Improving Explainability of Disentangled Representations using Multipath-Attribution Mappings

Lukas Klein, João B. S. Carvalho, Mennatallah El-Assady, Paolo Penna, Joachim M. Buhmann, Paul F Jaeger

Keywords: Explanatory AI, Disentangled Representations, Shortcut Detection, Medical Imaging

TL;DR: Novel framework combining disentangled representations with multipath-attribution, yielding enhanced interpretability and generalization on medical datasets.

Abstract: Explainable AI aims to render model behavior understandable by humans, which can be seen as an intermediate step in extracting causal relations from correlative patterns. Due to the high risk of possible fatal decisions in image-based clinical diagnostics, it is necessary to integrate explainable AI into these safety-critical systems. Current explanatory methods typically assign attribution scores to pixel regions in the input image, indicating their importance for a model's decision. However, they fall short when explaining why a visual feature is used. We propose a framework that utilizes interpretable disentangled representations for downstream-task prediction. Through visualizing the disentangled representations, we enable experts to investigate possible causation effects by leveraging their domain knowledge. Additionally, we deploy a multi-path attribution mapping for enriching and validating explanations. We demonstrate the effectiveness of our approach on a synthetic benchmark suite and two medical datasets. We show that the framework not only acts as a catalyst for causal relation extraction but also enhances model robustness by enabling shortcut detection without the need for testing under distribution shifts. Code available at https://github.com/IML-DKFZ/m-pax_lib.

Registration: I acknowledge that publication of this at MIDL and in the proceedings requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: both

Primary Subject Area: Interpretability and Explainable AI

Secondary Subject Area: Unsupervised Learning and Representation Learning

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.

Code And Data:

Code: https://github.com/IML-DKFZ/m-pax_lib

Data:
MNIST: http://yann.lecun.com/exdb/mnist/
DiagViB-6: https://github.com/boschresearch/diagvib-6
OCT Retina: https://data.mendeley.com/datasets/rscbjbr9sj/3
ISIC 2019: https://challenge.isic-archive.com/data/#2019