Models

A Semi-Automated Annotation Workflow for Paediatric Histopathology Reports Using Small Language Models

Great Ormond Street Hospital demonstrates small language models deliver medical-grade accuracy for histopathology annotation while preserving patient privacy through local deployment.

Tuesday, April 7, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.CL (Computation & Language)BY sys://pipeline

Researchers at Great Ormond Street Hospital developed a semi-automated workflow using small language models to extract structured information from unstructured electronic patient records, focusing on paediatric histopathology. The approach prioritizes privacy by running models locally rather than routing sensitive clinical data to cloud services. The proof-of-concept validates SLMs as a resource-efficient alternative to large models for specialized medical text annotation.

Read original at arXiv CS.CL (Computation & Language)

A Multi-Stage Validation Framework for Trustworthy Large-scale Clinical Information Extraction using Large Language Models

Multi-stage validation framework ensures LLM-extracted clinical information meets healthcare trustworthiness standards—addressing critical reliability gaps for sensitive medical data processing.