Difference between revisions of "2017 Winter Project Week/OCM-MRI"

From NAMIC Wiki
Jump to: navigation, search
 
(4 intermediate revisions by the same user not shown)
Line 22: Line 22:
 
<!-- Approach and Plan bullet points -->
 
<!-- Approach and Plan bullet points -->
 
* Look into different software frameworks, choose one
 
* Look into different software frameworks, choose one
* Investigate generative deep neural networks techniques
+
* Investigate generative deep neural network techniques
 
* Produce preliminary results on hybrid US+MRI data
 
* Produce preliminary results on hybrid US+MRI data
 
|
 
|
 
<!-- Progress and Next steps bullet points (fill out at the end of project week) -->
 
<!-- Progress and Next steps bullet points (fill out at the end of project week) -->
*
+
* Discussions
 +
* Installation of and finding our way around Keras and TensorFlow
 +
* Network candidates: CNN with logistic regression output, Generative Adversarial Networks (GAN), Variational Autoencoders
 +
* In particular, Variational Autoencoders are able to model conditional distributions as required here
 
|}
 
|}
  

Latest revision as of 15:41, 13 January 2017

Home < 2017 Winter Project Week < OCM-MRI

Key Investigators

  • Frank Preiswerk, Brigham and Women's Hospital, Harvard Medical School
  • Yaofei Wang (Vivian), Tianjin University, Beijing, China

Project Description

Objective Approach and Plan Progress and Next Steps
  • To investigate the use of Deep Learning for the generation of synthetic respiratory MR images.
  • Look into different software frameworks, choose one
  • Investigate generative deep neural network techniques
  • Produce preliminary results on hybrid US+MRI data
  • Discussions
  • Installation of and finding our way around Keras and TensorFlow
  • Network candidates: CNN with logistic regression output, Generative Adversarial Networks (GAN), Variational Autoencoders
  • In particular, Variational Autoencoders are able to model conditional distributions as required here

Background and References