Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Currently submitted to: JMIR Medical Informatics

Date Submitted: Feb 12, 2026
Open Peer Review Period: Feb 23, 2026 - Apr 20, 2026
(closed for review but you can still tweet)

NOTE: This is an unreviewed Preprint

Warning: This is a unreviewed preprint (What is a preprint?). Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn (a note "no longer under consideration" will appear above).

Peer review me: Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period (in this case, a "Peer Review Me" button to sign up as reviewer is displayed above). All preprints currently open for review are listed here. Outside of the formal open peer-review period we encourage you to tweet about the preprint.

Citation: Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).

Final version: If our system detects a final peer-reviewed "version of record" (VoR) published in any journal, a link to that VoR will appear below. Readers are then encourage to cite the VoR instead of this preprint.

Settings: If you are the author, you can login and change the preprint display settings, but the preprint URL/DOI is supposed to be stable and citable, so it should not be removed once posted.

Submit: To post your own preprint, simply submit to any JMIR journal, and choose the appropriate settings to expose your submitted version as preprint.

Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.

Topology-Aware Context Routing and Neuro-Symbolic Reasoning for Clinical Evidence Extraction: Framework Development and Validation

  • Yalan Huang; 
  • Longchao Wang; 
  • Xuemei Yang; 
  • Xiaoli Tang

ABSTRACT

Background:

The automatic extraction of Population, Intervention, Comparison, and Outcome (PICO) elements is foundational to evidence-based medicine (EBM). Despite the transformative potential of LLMs, these models consistently struggle with structural hallucinations—specifically the logical misalignment between interventions and outcomes—and semantic sparsity when processing complex, multi-arm clinical trial data, particularly in high-heterogeneity fields such as oncology.

Objective:

This study develops and validates the Topology-aware Dynamic Context Activation and Neuro-Symbolic (TaD-NeS) framework. The primary objective is to transition clinical evidence extraction from traditional keyword-based retrieval to a deterministic mapping paradigm, thereby ensuring the structural integrity of complex medical knowledge within a Comparison-Centric Ontology Hypergraph (CCOH).

Methods:

The TaD-NeS framework integrates a dual-source anchoring mechanism that aligns PubMed abstracts with ClinicalTrials.gov metadata. We implement a boundary defense strategy using negative constraints to restrict the generative space of the LLM to predefined logical topographies. The core architecture utilizes a topology-aware router and neuro-symbolic reasoning to parse unstructured text into N-ary relational units, effectively decoupling neural semantic perception from symbolic logical verification.

Results:

Evaluation on a curated breast cancer trial dataset, comprising 653 publications and 7,382 CCOH nodes, demonstrates superior performance. The framework achieves an Entity F1-score of 94.75% and a landmark Unit-Accuracy of 94.82%, significantly outperforming baseline models. The high precision (95.10%) and numerical tolerance accuracy (97.05%) confirm the ability of the framework to eliminate inter-group mismatching. Case studies validate the auditability of the system, providing a transparent digital fingerprint for every extracted evidence link.

Conclusions:

The TaD-NeS framework provides a robust solution to the black-box nature of LLMs in clinical informatics. By enforcing topological constraints and neuro-symbolic synergy, it achieves a paradigm shift toward deterministic evidence synthesis. This approach not only mitigates structural hallucinations but also establishes a verifiable foundation for automated systematic reviews and next-generation clinical decision support systems.


 Citation

Please cite as:

Huang Y, Wang L, Yang X, Tang X

Topology-Aware Context Routing and Neuro-Symbolic Reasoning for Clinical Evidence Extraction: Framework Development and Validation

JMIR Preprints. 12/02/2026:93396

DOI: 10.2196/preprints.93396

URL: https://preprints.jmir.org/preprint/93396

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.