$\Delta$t-Mamba3D: A Time-Aware Spatio-Temporal State-Space Model for Breast Cancer Risk Prediction

TL;DR

This paper introduces $ riangle$t-Mamba3D, a novel time-aware state-space model for longitudinal breast cancer risk prediction that effectively captures irregular time intervals and spatio-temporal features in sequential medical images.

Contribution

It proposes a continuous-time state-space architecture with a selective scanning mechanism and multi-scale fusion, enabling efficient and accurate modeling of irregularly timed longitudinal medical imaging data.

Findings

Outperforms existing models with 2-5% higher validation c-index.

Achieves superior 1-5 year AUC scores in breast cancer risk prediction.

Handles long patient histories efficiently with linear complexity.

Abstract

Longitudinal analysis of sequential radiological images is hampered by a fundamental data challenge: how to effectively model a sequence of high-resolution images captured at irregular time intervals. This data structure contains indispensable spatial and temporal cues that current methods fail to fully exploit. Models often compromise by either collapsing spatial information into vectors or applying spatio-temporal models that are computationally inefficient and incompatible with non-uniform time steps. We address this challenge with Time-Aware $\Delta$t-Mamba3D, a novel state-space architecture adapted for longitudinal medical imaging. Our model simultaneously encodes irregular inter-visit intervals and rich spatio-temporal context while remaining computationally efficient. Its core innovation is a continuous-time selective scanning mechanism that explicitly integrates the true time difference between exams into its state transitions. This is complemented by a multi-scale 3D neighborhood fusion module that robustly captures spatio-temporal relationships. In a comprehensive breast cancer risk prediction benchmark using sequential screening mammogram exams, our model shows superior performance, improving the validation c-index by 2-5 percentage points and achieving higher 1-5 year AUC scores compared to established variants of recurrent, transformer, and state-space models. Thanks to its linear complexity, the model can efficiently process long and complex patient screening histories of mammograms, forming a new framework for longitudinal image analysis.