Institute of Computational Biology, Helmholtz Center Munich, Munich 85764, Germany.
Department of Developmental Neuroscience, UCL Great Ormond Street Institute for Child Health, London WC1N 1EH, UK.
Brain. 2022 Nov 21;145(11):3859-3871. doi: 10.1093/brain/awac224.
One outstanding challenge for machine learning in diagnostic biomedical imaging is algorithm interpretability. A key application is the identification of subtle epileptogenic focal cortical dysplasias (FCDs) from structural MRI. FCDs are difficult to visualize on structural MRI but are often amenable to surgical resection. We aimed to develop an open-source, interpretable, surface-based machine-learning algorithm to automatically identify FCDs on heterogeneous structural MRI data from epilepsy surgery centres worldwide. The Multi-centre Epilepsy Lesion Detection (MELD) Project collated and harmonized a retrospective MRI cohort of 1015 participants, 618 patients with focal FCD-related epilepsy and 397 controls, from 22 epilepsy centres worldwide. We created a neural network for FCD detection based on 33 surface-based features. The network was trained and cross-validated on 50% of the total cohort and tested on the remaining 50% as well as on 2 independent test sites. Multidimensional feature analysis and integrated gradient saliencies were used to interrogate network performance. Our pipeline outputs individual patient reports, which identify the location of predicted lesions, alongside their imaging features and relative saliency to the classifier. On a restricted 'gold-standard' subcohort of seizure-free patients with FCD type IIB who had T1 and fluid-attenuated inversion recovery MRI data, the MELD FCD surface-based algorithm had a sensitivity of 85%. Across the entire withheld test cohort the sensitivity was 59% and specificity was 54%. After including a border zone around lesions, to account for uncertainty around the borders of manually delineated lesion masks, the sensitivity was 67%. This multicentre, multinational study with open access protocols and code has developed a robust and interpretable machine-learning algorithm for automated detection of focal cortical dysplasias, giving physicians greater confidence in the identification of subtle MRI lesions in individuals with epilepsy.
机器学习在诊断生物医学成像方面面临的一个突出挑战是算法的可解释性。一个关键的应用是从结构磁共振成像中识别细微的致癫痫灶性皮质发育不良（FCD）。FCD 在结构磁共振成像上难以可视化，但通常可以通过手术切除。我们旨在开发一种开源、可解释的基于表面的机器学习算法，以自动识别来自世界各地癫痫手术中心的异质结构磁共振成像数据中的 FCD。多中心癫痫病变检测（MELD）项目汇集并协调了来自全球 22 个癫痫中心的 1015 名参与者、618 名局灶性 FCD 相关癫痫患者和 397 名对照者的回顾性 MRI 队列。我们基于 33 个基于表面的特征创建了一个用于 FCD 检测的神经网络。该网络在总队列的 50%上进行训练和交叉验证，并在剩余的 50%以及 2 个独立的测试站点上进行测试。多维特征分析和集成梯度显着性用于检查网络性能。我们的管道输出患者的个别报告，识别预测病变的位置，以及其成像特征和相对于分类器的相对显着性。在一个有限的 FCD 类型 IIB 无癫痫发作患者的“金标准”亚队列中，该队列具有 T1 和液体衰减反转恢复 MRI 数据，MELD FCD 基于表面的算法的敏感性为 85%。在整个保留的测试队列中，敏感性为 59%，特异性为 54%。在包括病变周围的边界区域以解释手动勾画病变掩模边界的不确定性之后，敏感性为 67%。这项具有开放协议和代码的多中心、多国研究已经开发出一种用于自动检测局灶性皮质发育不良的强大且可解释的机器学习算法，为医生在识别癫痫患者的细微 MRI 病变时提供了更大的信心。
