Ultrasound examiners' ability to describe ovarian cancer spread using preacquired ultrasound videoclips from a selected patient sample with high prevalence of cancer spread

• DOI: 10.1002/uog.29208
• Publication Date: 2025-04-18T07:44:11Z
• Authors: D. Fischerova; P. Pinto; M. Pesta; M. Blasko; M. C. Moruzzi; A. C. Testa; D. Franchi; V. Chiappa; J. L. Alcázar; M. Wiesnerova; D. Cibula; L. Valentin

ABSTRACT Objectives To assess the ability, as well factors affecting of ultrasound examiners with different levels experience to detect correctly infiltration ovarian cancer in predefined anatomical locations, and evaluate inter‐rater agreement regarding presence or absence infiltration, using preacquired videoclips obtained a selected patient sample high prevalence spread. Methods This study forms part Imaging Study Advanced ovArian Cancer multicenter observational (NCT03808792). Ultrasound showing assessment were by principal investigator (an expert, who did not participate rating) at 19 sites abdomen pelvis, including five that, if infiltrated, would indicate tumor non‐resectability. For each site, there 10 no infiltration. The reference standard was either findings surgery histological confirmation response chemotherapy. statistical analysis, grouped into four regions: middle abdomen, upper lymph nodes. assessed raters comprising both senior gynecologists (mainly self‐trained expert perform preoperative spread almost daily) had undergone minimum 6 months' supervised training gynecological oncology center. classified highly experienced less based on annual individual caseload number years that they been performing evaluation Raters aware for site be without Each rater independently every videoclip rated image quality (on scale from 0 10) their diagnostic confidence 10). A generalized linear mixed model random effects used estimate which (including level experience, quality, region) affected likelihood correct classification We observed percentage correctly, expected (reliability) classifying being infiltrated cancer. Results Twenty‐five participated study, whom 13 12 experienced. ranged 70% 100% depending median 25 90% 100%. probability all 380 0.956 0.975 rater's experience. increased region. It highest second those third nodes lowest abdomen. substantial (Fleiss kappa, 0.68 (95% CI, 0.66–0.71)) very good 0.99 0.97–1.00)) site. 0.66–0.71) 0.97 0.94–0.99)) pelvis 0.94 0.92–0.97) 0.97–1.00)). Conclusions can classify video recordings an examiner, is substantial. However, owing design, our results accuracy are likely overoptimistic. © 2025 Author(s). Obstetrics & Gynecology published John Wiley Sons Ltd behalf International Society Gynecology.