DiagnoLLM：面向可解释疾病诊断的混合贝叶斯神经语言框架
DiagnoLLM: A Hybrid Bayesian Neural Language Framework for Interpretable Disease Diagnosis
AI总结：
针对临床AI需兼具准确性与可解释性的需求，提出融合贝叶斯建模、eQTL引导深度学习与LLM叙事生成的DiagnoLLM框架，在AD检测中达88%准确率，可生成适配医患的可信诊断报告。
摘要：
构建可信临床AI系统不仅需要准确预测，还需要透明、具备生物学依据的解释。我们提出DiagnoLLM，这一混合框架融合贝叶斯反卷积、eQTL引导的深度学习与基于大语言模型的叙事生成能力，用于可解释疾病诊断。DiagnoLLM首先采用基于高斯过程的分层模型GP-unmix，在建模生物不确定性的同时，从批量与单细胞RNA-seq数据中推断细胞类型特异性基因表达谱。这些特征结合eQTL分析得到的调控先验，为神经分类器提供支撑，使其在阿尔茨海默病（AD）检测任务中达到88.0%的准确率。为支持人类理解与信任，我们引入基于大语言模型的推理模块，将模型输出转化为面向特定受众的诊断报告，内容锚定临床特征、归因信号与领域知识。人工评估证实，这些报告准确、可执行，且适配医生与患者两类群体。我们的研究结果表明，当大语言模型作为事后推理器而非端到端预测器部署时，可在混合诊断流水线中充当有效的沟通载体。
英文摘要：
Building trustworthy clinical AI systems requires not only accurate predictions but also transparent, biologically grounded explanations. We present \texttt{DiagnoLLM}, a hybrid framework that integrates Bayesian deconvolution, eQTL-guided deep learning, and LLM-based narrative generation for interpretable disease diagnosis. DiagnoLLM begins with GP-unmix, a Gaussian Process-based hierarchical model that infers cell-type-specific gene expression profiles from bulk and single-cell RNA-seq data while modeling biological uncertainty. These features, combined with regulatory priors from eQTL analysis, power a neural classifier that achieves high predictive performance in Alzheimer's Disease (AD) detection (88.0\% accuracy). To support human understanding and trust, we introduce an LLM-based reasoning module that translates model outputs into audience-specific diagnostic reports, grounded in clinical features, attribution signals, and domain knowledge. Human evaluations confirm that these reports are accurate, actionable, and appropriately tailored for both physicians and patients. Our findings show that LLMs, when deployed as post-hoc reasoners rather than end-to-end predictors, can serve as effective communicators within hybrid diagnostic pipelines.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
