Weakly supervised semantic segmentation of tomographic images in the diagnosis of stroke

TL;DR

This paper introduces a weakly supervised segmentation method for stroke-affected brain regions in CT images, effectively handling inaccurate labels and improving segmentation accuracy with a modified U-Net architecture.

Contribution

It proposes novel techniques for segmentation with inaccurately labeled data in a weakly supervised setting, enhancing U-Net performance for stroke diagnosis.

Findings

Improved segmentation accuracy on real CT scans.

Effective handling of inaccurate radiologist labels.

Enhanced U-Net architecture for medical image segmentation.

Abstract

This paper presents an automatic algorithm for the segmentation of areas affected by an acute stroke on the non-contrast computed tomography brain images. The proposed algorithm is designed for learning in a weakly supervised scenario when some images are labeled accurately, and some images are labeled inaccurately. Wrong labels appear as a result of inaccuracy made by a radiologist in the process of manual annotation of computed tomography images. We propose methods for solving the segmentation problem in the case of inaccurately labeled training data. We use the U-Net neural network architecture with several modifications. Experiments on real computed tomography scans show that the proposed methods increase the segmentation accuracy.