From the Artificial Intelligence in Medicine (AIM) Program, Mass General Brigham, Harvard Medical School, Boston, Mass (A.B., Z.Y., Y.Z., A.Z., H.H., R.C., H.J.W.L.A., B.H.K.); Department of Radiation Oncology (A.B., Z.Y., M.C.T., Y.Z., A.Z., H.H., R.C., K.X.L., D.A.H.K., H.J.W.L.A., B.H.K.) and Department of Radiology (H.J.W.L.A.), Brigham and Women's Hospital, Dana-Farber Cancer Institute, Harvard Medical School, 75 Francis St, Boston, MA 02115; Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, Mass (S.P.P., S.V., T.Y.P.); Department of Biostatistics and Computational Biology, Harvard T.H. Chan School of Public Health, Boston, Mass (P.J.C.); Center for Data-Driven Discovery in Biomedicine (D3b) (A.N., A.C.R.) and Department of Neurosurgery (A.C.R.), Children's Hospital of Philadelphia, Philadelphia, Pa; Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa (A.N.); Departments of Neurology, Pediatrics, and Neurologic Surgery, University of California, San Francisco, San Francisco, Calif (S.M.); and Department of Radiology and Nuclear Medicine, CARIM & GROW, Maastricht University, Maastricht, the Netherlands (H.J.W.L.A.).
Radiol Artif Intell. 2024 Jul;6(4):e230254. doi: 10.1148/ryai.230254.
Purpose To develop, externally test, and evaluate clinical acceptability of a deep learning pediatric brain tumor segmentation model using stepwise transfer learning. Materials and Methods In this retrospective study, the authors leveraged two T2-weighted MRI datasets (May 2001 through December 2015) from a national brain tumor consortium ( = 184; median age, 7 years [range, 1-23 years]; 94 male patients) and a pediatric cancer center ( = 100; median age, 8 years [range, 1-19 years]; 47 male patients) to develop and evaluate deep learning neural networks for pediatric low-grade glioma segmentation using a stepwise transfer learning approach to maximize performance in a limited data scenario. The best model was externally tested on an independent test set and subjected to randomized blinded evaluation by three clinicians, wherein they assessed clinical acceptability of expert- and artificial intelligence (AI)-generated segmentations via 10-point Likert scales and Turing tests. Results The best AI model used in-domain stepwise transfer learning (median Dice score coefficient, 0.88 [IQR, 0.72-0.91] vs 0.812 [IQR, 0.56-0.89] for baseline model; = .049). With external testing, the AI model yielded excellent accuracy using reference standards from three clinical experts (median Dice similarity coefficients: expert 1, 0.83 [IQR, 0.75-0.90]; expert 2, 0.81 [IQR, 0.70-0.89]; expert 3, 0.81 [IQR, 0.68-0.88]; mean accuracy, 0.82). For clinical benchmarking ( = 100 scans), experts rated AI-based segmentations higher on average compared with other experts (median Likert score, 9 [IQR, 7-9] vs 7 [IQR 7-9]) and rated more AI segmentations as clinically acceptable (80.2% vs 65.4%). Experts correctly predicted the origin of AI segmentations in an average of 26.0% of cases. Conclusion Stepwise transfer learning enabled expert-level automated pediatric brain tumor autosegmentation and volumetric measurement with a high level of clinical acceptability. Stepwise Transfer Learning, Pediatric Brain Tumors, MRI Segmentation, Deep Learning . © RSNA, 2024.
目的 使用逐步迁移学习开发、外部测试和评估用于小儿脑肿瘤分割的深度学习模型的临床可接受性。
材料与方法 本回顾性研究利用来自国家脑肿瘤联盟的两个 T2 加权 MRI 数据集（2001 年 5 月至 2015 年 12 月）（n = 184；中位年龄 7 岁[范围，1-23 岁]；94 名男性患者）和一家儿科癌症中心（n = 100；中位年龄 8 岁[范围，1-19 岁]；47 名男性患者），开发和评估使用逐步迁移学习方法的小儿低级别胶质瘤分割的深度学习神经网络，以在有限数据情况下最大限度地提高性能。最佳模型在独立测试集上进行外部测试，并由三位临床医生进行随机盲法评估，他们通过 10 分李克特量表和图灵测试评估专家和人工智能（AI）生成的分割的临床可接受性。
结果 最佳 AI 模型使用了基于域的逐步迁移学习（中位数 Dice 评分系数为 0.88 [四分位距，0.72-0.91] vs 基线模型的 0.812 [四分位距，0.56-0.89]； =.049）。通过外部测试，该 AI 模型使用三位临床专家的参考标准得出了出色的准确性（专家 1 的中位 Dice 相似系数：0.83 [四分位距，0.75-0.90]；专家 2，0.81 [四分位距，0.70-0.89]；专家 3，0.81 [四分位距，0.68-0.88]；平均准确率，0.82）。对于临床基准测试（n = 100 次扫描），专家对基于 AI 的分割的平均评分高于其他专家（中位数李克特评分，9 [四分位距，7-9] vs 7 [四分位距 7-9]），并且认为更多的 AI 分割具有临床可接受性（80.2%比 65.4%）。专家平均正确预测了 AI 分割的起源在 26.0%的病例中。
结论 逐步迁移学习使小儿脑肿瘤自动分割和体积测量达到专家水平，并具有高度的临床可接受性。
逐步迁移学习，小儿脑肿瘤，MRI 分割，深度学习。
© 2024 RSNA。
