Brain Metastasis Segmentation Network Trained with Robustness to Annotations with Multiple False Negatives
This work addresses the challenge of labor-intensive expert annotation in medical imaging by enabling more efficient scaling of labeling processes, though it is incremental as it builds on existing robustness techniques.
The paper tackles the problem of training deep learning models for brain metastasis segmentation with noisy annotations containing false negatives, and achieves a sensitivity of 97% of baseline with a 50% false negative rate using a novel lopsided loss function, compared to 10% for standard methods.
Deep learning has proven to be an essential tool for medical image analysis. However, the need for accurately labeled input data, often requiring time- and labor-intensive annotation by experts, is a major limitation to the use of deep learning. One solution to this challenge is to allow for use of coarse or noisy labels, which could permit more efficient and scalable labeling of images. In this work, we develop a lopsided loss function based on entropy regularization that assumes the existence of a nontrivial false negative rate in the target annotations. Starting with a carefully annotated brain metastasis lesion dataset, we simulate data with false negatives by (1) randomly censoring the annotated lesions and (2) systematically censoring the smallest lesions. The latter better models true physician error because smaller lesions are harder to notice than the larger ones. Even with a simulated false negative rate as high as 50%, applying our loss function to randomly censored data preserves maximum sensitivity at 97% of the baseline with uncensored training data, compared to just 10% for a standard loss function. For the size-based censorship, performance is restored from 17% with the current standard to 88% with our lopsided bootstrap loss. Our work will enable more efficient scaling of the image labeling process, in parallel with other approaches on creating more efficient user interfaces and tools for annotation.