多中心评估人工智能生成的 DIR 和 PSIR 用于皮质和皮质下多发性硬化病变检测。
Multicenter Evaluation of AI-generated DIR and PSIR for Cortical and Juxtacortical Multiple Sclerosis Lesion Detection.
机构信息
From the MS Center Amsterdam, Anatomy & Neurosciences, Amsterdam Neuroscience, Amsterdam UMC, Vrije Universiteit Amsterdam, De Boelelaan 1117, Amsterdam, the Netherlands (P.M.B., S.N., F.A.N.S., M.M.S., J.J.G.G., M.D.S.); Translational Neuroradiology Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Md (E.S.B., D.S.R.); Department of Neurology, Icahn School of Medicine at Mount Sinai, New York, NY (E.S.B.); Bio-imaging Institute, University of Bordeaux, Bordeaux, France (G.B.); Neurology Section, Department of Neuroscience, Biomedicine and Movement Sciences, University of Verona, Verona, Italy (M. Castellaro, M. Calabrese); Department of Information Engineering, University of Padova, Padova, Italy (M. Castellaro); NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK (D.T.C.); National Institute for Health Research University College London Hospitals Biomedical Research Centre, London, UK (D.T.C.); Departments of Neuroradiology (P.E., B.W.) and Neurology (M.M.), School of Medicine, Klinikum Rechts der Isar, Technical University of Munich, Munich, Germany; Neuroimaging Research Unit, Division of Neuroscience Neurology Unit, IRCCS San Raffaele Scientific Institute Vita-Salute San Raffaele University, Milan, Italy (M.F., P.P., M.A.R.); Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genova, Genoa, Italy (M.I., C.L.); IRCCS Ospedale Policlinico San Martino, Largo Rosanna Benzi, Genoa, Italy (M.I., C.L.); Center for Neurologic Imaging, Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, Mass (A.M., A.M.P., C.R.G.G.); Department of Radiology and Nuclear Medicine, MS Center Amsterdam, Amsterdam Neurosciences, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands (B.M.); Department of Epidemiology and Data Science, Amsterdam University Medical Center, Amsterdam, the Netherlands (J.W.R.T.); Anatomy & Neurosciences, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands (L.E.J.); and Amsterdam Neuroscience, Brain Imaging and Neurodegeneration, Amsterdam, the Netherlands (L.E.J.).
出版信息
Radiology. 2023 Apr;307(2):e221425. doi: 10.1148/radiol.221425. Epub 2023 Feb 7.
Background Cortical multiple sclerosis lesions are clinically relevant but inconspicuous at conventional clinical MRI. Double inversion recovery (DIR) and phase-sensitive inversion recovery (PSIR) are more sensitive but often unavailable. In the past 2 years, artificial intelligence (AI) was used to generate DIR and PSIR from standard clinical sequences (eg, T1-weighted, T2-weighted, and fluid-attenuated inversion-recovery sequences), but multicenter validation is crucial for further implementation. Purpose To evaluate cortical and juxtacortical multiple sclerosis lesion detection for diagnostic and disease monitoring purposes on AI-generated DIR and PSIR images compared with MRI-acquired DIR and PSIR images in a multicenter setting. Materials and Methods Generative adversarial networks were used to generate AI-based DIR ( = 50) and PSIR ( = 43) images. The number of detected lesions between AI-generated images and MRI-acquired (reference) images was compared by randomized blinded scoring by seven readers (all with >10 years of experience in lesion assessment). Reliability was expressed as the intraclass correlation coefficient (ICC). Differences in lesion subtype were determined using Wilcoxon signed-rank tests. Results MRI scans of 202 patients with multiple sclerosis (mean age, 46 years ± 11 [SD]; 127 women) were retrospectively collected from seven centers (February 2020 to January 2021). In total, 1154 lesions were detected on AI-generated DIR images versus 855 on MRI-acquired DIR images (mean difference per reader, 35.0% ± 22.8; < .001). On AI-generated PSIR images, 803 lesions were detected versus 814 on MRI-acquired PSIR images (98.9% ± 19.4; = .87). Reliability was good for both DIR (ICC, 0.81) and PSIR (ICC, 0.75) across centers. Regionally, more juxtacortical lesions were detected on AI-generated DIR images than on MRI-acquired DIR images (495 [42.9%] vs 338 [39.5%]; < .001). On AI-generated PSIR images, fewer juxtacortical lesions were detected than on MRI-acquired PSIR images (232 [28.9%] vs 282 [34.6%]; = .02). Conclusion Artificial intelligence-generated double inversion-recovery and phase-sensitive inversion-recovery images performed well compared with their MRI-acquired counterparts and can be considered reliable in a multicenter setting, with good between-reader and between-center interpretative agreement. Published under a CC BY 4.0 license. See also the editorial by Zivadinov and Dwyer in this issue.
背景 皮质多发性硬化病变具有临床相关性，但在常规临床 MRI 上并不明显。双反转恢复（DIR）和相敏反转恢复（PSIR）更为敏感，但通常无法获得。在过去的 2 年中，人工智能（AI）被用于从标准临床序列（例如 T1 加权、T2 加权和液体衰减反转恢复序列）生成 DIR 和 PSIR，但多中心验证对于进一步实施至关重要。目的 评估皮质和皮质下多发性硬化病变的检测，以用于诊断和疾病监测目的，比较在多中心环境下基于 AI 生成的 DIR（=50）和 PSIR（=43）图像与 MRI 获得的 DIR 和 PSIR 图像。通过七位读者（均具有 >10 年的病变评估经验）进行随机盲评分，比较 AI 生成图像与 MRI 获得（参考）图像之间的病变检出数量。可靠性表示为组内相关系数（ICC）。使用 Wilcoxon 符号秩检验确定病变亚型的差异。结果 回顾性收集了来自七个中心的 202 名多发性硬化症患者（平均年龄，46 岁±11[标准差]；127 名女性）的 MRI 扫描（2020 年 2 月至 2021 年 1 月）。总共在 AI 生成的 DIR 图像上检测到 1154 个病变，而在 MRI 获得的 DIR 图像上检测到 855 个病变（每位读者的平均差异，35.0%±22.8；<0.001）。在 AI 生成的 PSIR 图像上，检测到 803 个病变，而在 MRI 获得的 PSIR 图像上检测到 814 个病变（98.9%±19.4；=0.87）。跨中心的 DIR（ICC，0.81）和 PSIR（ICC，0.75）的可靠性均良好。在区域上，AI 生成的 DIR 图像上比 MRI 获得的 DIR 图像上检测到更多的皮质下病变（495[42.9%]与 338[39.5%]；<0.001）。在 AI 生成的 PSIR 图像上，检测到的皮质下病变比 MRI 获得的 PSIR 图像少（232[28.9%]与 282[34.6%]；=0.02）。结论 与 MRI 获得的图像相比，人工智能生成的双反转恢复和相敏反转恢复图像表现良好，在多中心环境下可以被认为是可靠的，具有良好的读者间和中心间解释一致性。在知识共享署名 4.0 许可下发布。请参阅本期 Zivadinov 和 Dwyer 的社论。
Zotero 插件