Automated and Interpretable Survival Analysis from Multimodal Data
This addresses the need for accurate and transparent survival prediction models in oncology, though it is incremental as it builds on existing methods like Grad-CAM and genetic programming.
The paper tackled survival analysis in oncology by developing an interpretable multimodal AI framework that integrates clinical and imaging data, achieving a C-index of 0.838 for prediction and 0.826 for stratification on a head and neck cancer dataset.
Accurate and interpretable survival analysis remains a core challenge in oncology. With growing multimodal data and the clinical need for transparent models to support validation and trust, this challenge increases in complexity. We propose an interpretable multimodal AI framework to automate survival analysis by integrating clinical variables and computed tomography imaging. Our MultiFIX-based framework uses deep learning to infer survival-relevant features that are further explained: imaging features are interpreted via Grad-CAM, while clinical variables are modeled as symbolic expressions through genetic programming. Risk estimation employs a transparent Cox regression, enabling stratification into groups with distinct survival outcomes. Using the open-source RADCURE dataset for head and neck cancer, MultiFIX achieves a C-index of 0.838 (prediction) and 0.826 (stratification), outperforming the clinical and academic baseline approaches and aligning with known prognostic markers. These results highlight the promise of interpretable multimodal AI for precision oncology with MultiFIX.