for Creation of Reports for Quantitive Analysis]]
#[[Project_Week_25/Multimodal: | Multimodal, Multiresolution, Multivolume Data]]
#[[Project_Week_25/DICOM Segmentation Support for Cornerstone and OHIF Viewer | DICOM Segmentation Support for Cornerstone/ OHIF Viewer]]
#Slicer Export to VR (Juan Ruiz Alzola, Mike Halle)
#Breast Cancer segmentation in DCE-MRI via Deep Learning approaches (Gabriele Piantadosi)

=Registrants=

Do not add your name to this list - it is maintained by the organizers based on your paid registration. To register, visit this [http://www.imagenglab.com/newsite/project-week/ registration site].

# Kikinis, Ron :: Brigham and Women's Hospital, Harvard Medical School, USA
# Pieper, Steve :: Isomics, Inc., USA
# Kapur, Tina :: Brigham and Women's Hospital, Harvard Medical School, USA
# Spadea, Maria Francesca :: Magna Graecia University, Italy
# Zaffino, Paolo :: Magna Graecia University, Italy
# Scaramuzzino, Salvatore :: Magna Graecia University/ASL Vercelli, Italy
# Pileggi, Giampaolo :: Magna Graecia University, Italy/German Cancer Research Center (DKFZ), Germany
# Rackerseder, Julia :: Technical University of Munich, Germany
# Pinter, Csaba :: Queen's University, Canada
# Kraß, Scheherazade :: University of Bremen, Germany
# Gerig, Guido :: NYU Tandon School of Engineering, USA
# Punzo, Davide :: Kapteyn Astronomical Institute, University of Groningen, The Netherlands
# Drouin, Simon :: NeuroImaging and Surgical Technologies (NIST) Lab, Canada
# Lasso, Andras :: School of Computing, Queen's University, Canada
# Favaro, Alberto :: Politecnico di Milano, Italy
# Leger, Etienne :: Concordia University, Canada
# Ziegler, Erik :: Ziegler Consult SAS
# Onken, Michael :: Open Connections GmbH, Germany
# Pinzi, Marlene :: Imperial College, UK
# Nitsch, Jennifer :: University of Bremen, Germany
# Moccia, Sara :: Politecnico di Milano, Italy
# Black, David :: Fraunhofer Institute for Medical Image Computing MEVIS, Bremen, Germany
# Penzkofer, Tobias :: Charité Universitätsmedizin, Berlin, Germany
# Hansen, Christian :: University of Magdeburg, Germany
# Vegard Solberg, Ole :: Norway
# Heinrich, Florian :: University of Magdeburg, Germany
# Mewes, André :: University of Magdeburg, Germany
# Hatscher, Benjamin :: University of Magdeburg, Germany
# Hettig, Julian :: University of Magdeburg, Germany
# Meyer, Anneke :: University of Magdeburg, Germany
# Gulamhussene, Gino :: University of Magdeburg, Germany
# Cassetta, Roberto :: Politecnico di Milano, Italy
# Fillion-Robin, Jean-Christophe :: Kitware, Inc., USA
# Metzger, Jasmin :: German Cancer Research Center (DKFZ), Germany
# Fishbaugh, James :: NYU Tandon School of Engineering, USA
# Nolden, Marco :: German Cancer Research Center (DKFZ), Germany
# Nehrkorn, Jorge Quintero :: Canary Islands, Spain
# Perez Garcia, Fernando :: ICM Brain & Spine Institute, Paris, France
# De Momi, Elena :: Politecnico di Milano, Italy
# Piantadosi, Gabriele :: DIETI, Federico II di Napoli, Italy
# Pernelle, Guillaume :: Imperial College, UK
# San Jose, Raul :: Brigham and Women's Hospital, Harvard Medical School, USA
# Nardelli, Pietro :: Brigham and Women's Hospital, Harvard Medical School, USA
# Fernandez Vidal, Sara :: ICM Brain & Spine Institute, Paris, France
# Sharp, Gregory :: Massachusetts General Hospital, Harvard Medical School, USA
# Moiraghi, Alessandro :: Università degli Studi di Milano, Italy
# Seco, Joao :: German Cancer Research Center (DKFZ), Germany
# Pumar Carreras, Nayra :: Canary Islands, Spain
# Afonso Suarez, Maria Dolores :: Canary Islands, Spain
# Alzola Ruiz, Juan :: Canary Islands, Spain
# Pujol, Sonia :: Brigham and Women's Hospital, Harvard Medical School, USA