18F-FDG PET/CT鉴别淋巴瘤与结节病:影像组学特征引导的机器学习与人类阅片者表现的评估
该研究回顾性纳入420例(169例结节病、140例霍奇金淋巴瘤、111例弥漫大B细胞淋巴瘤)治疗前18F-FDG PET/CT,比较4名医师诊断表现与基于215个影像组学特征联合临床数据的机器学习模型。机器学习模型鉴别淋巴瘤与结节病的AUC在病灶层面为0.94、患者层面为0.85,鉴别霍奇金淋巴瘤与弥漫大B细胞淋巴瘤的AUC分别为0.95和0.86,表现与医师相当或更优。
要点
- 医师汇总鉴别淋巴瘤与结节病的灵敏度0.99、特异度0.75、AUC 0.87、准确率89.3%。
- 医师鉴别霍奇金淋巴瘤与弥漫大B细胞淋巴瘤的灵敏度0.58、特异度0.82、AUC 0.70、准确率68.5%。
- 最佳机器学习模型鉴别淋巴瘤与结节病病灶层面AUC 0.94、患者层面AUC 0.85。
核心结论
- 利用机器学习和影像组学表征结节病与淋巴瘤病灶是可行的,表现非常好至极好,等同于或优于医师,而医师评估显示显著的观察者间变异。
研究设计与样本
- 回顾性收集在列日大学医院接受治疗前18F-FDG PET/CT的420例患者,其中结节病169例、霍奇金淋巴瘤140例、弥漫大B细胞淋巴瘤111例。
- 研究随机分配给4名医师给出3种疾病的诊断建议,并计算个体和汇总表现;使用34项研究由所有医师解读以评估观察者间变异。
- 在病灶和肝脏上勾画感兴趣体积,使用RadiomiX工具箱提取215个影像组学特征。
- 结合临床数据(年龄、性别和体重)与影像组学(原始及肿瘤-肝脏TLR影像组学),采用7种特征选择方法和4种机器学习分类器,在病灶层面和患者层面建立模型。
主要发现
- 医师汇总鉴别淋巴瘤与结节病的灵敏度0.99(95%CI 0.97-1.00)、特异度0.75(95%CI 0.68-0.81)、AUC 0.87(95%CI 0.84-0.90)、准确率89.3%。
- 医师在肿瘤人群中鉴别霍奇金淋巴瘤的灵敏度0.58(95%CI 0.49-0.66)、特异度0.82(95%CI 0.74-0.89)、AUC 0.70(95%CI 0.64-0.75)、准确率68.5%。
- 观察者间诊断淋巴瘤与结节病、霍奇金淋巴瘤与弥漫大B细胞淋巴瘤的Fleiss κ值分别为0.66(95%CI 0.45-0.87)和0.69(95%CI 0.45-0.93),为中等一致性。
- 鉴别淋巴瘤与结节病的最佳机器学习模型在病灶层面和患者层面的AUC分别为0.94(95%CI 0.93-0.95)和0.85(95%CI 0.82-0.88),使用TLR影像组学(患者层面模型加年龄)。
- 鉴别霍奇金淋巴瘤与弥漫大B细胞淋巴瘤,病灶层面使用TLR影像组学的AUC为0.95(95%CI 0.93-0.96),患者层面使用原始影像组学加年龄的AUC为0.86(95%CI 0.80-0.91)。
关键数值与推荐
- 医师汇总鉴别淋巴瘤与结节病:灵敏度0.99(95%CI 0.97-1.00)、特异度0.75(95%CI 0.68-0.81)、AUC 0.87(95%CI 0.84-0.90)、准确率89.3%。
- 医师在肿瘤人群中鉴别霍奇金淋巴瘤:灵敏度0.58(95%CI 0.49-0.66)、特异度0.82(95%CI 0.74-0.89)、AUC 0.70(95%CI 0.64-0.75)、准确率68.5%。
- 观察者间一致性Fleiss κ:淋巴瘤与结节病0.66(95%CI 0.45-0.87),霍奇金淋巴瘤与弥漫大B细胞淋巴瘤0.69(95%CI 0.45-0.93)。
- 最佳机器学习模型鉴别淋巴瘤与结节病:病灶层面AUC 0.94(95%CI 0.93-0.95),患者层面AUC 0.85(95%CI 0.82-0.88)。
- 机器学习模型鉴别霍奇金淋巴瘤与弥漫大B细胞淋巴瘤:病灶层面AUC 0.95(95%CI 0.93-0.96),患者层面AUC 0.86(95%CI 0.80-0.91)。
临床怎么用
- 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)。
原文
Sarcoidosis and lymphoma often share common features on 18F-FDG PET/CT, such as intense hypermetabolic lesions in lymph nodes and multiple organs. We aimed at developing and validating radiomics signatures to differentiate sarcoidosis from Hodgkin lymphoma (HL) and diffuse large B-cell lymphoma (DLBCL). Methods: We retrospectively collected 420 patients (169 sarcoidosis, 140 HL, and 111 DLBCL) who underwent pretreatment 18F-FDG PET/CT at the University Hospital of Liege. The studies were randomly distributed to 4 physicians, who gave their diagnostic suggestion among the 3 diseases. The individual and pooled performance of the physicians was then calculated. Interobserver variability was evaluated using a sample of 34 studies interpreted by all physicians. Volumes of interest were delineated over the lesions and the liver using MIM software, and 215 radiomics features were extracted using the RadiomiX Toolbox. Models were developed combining clinical data (age, sex, and weight) and radiomics (original and tumor-to-liver TLR radiomics), with 7 different feature selection approaches and 4 different machine-learning (ML) classifiers, to differentiate sarcoidosis and lymphomas on both lesion-based and patient-based approaches. Results: For identifying lymphoma versus sarcoidosis, physicians' pooled sensitivity, specificity, area under the receiver-operating-characteristic curve (AUC), and accuracy were 0.99 (95% CI, 0.97-1.00), 0.75 (95% CI, 0.68-0.81), 0.87 (95% CI, 0.84-0.90), and 89.3%, respectively, whereas for identifying HL in the tumor population, it was 0.58 (95% CI, 0.49-0.66), 0.82 (95% CI, 0.74-0.89), 0.70 (95% CI, 0.64-0.75) and 68.5%, respectively. Moderate agreement was found among observers for the diagnosis of lymphoma versus sarcoidosis and HL versus DLBCL, with Fleiss κ-values of 0.66 (95% CI, 0.45-0.87) and 0.69 (95% CI, 0.45-0.93), respectively. The best ML models for identifying lymphoma versus sarcoidosis showed an AUC of 0.94 (95% CI, 0.93-0.95) and 0.85 (95% CI, 0.82-0.88) in lesion- and patient-based approaches, respectively, using TLR radiomics (plus age for the second). To differentiate HL from DLBCL, we obtained an AUC of 0.95 (95% CI, 0.93-0.96) in the lesion-based approach using TLR radiomics and 0.86 (95% CI, 0.80-0.91) in the patient-based approach using original radiomics and age. Conclusion: Characterization of sarcoidosis and lymphoma lesions is feasible using ML and radiomics, with very good to excellent performance, equivalent to or better than that of physicians, who showed significant interobserver variability in their assessment.
引用
- Lovinfosse P, Ferreira M, Withofs N, et al. Distinction of Lymphoma from Sarcoidosis on 18F-FDG PET/CT: Evaluation of Radiomics-Feature-Guided Machine Learning Versus Human Reader Performance. J Nucl Med. 2022. PMID: 35589406. doi:10.2967/jnumed.121.263598
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。