![[2310.15741] Interpretable Medical Image Classification using Prototype Learning and Privileged Information](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2310.15741] Interpretable Medical Image Classification using Prototype Learning and Privileged Information
Summary
This article presents a novel approach to medical image classification using prototype learning and privileged information, enhancing interpretability and accuracy in predictions.
Why It Matters
As medical imaging increasingly relies on AI, the need for interpretable models is critical for clinical trust and decision-making. This research introduces Proto-Caps, which improves both the understanding and performance of image classification, addressing a key challenge in the field.
Key Takeaways
- Proto-Caps combines capsule networks and prototype learning for enhanced interpretability.
- The model achieves over 93% accuracy in predicting malignancy in lung nodules.
- Case-based reasoning is facilitated through prototype representations, allowing for visual validation.
- Utilizing privileged information during training significantly boosts model performance.
- This approach sets a new benchmark for explainability in medical image classification.
Computer Science > Computer Vision and Pattern Recognition arXiv:2310.15741 (cs) [Submitted on 24 Oct 2023] Title:Interpretable Medical Image Classification using Prototype Learning and Privileged Information Authors:Luisa Gallee, Meinrad Beer, Michael Goetz View a PDF of the paper titled Interpretable Medical Image Classification using Prototype Learning and Privileged Information, by Luisa Gallee and 2 other authors View PDF Abstract:Interpretability is often an essential requirement in medical imaging. Advanced deep learning methods are required to address this need for explainability and high performance. In this work, we investigate whether additional information available during the training process can be used to create an understandable and powerful model. We propose an innovative solution called Proto-Caps that leverages the benefits of capsule networks, prototype learning and the use of privileged information. Evaluating the proposed solution on the LIDC-IDRI dataset shows that it combines increased interpretability with above state-of-the-art prediction performance. Compared to the explainable baseline model, our method achieves more than 6 % higher accuracy in predicting both malignancy (93.0 %) and mean characteristic features of lung nodules. Simultaneously, the model provides case-based reasoning with prototype representations that allow visual validation of radiologist-defined attributes. Comments: Subjects: Computer Vision and Pattern Recognition (cs.CV); ...