Extract Findings in Radiology Reports

This model is engineered for radiology texts and reports, adeptly identifying entities such as imaging tests, imaging techniques, imaging findings, and more.

Use the provided Streamlit playground application to test this service.

<p><br/></p>

**Covered Entities:** Imaging_Test, Imaging_Technique, ImagingFindings, OtherFindings, BodyPart, Direction, Test, Symptom, Disease_Syndrome_Disorder, Medical_Device, Procedure, Measurements, Units, Gender, Metastasis, Invasion, Route, Treatment, Drug, Form, Frequency, Dosage, Date, Test_Result, Medical_Device

**Assertion Labels**: Confirmed, Suspected, Negative.

C**overed Relations:** is_related, not_related

Developed with radiologists, technicians, and medical researchers in mind, the model brings high accuracy to the extraction of pivotal data points from radiological documentation. Harness the power of this pipeline to enhance diagnostic precision, streamline radiological workflows, and support data-driven clinical decision-making.

搜集汇总

背景与挑战

背景概述

该数据集专为放射学文本设计，可精准识别24类医疗实体（如影像检查、技术、结果等）及3种断言状态，支持2种关系分析，旨在辅助医疗专业人员提升诊断效率并优化临床决策流程。