EndoExtract: Co-Designing Structured Text Extraction from Endometriosis Ultrasound Reports

2026-01-26

AI Summary

This study addresses the inefficiency in downstream analysis and modeling caused by the prevalence of unstructured text in endometriosis ultrasound reports, which traditionally rely on manual abstraction. The authors propose a locally deployed large language model (LLM)-driven system that automatically extracts structured clinical data while innovatively integrating key pain points from clinical workflows into its interactive design. Key features include mandatory review of interpretable fields, automatic highlighting of source evidence, and decoupling the pace of batch extraction from human verification. This approach shifts users from field-by-field data entry to a supervisory validation paradigm, substantially improving efficiency. The work also uncovers emerging challenges, such as risks associated with rapid skimming and handling missing data, thereby highlighting new directions for human-in-the-loop clinical NLP systems.

Technology Category

• Natural Language Processing: Summarization
• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Web Mining and Content Analysis: Bridging structured and unstructured data

Abstract

Endometriosis ultrasound reports are often unstructured free-text documents that require manual abstraction for downstream tasks such as analytics, machine learning model training, and clinical auditing. We present \textbf{EndoExtract}, an on-premise LLM-powered system that extracts structured data from these reports and surfaces interpretive fields for human review. Through contextual inquiry with research assistants, we identified key workflow pain points: asymmetric trust between numerical and interpretive fields, repetitive manual highlighting, fatigue from sustained comparison, and terminology inconsistency across radiologists. These findings informed an interface that surfaces only interpretive fields for mandatory review, automatically highlights source evidence within PDFs, and separates batch extraction from human-paced verification. A formative workshop revealed that \textbf{EndoExtract} supports a shift from field-by-field data entry to supervisory validation, though participants noted risks of over-skimming and challenges in managing missing data.

Problem

• endometriosis
• ultrasound reports
• structured text extraction
• manual abstraction
• clinical documentation

Innovation

• structured text extraction
• large language model (LLM)
• human-in-the-loop
• clinical NLP
• co-design