Digital Registrar: A Schema-First Framework for Multi-Cancer Privacy-Preserving Pathology Abstraction via Local LLMs
Diagnostics · 27.05.2026 · 16 (11) · 1644
DOI: 10.3390/diagnostics16111644 ↗Kurz gefasst
Fachartikel in der Zeitschrift Diagnostics, online erschienen am 27.05.2026. Verfasst von Nan-Haw Chow und 9 weiteren Autor:innen. Beteiligte Einrichtung: Center for Precision Medicine, China Medical University Hospital, Taichung 404327, Taiwan. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 39 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Science and Technology Council, China Medical University Hospital, National Science and Technology Council of Taiwan.
Abstract (Originalsprache)
Background/Objectives: Free-text surgical pathology reports hinder automated cancer registry entry and secondary analytics. This study introduces a clinically governed schema layer for interoperability, testing whether a locally-deployable Large Language Model (LLM) pipeline can deliver robust registry-grade extraction across institutions.
Methods
We developed a College of American Pathologists (CAP)-aligned clinical ontology encompassing 10 cancer types, 192 per-organ scalar fields, key biomarkers, and nested structures for lymph nodes and margins. Encoded via Declarative Self-improving Python (DSPy) signatures with grammar-constrained decoding using DSPy v3.2.1, this model-agnostic pipeline was benchmarked on 893 internal reports against a pathologist-adjudicated gold standard. External validation utilized 242 The Cancer Genome Atlas (TCGA) reports. Hardware feasibility was confirmed on a single 48-gigabyte (GB) Graphics Processing Unit (GPU), ensuring suitability for privacy-preserving on-premises deployment.
Results
Using the gpt-oss-20b model, the framework achieved 92.0% macro-mean exact-match accuracy on internal data, demonstrating near-perfect run-to-run reliability. Critical prognostic indicators, including breast estrogen receptor/progesterone receptor (ER/PR) (98.7%) and margin positivity (>93%), maintained high fidelity. On the external TCGA cohort, accuracy was 77.5%, rising to 88.0% after excluding structurally silent fields absent in older narratives. Operationally, the model processed reports in 40–70 s, optimally balancing speed and accuracy.
Conclusions
This schema-first abstraction layer successfully decouples clinical logic from specific Artificial Intelligence (AI) models. By reliably transforming narrative reports into machine-readable structures, it establishes a portable privacy-preserving foundation for automated cancer surveillance, institutional data reuse, and future multimodal clinical systems.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Center for Precision Medicine, China Medical University Hospital, Taichung 404327, Taiwan.
- School of Medicine, China Medical University, Taichung 404333, Taiwan.
- Department of Pathology, China Medical University Hospital, Taichung 404327, Taiwan.
- Signal 1, Toronto, ON M5R 2E3, Canada.
- Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taipei 106990, Taiwan.
- Department of Foreign Languages and Literature, National Chi Nan University, Nantou 545301, Taiwan.
ORCID-Profile: Nan-Haw Chow, Han Chang, Yen-Wei Chu, Pau-Choo Chung, Kai-Po Chang
Förderung
- National Science and Technology Council
- China Medical University Hospital
- National Science and Technology Council of Taiwan
Zitieren
APA
Chow, N. H., Chang, H., Chen, H. K., Lin, C. Y., Liu, Y. L., Tseng, P. Y., Shiu, L. J., Chu, Y. W., Chung, P. C., & Chang, K. P. (2026). Digital Registrar: A Schema-First Framework for Multi-Cancer Privacy-Preserving Pathology Abstraction via Local LLMs. Diagnostics, 16(11), 1644. https://doi.org/10.3390/diagnostics16111644
Vancouver
Chow NH, Chang H, Chen HK, Lin CY, Liu YL, Tseng PY, et al. Digital Registrar: A Schema-First Framework for Multi-Cancer Privacy-Preserving Pathology Abstraction via Local LLMs. Diagnostics. 2026;16(11):1644. doi:10.3390/diagnostics16111644
BibTeX
@article{Chow2026,
author = {Chow, Nan-Haw and Chang, Han and Chen, Hung-Kai and Lin, Chen-Yuan and Liu, Ying-Lung and Tseng, Po-Yen and Shiu, Li-Ju and Chu, Yen-Wei and Chung, Pau-Choo and Chang, Kai-Po},
title = {Digital Registrar: A Schema-First Framework for Multi-Cancer Privacy-Preserving Pathology Abstraction via Local LLMs},
journal = {Diagnostics},
year = {2026},
volume = {16},
number = {11},
pages = {1644},
doi = {10.3390/diagnostics16111644}
}
EUVIMED