Weakly Supervised Learning Significantly Reduces the Number of Labels Required for Intracranial Hemorrhage Detection on Head CT

TL;DR

This study demonstrates that weakly supervised learning using only examination-level labels can achieve comparable intracranial hemorrhage detection performance to strongly supervised methods requiring detailed image-level annotations, significantly reducing labeling effort.

Contribution

The paper shows that weak supervision with examination-level labels suffices for effective hemorrhage detection, challenging the need for detailed image annotations in medical imaging.

Findings

Weak supervision achieves similar detection performance as strong supervision.

Local image-level labels may not be necessary for accurate hemorrhage detection.

Reducing label granularity decreases data annotation costs significantly.

Abstract

Modern machine learning pipelines, in particular those based on deep learning (DL) models, require large amounts of labeled data. For classification problems, the most common learning paradigm consists of presenting labeled examples during training, thus providing strong supervision on what constitutes positive and negative samples. This constitutes a major obstacle for the development of DL models in radiology--in particular for cross-sectional imaging (e.g., computed tomography [CT] scans)--where labels must come from manual annotations by expert radiologists at the image or slice-level. These differ from examination-level annotations, which are coarser but cheaper, and could be extracted from radiology reports using natural language processing techniques. This work studies the question of what kind of labels should be collected for the problem of intracranial hemorrhage detection in brain CT. We investigate whether image-level annotations should be preferred to examination-level ones. By framing this task as a multiple instance learning problem, and employing modern attention-based DL architectures, we analyze the degree to which different levels of supervision improve detection performance. We find that strong supervision (i.e., learning with local image-level annotations) and weak supervision (i.e., learning with only global examination-level labels) achieve comparable performance in examination-level hemorrhage detection (the task of selecting the images in an examination that show signs of hemorrhage) as well as in image-level hemorrhage detection (highlighting those signs within the selected images). Furthermore, we study this behavior as a function of the number of labels available during training. Our results suggest that local labels may not be necessary at all for these tasks, drastically reducing the time and cost involved in collecting and curating datasets.