Projects:DTI DWI QualityControl

From NAMIC Wiki
Jump to: navigation, search
Home < Projects:DTI DWI QualityControl
Back to UNC Algorithms


Diffusion Tensor and Diffusion Weighted Imaging Quality Control

DWIs data suffer from inherent low SNR, overall long scanning time of multiple directional encoding with correspondingly large risk to encounter several kinds of artifacts. These artifacts can be too severe for a correct and stable estimation of the diffusion tensor. Thus, a quality control (QC) procedure is absolutely necessary for DTI studies. We are developing a framework, called DTIPrep for assessing and correcting DWIs and DTI.

Description

DTIPrep as the first comprehensive and fully automatic pre-processing tool for DWI and DTI quality control can provide a crucial piece for robust DTI analysis studies. The protocoling, reporting, visual controlling and data correction capabilities are used to produce high consistence and inter-rater reliable QC results. This framework is organized by pipeline steps include: 1) image info checking, 2) diffusion info checking, 3) Slice-wise intensity checking, 4) Interlace-wise intensity checking, 5) Averaging baselines, 6) Eddy-motion correction, 7) gradient-wise checking, 8) Computing DTI measurements and saving.


alt Diffusion Weighted Imaging and Diffusion Tensor Imaging Quality Control_DTIPrep
alt 3D view of gradients before and after Quality Control procedures

As further extended-DTIprep, our experiments show that the residual artifacts presence after DTIPrep DWIs QC can be detected and corrected using knowledge of DTI. We introduce new approach in DTI QC including the detection and the corrections steps. Our approach use new entropy-based benchmark comes up with Principal Directions (PDs) histogram implemented by multi-level subdivision icosahedron within different regions of brain. Given training of the measurement, the quality of DTI is categorized into acceptable, suspicious and unacceptable groups using calculated the standard scores.

File:Diagram2.pdf

We employ the correction step by excluding gradients which have the most contribution in the artifacts. We continue ex- cluding gradients till the z-score of whole brain of updated image will be close enough to the trained

Publications

Key Investigators

  • UNC Algorithms: Mahshid Farzinfar, Zhexing Liu, Martin Styner, Clement Vachet
  • Utah Algorithms: Tom Fletcher, Ross Whitaker, Guido Gerig, Sylvain Gouttard

Links