Interpretable machine learning for time-to-event prediction in medicine and healthcare

Abstract

Time-to-event prediction, e.g. cancer survival analysis or hospital length of stay, is a highly prominent machine learning task in medical and healthcare applications. However, only a few interpretable machine learning methods comply with its challenges. To facilitate a comprehensive explanatory analysis of survival models, we formally introduce time-dependent feature effects and global feature importance explanations. We show how post-hoc interpretation methods allow for finding biases in AI systems predicting length of stay using a novel multi-modal dataset created from 1235 X-ray images with textual radiology reports annotated by human experts. Moreover, we evaluate cancer survival models beyond predictive performance to include the importance of multi-omics feature groups based on a large-scale benchmark comprising 11 datasets from The Cancer Genome Atlas (TCGA). Model developers can use the proposed methods to debug and improve machine learning algorithms, while physicians can discover disease biomarkers and assess their significance. We hope the contributed open data and code resources facilitate future work in the emerging research direction of explainable survival analysis.

An extended version of an AIME 2023 paper submitted to Artificial Intelligence in Medicine

References in corpus (12)

• CE-Net: Context Encoder Network for 2D Medical Image Segmentation
• Reading Race: AI Recognises Patient's Racial Identity In Medical Images
• Interpretability of machine learning based prediction models in healthcare
• Adversarial attacks and defenses in explainable artificial intelligence: A survey
• mlr3proba: An R Package for Machine Learning in Survival Analysis
• Large-scale benchmark study of survival prediction methods using multi-omics data
• SurvSHAP(t): Time-dependent explanations of machine learning survival models
• Model-agnostic Feature Importance and Effects with Dependent Features -- A Conditional Subgroup Approach
• Grouped Feature Importance and Combined Features Effect Plot
• survex: an R package for explaining machine learning survival models
• The Grammar of Interactive Explanatory Model Analysis
• Explainable AI for survival analysis: a median-SHAP approach

Cited by in corpus (2)

• CTSL: Codebook-based Temporal-Spatial Learning for Accurate Non-Contrast Cardiac Risk Prediction Using Cine MRIs
• Birds look like cars: Adversarial analysis of intrinsically interpretable deep learning