Edit this page

ProjectWeek

Back to Projects List

Multi-Label Segmentation of Prostate Zones with Volumetric CNN

Key Investigators

Project Description

The goal of this project is to create and evaluate variants of a CNN for multi-label segmentation of prostate zones in MR images. The prostate zones are essential for lesion classification and therapy planning. After successful segmentation, a sector map could be extracted that is used for PI-RADS reporting. This has the potential to automate and better standardize prostate lesion location reporting.

Objective

  1. Overlap-free segmentations of prostate zones.
  2. Gap-free segmentations of prostate zones.
  3. Improvement of current segmentation result, especially for the anterior fibromuscular stroma (AFS)

Approach and Plan

  1. Apply variants of volumentric CNN architectures.
  2. Discuss ways to obtain overlap- and gap-free segmentations.
  3. Discuss methods to create sector map. Which landmarks should be used?

Progress and Next Steps

  1. first results on more training data and with different models look promising
  2. Obtained meaningful results for the AFS .
  3. disucussions with people how to further improve the outcome.

Illustrations

Segmentation Example

Segmentation Example 2

Background and References