Exploring Structure-Wise Uncertainty for 3D Medical Image Segmentation
This addresses the need for more precise and semantic-aware uncertainty estimation in medical imaging, particularly for segmenting distinct structures like tumors, though it appears incremental as it builds on existing uncertainty methods.
The paper tackled the problem of estimating uncertainty for individual structures in 3D medical image segmentation, proposing a framework that identifies the best uncertainty estimation method to improve segmentation quality, tested on datasets including LIDC-IDRI, LiTS, and a private brain metastases dataset.
When applying a Deep Learning model to medical images, it is crucial to estimate the model uncertainty. Voxel-wise uncertainty is a useful visual marker for human experts and could be used to improve the model's voxel-wise output, such as segmentation. Moreover, uncertainty provides a solid foundation for out-of-distribution (OOD) detection, improving the model performance on the image-wise level. However, one of the frequent tasks in medical imaging is the segmentation of distinct, local structures such as tumors or lesions. Here, the structure-wise uncertainty allows more precise operations than image-wise and more semantic-aware than voxel-wise. The way to produce uncertainty for individual structures remains poorly explored. We propose a framework to measure the structure-wise uncertainty and evaluate the impact of OOD data on the model performance. Thus, we identify the best UE method to improve the segmentation quality. The proposed framework is tested on three datasets with the tumor segmentation task: LIDC-IDRI, LiTS, and a private one with multiple brain metastases cases.