Class Activation Map-based Weakly supervised Hemorrhage Segmentation using Resnet-LSTM in Non-Contrast Computed Tomography images

TL;DR

This paper introduces a weakly supervised deep learning approach for intracranial hemorrhage segmentation in non-contrast CT scans, utilizing class activation maps and pseudo-masks to achieve competitive results with less manual labeling.

Contribution

The novel method leverages image-level labels and class activation maps for hemorrhage segmentation, reducing the need for detailed annotations and improving efficiency.

Findings

Achieved a Dice score of 0.55 on validation data

Comparable performance to existing weakly supervised methods

Effective segmentation with less training data

Abstract

In clinical settings, intracranial hemorrhages (ICH) are routinely diagnosed using non-contrast CT (NCCT) for severity assessment. Accurate automated segmentation of ICH lesions is the initial and essential step, immensely useful for such assessment. However, compared to other structural imaging modalities such as MRI, in NCCT images ICH appears with very low contrast and poor SNR. Over recent years, deep learning (DL)-based methods have shown great potential, however, training them requires a huge amount of manually annotated lesion-level labels, with sufficient diversity to capture the characteristics of ICH. In this work, we propose a novel weakly supervised DL method for ICH segmentation on NCCT scans, using image-level binary classification labels, which are less time-consuming and labor-efficient when compared to the manual labeling of individual ICH lesions. Our method initially determines the approximate location of ICH using class activation maps from a classification network, which is trained to learn dependencies across contiguous slices. We further refine the ICH segmentation using pseudo-ICH masks obtained in an unsupervised manner. The method is flexible and uses a computationally light architecture during testing. On evaluating our method on the validation data of the MICCAI 2022 INSTANCE challenge, our method achieves a Dice value of 0.55, comparable with those of existing weakly supervised method (Dice value of 0.47), despite training on a much smaller training data.