基于CT的深度学习模型在鉴别≤8 mm实性肺结节良恶性中的价值
研究纳入719例经手术切除的小实性肺结节患者,构建基于CT的MDANet深度学习模型鉴别良恶性及良性肿瘤与炎性结节。模型4(结节+15 mm瘤周区域)外部验证AUC 0.730,优于四种常规算法;模型8(结节+10 mm瘤周区域)鉴别良性肿瘤与炎性结节外部验证AUC 0.871。
要点
- 结节+15 mm瘤周区域模型鉴别良恶性SPN外部验证AUC 0.730。
- 结节+10 mm瘤周区域模型鉴别良性肿瘤与炎性结节外部验证AUC 0.871。
- 最佳模型外部验证表现优于VGG19、ResNet50、ResNeXt50和DenseNet121。
核心结论
- 基于MDANet构建的CT深度学习模型可准确鉴别小实性肺结节良恶性及良性肿瘤与炎性结节。
研究设计与样本
- 研究纳入719例具有小实性肺结节、术前接受胸部CT并接受手术切除的患者,分为内部训练、内部验证和外部验证队列。
诊断标准与关键数值
- 模型4(结节+15 mm瘤周区域)外部验证鉴别良恶性SPN的AUC 0.730、准确率0.724、召回率0.711、精确率0.705、F1分数0.707。
- 模型8(结节+10 mm瘤周区域)外部验证鉴别良性肿瘤与炎性结节的AUC 0.871、准确率0.938、召回率0.863、精确率0.904、F1分数0.882。
临床应用与阅片要点
- 作者认为基于MDANet的CT深度学习模型可用于鉴别小实性肺结节良恶性及良性肿瘤与炎性结节。
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: We examined the effectiveness of computed tomography (CT)-based deep learning (DL) models in differentiating benign and malignant solid pulmonary nodules (SPNs) ≤ 8 mm.
MATERIALS AND METHODS: The study patients (n = 719) were divided into internal training, internal validation, and external validation cohorts; all had small SPNs and had undergone preoperative chest CTs and surgical resection. We developed five DL models incorporating features of the nodule and five different peri-nodular regions with the Multiscale Dual Attention Network (MDANet) to differentiate benign and malignant SPNs. We selected the best-performing model, which was then compared to four conventional algorithms (VGG19, ResNet50, ResNeXt50, and DenseNet121). Furthermore, another five DL models were constructed using MDANet to distinguish benign tumors from inflammatory nodules and the one performed best was selected out.
RESULTS: Model 4, which incorporated the nodule and 15 mm peri-nodular region, best differentiated benign and malignant SPNs. The model had an area under the curve (AUC), accuracy, recall, precision, and F1-score of 0.730, 0.724, 0.711, 0.705, and 0.707 in the external validation cohort. Model 4 also performed better than the other four conventional algorithms. Model 8, which incorporated the nodule and 10 mm peri-nodular region, was the best model for distinguishing benign tumors from inflammatory nodules. The model had an AUC, accuracy, recall, precision, and F1-score of 0.871, 0.938, 0.863, 0.904, and 0.882 in the external validation cohort.
CONCLUSION: The study concludes that CT-based DL models built with MDANet can accurately discriminate among small benign and malignant SPNs, benign tumors and inflammatory nodules.
引用
- Li Y, Huang XT, Feng YB, et al. Value of CT-Based Deep Learning Model in Differentiating Benign and Malignant Solid Pulmonary Nodules ≤ 8 mm. Acad Radiol. 2024. PMID: 38806374. doi:10.1016/j.acra.2024.05.021
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。