Extraction of clinical information from faxed medical records using a small local large language model pipeline on consumer hardware
- 1University of Texas at Austin
Abstract
ObjectiveSpecialty referral packets often arrive by fax and clinicians must manually read and organize a large amount of disparate information to affect continuity of care. Commercial large language models (LLMs) can organize packet information, but protected health information (PHI) must be secured. We evaluated whether an open-weight pipeline running entirely on one consumer graphics card can accurately extract clinical information from faxed referral packets.
Materials and MethodsWe generated 200 synthetic colorectal cancer referral packets from simulated patient records, rendered them as faxes without a text layer. A local pipeline transcribed each packet with an optical character recognition model (OvisOCR2) and extracted structured data with an open-weight LLM (Qwen3.8-27B). Twenty outcomes spanning patient identifiers, laboratory results, dates, diagnoses, procedures, medications, imaging, and event chronology were scored against ground truth on 140 validation packets.
ResultsLaboratory values were extracted exactly in 98.9% of 7879 instances, dates in 99.7%, diagnoses in 100%, and procedures in 92.9%. Timeline recall was 96.7% and precision 99.4%. Weighted accuracy was 0.990 (95% CI 0.987-0.992). None of 11,993 extracted facts were fabricated. Median processing time was 90 seconds per packet.
ConclusionAn open-weight pipeline on consumer hardware accurately extracted multi-domain clinical information from synthetic faxed referrals.
Related
- Beyond Billing Codes: An Agentic Large Language Model Pipeline for Extracting Glaucoma Procedural Data from Free-Text Clinical Notes
- Bern Sleep-Wake Registry: temporal extraction of comorbidities from unstructured clinical text using a large language model
- Comparative analysis of open-source locally deployed LLMs for extraction of diagnostic information from clinical CMR imaging reports
- Zero-Shot Extraction of Structured Clinical Data from Free-Text CT Reports Using Foundational Large Language Models.
- Leveraging Large Language Models for Colorectal Cancer Symptom Extraction from MIMIC-IV Clinical Notes
BibTeX
@misc{tang2026extraction,
title = {{Extraction of clinical information from faxed medical records using a small local large language model pipeline on consumer hardware}},
author = {Z. Tang},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.03.26364667},
url = {https://www.medrxiv.org/content/10.64898/2026.10.03.26364667v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Extraction of clinical information from faxed medical records using a small local large language model pipeline on consumer hardware
AU - Z. Tang
PY - 2026
DA - 2026/10/03
PB - medRxiv
DO - 10.64898/2026.10.03.26364667
UR - https://www.medrxiv.org/content/10.64898/2026.10.03.26364667v1
AB - ObjectiveSpecialty referral packets often arrive by fax and clinicians must manually read and organize a large amount of disparate information to affect continuity of care. Commercial large language models (LLMs) can organize packet information, but protected health information (PHI) must be secured. We evaluated whether an open-weight pipeline running entirely on one consumer graphics card can accurately extract clinical information from faxed referral packets. Materials and MethodsWe generated 200 synthetic colorectal cancer referral packets from simulated patient records, rendered them as faxes without a text layer. A local pipeline transcribed each packet with an optical character recognition model (OvisOCR2) and extracted structured data with an open-weight LLM (Qwen3.8-27B). Twenty outcomes spanning patient identifiers, laboratory results, dates, diagnoses, procedures, medications, imaging, and event chronology were scored against ground truth on 140 validation packets. ResultsLaboratory values were extracted exactly in 98.9% of 7879 instances, dates in 99.7%, diagnoses in 100%, and procedures in 92.9%. Timeline recall was 96.7% and precision 99.4%. Weighted accuracy was 0.990 (95% CI 0.987-0.992). None of 11,993 extracted facts were fabricated. Median processing time was 90 seconds per packet. ConclusionAn open-weight pipeline on consumer hardware accurately extracted multi-domain clinical information from synthetic faxed referrals.
ER -