Skip to content Skip to navigation

PhD student Chaowei Tan and colleagues receive Best Student Paper award at ISBI 2018

Friday, April 6, 2018

Chaowei Tan, a PhD student in the Computer Science department, advised by Prof. Dimitri Metaxas, has been awarded a Best Student Paper award at the latest International Symposium on Biomedical Imaging (ISBI) in DC for the paper "Deep Multi-task and Task-specific Feature Learning Network for Robust Shape Preserved Organ Segmentation"  (authors: Chaowei Tan, Liang Zhao, Zhennan Yan, Kang Li, Dimitris Metaxas, and Yiqiang Zhan).

The IEEE International Symposium on Biomedical Imaging (ISBI) is a scientific conference dedicated to mathematical, algorithmic, and computational aspects of biological and biomedical imaging, across all scales of observation. Three Best Student Paper awards were offered during its latest iteration and the criteria were identified as follows:

1. novelty of medical image analysis;
2. the work should present novel contributions in methodology, and can be employed to solve basic and important problems in medical image analysis;
3. the experimental settings and results are complete and persuasive.

In particular, the awarded paper proposes a deep end-to-end network with two task specific branches to ensure continuousness, smoothness and shape-preservation in segmented structure without additionally sophisticated shape adjustment. The novelty of the proposed method lies in three aspects. First, it provides a formulation of the organ segmentation as a multi-task learning process that combines both region and boundary identification tasks, which can alleviate spatially isolated segmentation errors. Second, it introduces a boundary distance regression to ensure the smoothness of the segmented contours, instead of formulating boundary identification as a classification problem. Third, the deep network is designed to have a “Y” shape, i.e., the first half of the network is shared by both region and boundary identification tasks, while the second half is branched for each task independently. This architecture enables the task-specific feature learning for better region and boundary identification, and offers information for segmentation refinement based on a fusion scheme using energy functional.

Referenced Faculty: