Weakly supervised deep learning-based intracranial hemorrhage localization

TL;DR

This paper introduces a weakly supervised deep learning method for precise intracranial hemorrhage localization in CT slices using only slice-level labels, enabling faster diagnosis with minimal annotation effort.

Contribution

It proposes a novel multiple instance learning approach for hemorrhage localization that does not require detailed annotations, improving efficiency and accuracy.

Findings

Achieved a Dice coefficient of 58.08% on public dataset

Generated hemorrhage likelihood maps and identified bleeding coordinates

Demonstrated effectiveness of weakly supervised localization

Abstract

Intracranial hemorrhage is a life-threatening disease, which requires fast medical intervention. Owing to the duration of data annotation, head CT images are usually available only with slice-level labeling. This paper presents a weakly supervised method of precise hemorrhage localization in axial slices using only position-free labels, which is based on multiple instance learning. An algorithm is introduced that generates hemorrhage likelihood maps and finds the coordinates of bleeding. The Dice coefficient of 58.08 % is achieved on data from a publicly available dataset.