基於強化學習優化的大型語言模型推理之可解釋性神經退行性失智症診斷框架

An Explainable Diagnostic Framework for Neurodegenerative Dementias via Reinforcement-Optimized LLM Reasoning

May 26, 2025

作者: Andrew Zamai, Nathanael Fijalkow, Boris Mansencal, Laurent Simon, Eloi Navet, Pierrick Coupe

cs.AI

摘要

神經退行性失智症的鑑別診斷是一項具有挑戰性的臨床任務，主要由於症狀表現的重疊性以及結構性神經影像中觀察到的模式相似性。為提高診斷效率和準確性，基於深度學習的方法，如卷積神經網絡和視覺變壓器，已被提出用於腦部磁共振成像的自動分類。然而，儘管這些模型具有強大的預測性能，由於其決策過程的不透明性，它們在臨床應用中受到限制。在本研究中，我們提出了一個框架，整合了兩個核心組件以增強診斷的透明度。首先，我們引入了一個模組化流程，將3D T1加權腦部磁共振成像轉換為文本形式的放射學報告。其次，我們探討了現代大型語言模型（LLMs）在基於生成報告進行額顳葉失智症亞型、阿茲海默症與正常老化之間鑑別診斷的潛力。為彌補預測準確性與可解釋性之間的差距，我們採用強化學習來激勵LLMs進行診斷推理。無需監督推理軌跡或從更大模型進行蒸餾，我們的方法促成了基於神經影像發現的結構化診斷理據的出現。與事後解釋方法不同，這些方法回顧性地為模型決策提供合理性，我們的框架在推理過程中生成診斷理據——產生因果基礎的解釋，這些解釋不僅告知並引導模型的決策過程，而且使我們的框架在匹配現有深度學習方法診斷性能的同時，提供了支持其診斷結論的理據。

English

The differential diagnosis of neurodegenerative dementias is a challenging clinical task, mainly because of the overlap in symptom presentation and the similarity of patterns observed in structural neuroimaging. To improve diagnostic efficiency and accuracy, deep learning-based methods such as Convolutional Neural Networks and Vision Transformers have been proposed for the automatic classification of brain MRIs. However, despite their strong predictive performance, these models find limited clinical utility due to their opaque decision making. In this work, we propose a framework that integrates two core components to enhance diagnostic transparency. First, we introduce a modular pipeline for converting 3D T1-weighted brain MRIs into textual radiology reports. Second, we explore the potential of modern Large Language Models (LLMs) to assist clinicians in the differential diagnosis between Frontotemporal dementia subtypes, Alzheimer's disease, and normal aging based on the generated reports. To bridge the gap between predictive accuracy and explainability, we employ reinforcement learning to incentivize diagnostic reasoning in LLMs. Without requiring supervised reasoning traces or distillation from larger models, our approach enables the emergence of structured diagnostic rationales grounded in neuroimaging findings. Unlike post-hoc explainability methods that retrospectively justify model decisions, our framework generates diagnostic rationales as part of the inference process-producing causally grounded explanations that inform and guide the model's decision-making process. In doing so, our framework matches the diagnostic performance of existing deep learning methods while offering rationales that support its diagnostic conclusions.