Medical Heuristic Learning: An LLM-Driven Framework for Interpretable and Auditable Clinical Decision Rules

2026-06-15

AI Summary

This work addresses the dual challenge of achieving high predictive accuracy and interpretability in clinical tabular data, where existing black-box models struggle with limited sample sizes, severe class imbalance, and evolving feature distributions. To overcome these limitations, we propose the Medical Heuristic Learning (MHL) framework, which pioneers the use of large language models to drive non-gradient-based heuristic optimization. MHL integrates statistical and medical knowledge through probing, rule synthesis, and code-level iterative refinement to generate executable, interpretable, and auditable clinical decision rules in pure Python. The framework supports versioned rule management and enables continual learning under data drift while mitigating catastrophic forgetting. Empirical evaluation demonstrates that MHL matches or exceeds state-of-the-art performance across multiple clinical datasets, exhibits robustness in low-data and highly imbalanced settings, and effectively adapts to feature evolution over time.

Abstract

Predictive modeling for clinical tabular data is central to clinical decision support and therefore requires not only strong predictive performance but also transparent decision logic. Although deep learning and tree-based ensemble methods can achieve high accuracy, their black-box nature remains a major obstacle to clinical deployment. This challenge is further compounded by common characteristics of medical data, including limited sample sizes, severe class imbalance, and feature evolution arising from changes in diagnostic criteria and clinical documentation. To address these issues, we propose Medical Heuristic Learning (MHL), an instantiation of the learning-beyond-gradients paradigm for clinical tabular prediction. Instead of relying on neural network weight updates, MHL uses a large language model (LLM)-driven workflow that integrates statistical probes, medical knowledge probes, rule synthesis, and code-level iterative refinement to optimize a deterministic and executable decision system. The resulting model is expressed not as opaque parameters, but as versioned pure-Python decision rules that are explicitly interpretable, fully auditable, and clinically grounded. MHL also supports continual learning by starting from previously validated rules and iteratively revising them using updated feature information under data drift or feature evolution. Comprehensive experiments on medical datasets show that MHL achieves performance comparable to state-of-the-art methods while maintaining strong behavior in small-sample and highly imbalanced settings. The results further indicate that this explicit rule update mechanism can help alleviate catastrophic forgetting under feature evolution. Overall, these findings suggest that non-gradient-based heuristic systems offer a transparent and adaptable alternative for high-stakes clinical decision support.

Problem

• clinical decision support
• interpretable models
• tabular data
• class imbalance
• feature evolution

Innovation

• Medical Heuristic Learning
• LLM-driven workflow
• interpretable decision rules
• continual learning
• non-gradient-based optimization