Patient-Conditioned Dual Hypergraph Reasoning for Auditable Traditional Chinese Medicine Prescription Support

• URL: http://arxiv.org/abs/2607.04025v1
• Date: Sat, 04 Jul 2026 20:48:53 GMT
• Title: Patient-Conditioned Dual Hypergraph Reasoning for Auditable Traditional Chinese Medicine Prescription Support
• Abstract summary: We propose a patient-conditioned dual hypergraph framework for auditable TCM prescription support.<n>The first hypergraph organizes symptom, tongue, pulse, and other clinical evidence around syndrome and treatment-principle reasoning.<n>The second hypergraph organizes syndrome, treatment, disease-context, herb, retrieval, and dose-prior evidence for prescription construction.
• Score: 31.95109204556091
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Traditional Chinese medicine (TCM) prescription support requires patient-specific reasoning from clinical narratives to syndromes, treatment principles, herbs, and doses. Direct language-model generation can produce fluent prescriptions, but its decisions are difficult to audit against explicit clinical evidence. Static TCM knowledge resources provide useful priors, but they cannot determine which diagnostic and prescription relations should be emphasized for an individual patient. We propose a patient-conditioned dual hypergraph framework for auditable TCM prescription support. The first hypergraph organizes symptom, tongue, pulse, and other clinical evidence around syndrome and treatment-principle reasoning. The second hypergraph organizes syndrome, treatment, disease-context, herb, retrieval, and dose-prior evidence for prescription construction. Unlike static knowledge graphs or fixed hypergraphs, both hypergraphs are dynamically weighted by the patient representation. This design enables individualized activation of diagnostic and prescription paths, supporting personalized syndrome differentiation and herb-dose recommendation while preserving case-level auditability. Experiments on TCM-SD show that dynamic weighting in the first hypergraph improves MacBERT syndrome differentiation to 0.8297 accuracy and 0.3288 macro-F1. On TCM-BEST4SDT, the second hypergraph achieves the best mean Herb-F1 of 0.3111 across three seeds, and the full connected pipeline reaches 0.3074 Herb-F1, close to the oracle setting. A 50-case real-world CAP audit further suggests practical review potential, while highlighting the need for prospective dose-safety validation.

Related papers

• Disease-Centric Vision-Language Pretraining with Hybrid Visual Encoding for 3D Computed Tomography [75.7789001619107]
  Vision-language pre-training holds great promise for general-purpose medical AI.<n>We propose a tailored framework featuring three key components.<n>A diagnosis-aware prompt strategy employs real clinical phrases and aggregated disease prototypes to bridge the pre-training-inference gap.
  arXiv  Detail & Related papers  (2026-06-24T08:24:45Z)
• Analyzing and Improving Fine-grained Preference Optimization in Medical LVLMs [53.697403481143404]
  Large Vision-Language Models (LVLMs) have achieved strong performance across medical imaging tasks.<n>But they remain prone to factual inconsistencies, poor visual grounding, and misalignment with clinically meaningful feedback.
  arXiv  Detail & Related papers  (2026-06-10T18:35:36Z)
• Benchmarking Multi-turn Medical Diagnosis: Hold, Lure, and Self-Correction [72.89352076103889]
  Large language models (LLMs) achieve high accuracy in medical diagnosis when all clinical information is provided in a single turn.<n>We introduce MINT, a high-fidelity, multi-turn medical diagnosis benchmark comprising 1,035 cases.<n>We uncover three persistent behavioral patterns that significantly impact diagnostic decisions.
  arXiv  Detail & Related papers  (2026-04-06T00:23:10Z)
• MedEinst: Benchmarking the Einstellung Effect in Medical LLMs through Counterfactual Differential Diagnosis [13.241795322837861]
  We introduce MedEinst, a counterfactual benchmark with 5,383 paired clinical cases across 49 diseases.<n>We measure susceptibility via Bias Trap Rate--probability of misdiagnosing traps despite correctly diagnosing controls.
  arXiv  Detail & Related papers  (2026-01-10T17:39:25Z)
• A Hierarchical Structure-Enhanced Personalized Recommendation Model for Traditional Chinese Medicine Formulas Based on KG Diffusion Guidance [3.17076026949853]
  We propose a novel hierarchical structure-enhanced personalized recommendation model for TCM formulas based on knowledge graph diffusion guidance, namely TCM-HEDPR.<n>Specifically, we pre-train symptom representations using patient-personalized prompt sequences and apply prompt-oriented contrastive learning for data augmentation. Furthermore, we employ a KG-guided homogeneous graph diffusion method integrated with a self-attention mechanism to globally capture the non-linear symptom-herb relationship.
  arXiv  Detail & Related papers  (2025-09-28T01:40:01Z)
• Generalised Label-free Artefact Cleaning for Real-time Medical Pulsatile Time Series [3.8195510803972454]
  Artefacts compromise clinical decision-making in the use of medical time series.<n>We introduce a generalised label-free framework, GenClean, for real-time artefact detection.
  arXiv  Detail & Related papers  (2025-04-29T22:28:06Z)
• Finetuning and Quantization of EEG-Based Foundational BioSignal Models on ECG and PPG Data for Blood Pressure Estimation [46.36100528165335]
  Photoplethysmography and electrocardiography can potentially enable continuous blood pressure (BP) monitoring.<n>Yet accurate and robust machine learning (ML) models remains challenging due to variability in data quality and patient-specific factors.<n>In this work, we investigate whether a model pre-trained on one modality can effectively be exploited to improve the accuracy of a different signal type.<n>Our approach achieves near state-of-the-art accuracy for diastolic BP and surpasses by 1.5x the accuracy of prior works for systolic BP.
  arXiv  Detail & Related papers  (2025-02-10T13:33:12Z)
• Natural Language-Assisted Multi-modal Medication Recommendation [97.07805345563348]
  We introduce the Natural Language-Assisted Multi-modal Medication Recommendation(NLA-MMR)<n>The NLA-MMR is a multi-modal alignment framework designed to learn knowledge from the patient view and medication view jointly.<n>In this vein, we employ pretrained language models(PLMs) to extract in-domain knowledge regarding patients and medications.
  arXiv  Detail & Related papers  (2025-01-13T09:51:50Z)
• TCM-FTP: Fine-Tuning Large Language Models for Herbal Prescription Prediction [17.041413449854915]
  Traditional Chinese medicine (TCM) has relied on specific combinations of herbs in prescriptions to treat various symptoms and signs for thousands of years.<n> Predicting TCM prescriptions poses a fascinating technical challenge with significant practical implications.<n>We introduce textitDigestDS, a novel dataset comprising practical medical records from experienced experts in digestive system diseases.<n>We also propose a method, TCM-FTP (TCM Fine-Tuning Pre-trained), to leverage pre-trained large language models (LLMs) via supervised fine-tuning on textitDigestDS.
  arXiv  Detail & Related papers  (2024-07-15T08:06:37Z)
• Fusing Echocardiography Images and Medical Records for Continuous Patient Stratification [16.93115087698284]
  We propose a method to learn the representation of a cardiovascular pathology with a difficult-to-characterize continuum, namely hypertension.<n>Our method first projects each variable into its own representation space using modality-specific approaches.<n>These standardized representations of multimodal data are then fed to a Transformer encoder, which learns to merge them into a comprehensive representation of the patient through the task of predicting a clinical rating.<n>We observe the major trends along this continuum on a cohort of 239 hypertensive patients, providing unprecedented details in the description of hypertension's impact on various cardiac function descriptors.
  arXiv  Detail & Related papers  (2024-01-15T16:04:46Z)
• Cross-modal Clinical Graph Transformer for Ophthalmic Report Generation [116.87918100031153]
  We propose a Cross-modal clinical Graph Transformer (CGT) for ophthalmic report generation (ORG) CGT injects clinical relation triples into the visual features as prior knowledge to drive the decoding procedure. Experiments on the large-scale FFA-IR benchmark demonstrate that the proposed CGT is able to outperform previous benchmark methods.
  arXiv  Detail & Related papers  (2022-06-04T13:16:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.