医学
协议(科学)
唾液腺
医学物理学
放射科
超声波
超声科
回顾性队列研究
病理
疾病
梅德林
唾液腺疾病
适宜性标准
医学影像学
机构审查委员会
作者
Ali H. Dhanaliwala,Dana DiRenzo
标识
DOI:10.1097/rhu.0000000000002375
摘要
OBJECTIVES: Salivary gland ultrasound (SGUS) plays a vital role in diagnosing and monitoring glandular disease activity, including the risk of lymphomagenesis in Sjogren's disease (SjD). Despite the importance of SGUS, access is often limited as rheumatologists do not always have the resources to perform these exams. Radiology, which specializes in image acquisition and interpretation at volume, has the potential to improve access to SGUS for patients undergoing SjD workup. The goal of this study was to evaluate the feasibility of radiology to perform and report SGUS exams. METHODS: A retrospective analysis of SGUS exams completed at a large urban academic radiology department was conducted. Evaluation was limited to SGUS examinations performed after the implementation of a standardized imaging protocol and reporting template for OMERACT (Outcome Measures in Rheumatology) scoring, developed through a collaborative initiative with rheumatology. Data were extracted from the ultrasound Digital Imaging and Communications in Medicine (DICOM) header, and a large language model (LLM) was used to evaluate the study reports. RESULTS: Over a 42-month period, a total of 811 SGUS exams specifically requesting OMERACT scoring of the salivary glands were completed by radiology. The radiology sonographers took an average of 17 minutes to acquire the images, and the exams were reported by 53 different radiologists. The majority of reports (64%) adhered to the template, and OMERACT scores and gland volumes were documented in 87% of reports. CONCLUSIONS: We describe the implementation of a standardized SGUS protocol at a large urban academic health system that includes a wide range of radiologists with diverse experience, demonstrating the ability of radiologists to efficiently and effectively perform SGUS for OMERACT scoring as part of routine care.
科研通智能强力驱动
Strongly Powered by AbleSci AI