Suppr超能文献

用于睾丸病理学结构化报告的超声成像算法的开发与验证

Development and Validation of an Ultrasound Imaging Algorithm for Structured Reporting in Testicular Pathology.

作者信息

Pintican Roxana, Negrea Alexandru, Boll Isabell, Boca Bianca, Gherman Diana, Bora Marilena, Dudea Sorin, Ciurea Anca

机构信息

Department of Radiology, "Iuliu Hatieganu" University of Medicine and Pharmacy, 400347 Cluj-Napoca, Romania.

Department of Radiology, Prof Dr Ion Chiricuta Oncology Institute, 400015 Cluj-Napoca, Romania.

出版信息

Diagnostics (Basel). 2025 Apr 9;15(8):951. doi: 10.3390/diagnostics15080951.

Abstract

: Testicular ultrasound (US) imaging is a critical modality for diagnosing a variety of testicular pathologies, including malignancies. This study aimed to develop and validate a standardized diagnostic algorithm to enhance diagnostic accuracy and consistency in evaluating testicular lesions, particularly for distinguishing between benign and malignant conditions. : The algorithm was applied retrospectively to 110 testicular imaging cases, including 90 abnormal and 20 normal cases, analyzed by three radiologists with varying experience levels. Key diagnostic features, including lesion morphology, vascularity, and echotexture, were evaluated to guide the differentiation process. : demonstrated high diagnostic accuracy, with sensitivity reaching 100% for detecting abnormal cases and specificity ranging between 80% and 95%. In distinguishing benign from malignant lesions, the algorithm achieved an area under the curve (AUC) of up to 0.917, with specificities exceeding 93%. Notably, strong inter-rater agreement was observed, underscoring the algorithm's reliability across different expertise levels. While the algorithm significantly improved standardization and diagnostic performance, some variability in sensitivity for less experienced evaluators highlights the need for further refinement. : This study shows that the proposed diagnostic algorithm is an effective tool for testicular US, facilitating accurate and reproducible assessments, which are crucial for early detection and optimal management of testicular pathologies.

摘要

睾丸超声(US)成像对于诊断包括恶性肿瘤在内的多种睾丸疾病是一种关键的检查方法。本研究旨在开发并验证一种标准化诊断算法,以提高在评估睾丸病变,特别是区分良性和恶性情况时的诊断准确性和一致性。该算法被回顾性应用于110例睾丸成像病例,包括90例异常病例和20例正常病例,由三位经验水平不同的放射科医生进行分析。评估关键诊断特征,包括病变形态、血管分布和回声纹理,以指导鉴别过程。结果显示该算法具有较高的诊断准确性,检测异常病例的敏感性达到100%,特异性在80%至95%之间。在区分良性和恶性病变方面,该算法的曲线下面积(AUC)高达0.917,特异性超过93%。值得注意的是,观察者间一致性较强,这突出了该算法在不同专业水平上的可靠性。虽然该算法显著提高了标准化程度和诊断性能,但经验较少的评估者在敏感性方面存在一些差异,这凸显了进一步优化的必要性。这项研究表明,所提出的诊断算法是睾丸超声检查的有效工具,有助于进行准确且可重复的评估,这对于睾丸疾病的早期检测和最佳管理至关重要。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/658c/12025588/7dcc9124c11e/diagnostics-15-00951-g001.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验