基于YOLOv9的多中心MRI研究:冈上肌腱撕裂的自动检测与分类
本研究基于五家医院1698例患者的MRI数据,开发YOLOv9深度学习模型用于冈上肌腱撕裂的自动检测与分类。回顾性纳入患者并分为训练集1047例、验证集299例、测试集154例和外部测试集198例,使用冠状位脂肪抑制T2加权图像。模型在外部测试集检测准确率为75.64%,测试集分类宏F1为77.7%,优于所有放射科医生。
要点
- YOLOv9模型在所有验证、测试和外部测试集图像中成功识别冈上肌腱层,准确率达100%。
- 冈上肌腱撕裂检测准确率:验证集69.0%、测试集73.9%、外部测试集75.64%。
- 测试集部分与全层撕裂分类宏F1为77.7%,优于所有经验水平的放射科医生。
核心结论
- 基于MRI的YOLOv9模型在诊断冈上肌腱撕裂方面表现优异,超过了不同经验水平的放射科医生。
研究设计与样本
- 回顾性研究,纳入五家医院1698例患者,分为训练集1047例、验证集299例、测试集154例和外部测试集198例,使用冠状位脂肪抑制T2加权图像开发YOLOv9自动模型,并以交并比和混淆矩阵评估性能,采用McNemar检验比较模型与放射科医生判读,采用Cohen kappa评估放射科医生间一致性。
关键影像表现/征象
- YOLOv9模型在所有验证、测试和外部测试集图像中成功识别冈上肌腱层,准确率达100%。
诊断标准与关键数值
- 冈上肌腱撕裂检测准确率:验证集69.0%(755/1094)、测试集73.9%(414/560)、外部测试集75.64%(559/739)。
- 测试集部分与全层撕裂分类的宏F1为77.7%(95% CI:67.4-90.5),优于所有放射科医生(均P<0.05)。
临床应用与阅片要点
- 原文未据此给出推荐。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVES: This study develops a deep learning model using the You Only Look Once (YOLO) framework for the automated diagnosis of supraspinatus tendon tears (ST) based on multicenter MRI data.
MATERIALS AND METHODS: In this retrospective study, 1698 patients from five hospitals were included and allocated to training (n=1047), validation (n=299), test (n=154), and external test (n=198) sets. A YOLOv9-based automated model was developed using coronal fat-suppressed T2-weighted images for lesion detection, localization, and classification. Model performance was assessed using Intersection over Union and confusion matrices. Comparisons between model outputs and radiologist interpretations were performed with McNemar's test, and interobserver agreement among radiologists was evaluated using Cohen's kappa.
RESULTS: The YOLOv9 model successfully identified the supraspinatus tendon layer in all images across the validation, test, and external test sets, achieving 100% accuracy. For ST tear detection, the model achieved accuracies of 69.0% (755/1094) in the validation set, 73.9% (414/560) in the test set, and 75.64% (559/739) in the external test set. For classification of partial- and full-thickness tears on the test set, the model demonstrated a macro F1 score of 77.7% (95% CI: 67.4-90.5), outperforming all radiologists (all P<0.05).
CONCLUSION: The MRI-based YOLOv9 model excelled in diagnosing supraspinatus tendon tears, surpassing radiologists with varying levels of experience.
引用
- Yang X, Liu Z, Jiang H, et al. Multicenter Study of YOLOv9 for Automated Detection and Classification of Supraspinatus Tendon Tears on MRI. Acad Radiol. 2025. PMID: 41173764. doi:10.1016/j.acra.2025.10.022
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。