Attention-enabled Explainable AI for Bladder Cancer Recurrence Prediction

2025-04-30

AI Summary

Non-muscle-invasive bladder cancer (NMIBC) exhibits high recurrence rates (70–80%), yet existing predictive models suffer from risk overestimation, poor individualization, and limited interpretability—hindering precision clinical decision-making. Method: We propose an interpretable deep learning framework tailored for tabular clinical data, integrating categorical variable embeddings with attention mechanisms to enable patient-level recurrence risk prediction and visual attribution of key contributing factors. Contribution/Results: Our model identifies novel prognostic indicators—including surgical duration and length of hospital stay—for the first time. By transcending the “black-box” limitations of conventional statistical models, it delivers clinically intuitive, patient-specific explanations. Evaluated on real-world data, it achieves 70% accuracy—significantly outperforming standard approaches—and provides a reliable, performance-robust tool for NMIBC risk stratification and dynamic clinical management.

Technology Category

• Machine Learning: Transparent, Interpretable, Explainable ML
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Non-muscle-invasive bladder cancer (NMIBC) is a relentless challenge in oncology, with recurrence rates soaring as high as 70-80%. Each recurrence triggers a cascade of invasive procedures, lifelong surveillance, and escalating healthcare costs - affecting 460,000 individuals worldwide. However, existing clinical prediction tools remain fundamentally flawed, often overestimating recurrence risk and failing to provide personalized insights for patient management. In this work, we propose an interpretable deep learning framework that integrates vector embeddings and attention mechanisms to improve NMIBC recurrence prediction performance. We incorporate vector embeddings for categorical variables such as smoking status and intravesical treatments, allowing the model to capture complex relationships between patient attributes and recurrence risk. These embeddings provide a richer representation of the data, enabling improved feature interactions and enhancing prediction performance. Our approach not only enhances performance but also provides clinicians with patient-specific insights by highlighting the most influential features contributing to recurrence risk for each patient. Our model achieves accuracy of 70% with tabular data, outperforming conventional statistical methods while providing clinician-friendly patient-level explanations through feature attention. Unlike previous studies, our approach identifies new important factors influencing recurrence, such as surgical duration and hospital stay, which had not been considered in existing NMIBC prediction models.

Problem

• Predicting high recurrence rates in non-muscle-invasive bladder cancer (NMIBC)
• Overcoming flaws in existing clinical prediction tools for NMIBC
• Providing interpretable, personalized insights for patient management

Innovation

• Interpretable deep learning with attention mechanisms
• Vector embeddings for categorical variables
• Identifies new factors like surgical duration