XSRD-Net: EXplainable Stroke Relapse Detection

2025-09-09

AI Summary

This study addresses the challenge of early, precise prediction of ischemic stroke recurrence risk. We propose XSRD-Net, an interpretable multimodal deep learning framework that jointly models recurrence risk (binary classification) and recurrence-free survival time (survival regression) by fusing 3D intracranial computed tomographic angiography (CTA) images with structured clinical data. Our work innovatively uncovers the synergistic role of cardiac disease phenotypes and carotid artery imaging features in recurrence pathophysiology. Through modality contribution analysis and model-agnostic interpretability techniques, we ensure transparent, clinically actionable decision-making. Experiments demonstrate that a tabular-only baseline achieves an AUC of 0.84 for recurrence classification; in contrast, the multimodal XSRD-Net attains a c-index of 0.68 and AUC of 0.71 for survival prediction—significantly outperforming unimodal baselines. These results validate the clinical utility of integrating heterogeneous multimodal data for personalized secondary prevention strategies.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Stroke is the second most frequent cause of death world wide with an annual mortality of around 5.5 million. Recurrence rates of stroke are between 5 and 25% in the first year. As mortality rates for relapses are extraordinarily high (40%) it is of utmost importance to reduce the recurrence rates. We address this issue by detecting patients at risk of stroke recurrence at an early stage in order to enable appropriate therapy planning. To this end we collected 3D intracranial CTA image data and recorded concomitant heart diseases, the age and the gender of stroke patients between 2010 and 2024. We trained single- and multimodal deep learning based neural networks for binary relapse detection (Task 1) and for relapse free survival (RFS) time prediction together with a subsequent classification (Task 2). The separation of relapse from non-relapse patients (Task 1) could be solved with tabular data (AUC on test dataset: 0.84). However, for the main task, the regression (Task 2), our multimodal XSRD-net processed the modalities vision:tabular with 0.68:0.32 according to modality contribution measures. The c-index with respect to relapses for the multimodal model reached 0.68, and the AUC is 0.71 for the test dataset. Final, deeper interpretability analysis results could highlight a link between both heart diseases (tabular) and carotid arteries (vision) for the detection of relapses and the prediction of the RFS time. This is a central outcome that we strive to strengthen with ongoing data collection and model retraining.

Problem

• Detecting patients at risk of stroke recurrence early
• Predicting relapse-free survival time using multimodal data
• Linking heart diseases and carotid arteries to relapse prediction

Innovation

• Multimodal deep learning for relapse detection
• Combining 3D CTA images with clinical tabular data
• Explainable AI linking heart and artery features