Sample Efficient Learning of Image-Based Diagnostic Classifiers Using Probabilistic Labels
This addresses the challenge of limited training data in sensitive medical diagnosis tasks, providing a method to improve accuracy and calibration, though it is incremental as it builds on existing probabilistic labeling techniques.
The paper tackled the problem of training accurate and calibrated deep networks for image-based medical diagnosis with small datasets by using probabilistic labels, achieving up to 22% accuracy gains in tasks like hip dysplasia, fatty liver, and glaucoma classification.
Deep learning approaches often require huge datasets to achieve good generalization. This complicates its use in tasks like image-based medical diagnosis, where the small training datasets are usually insufficient to learn appropriate data representations. For such sensitive tasks it is also important to provide the confidence in the predictions. Here, we propose a way to learn and use probabilistic labels to train accurate and calibrated deep networks from relatively small datasets. We observe gains of up to 22% in the accuracy of models trained with these labels, as compared with traditional approaches, in three classification tasks: diagnosis of hip dysplasia, fatty liver, and glaucoma. The outputs of models trained with probabilistic labels are calibrated, allowing the interpretation of its predictions as proper probabilities. We anticipate this approach will apply to other tasks where few training instances are available and expert knowledge can be encoded as probabilities.