利用图神经网络检测致痫性局灶性皮质发育不良:MELD研究
这项多中心诊断研究基于2018至2022年23个癫痫中心的回顾性MRI数据,训练图神经网络MELD Graph自动检测局灶性皮质发育不良。测试集中阳性预测值为67%(敏感度70%、特异度60%),优于基线算法的39%(敏感度67%、特异度54%);独立测试队列阳性预测值为76%(敏感度72%、特异度56%),优于基线算法的46%(敏感度77%、特异度47%)。
要点
- MELD Graph在测试集检测FCD的阳性预测值为67%,基线算法为39%。
- 组织病理确诊且术后1年无发作患者中,MELD Graph敏感度为81.6%。
- MRI阴性FCD患者中,MELD Graph敏感度为63.7%。
核心结论
- MELD Graph是开放获取且可解释的FCD检测工具,阳性预测值较基线算法显著提高。
- 作者认为其临床实施有望用于局灶性癫痫的早期诊断和改善管理,可能带来更好的患者结局。
研究设计与样本
- 多中心诊断研究,回顾性收集2018至2022年全球23个癫痫中心的MRI数据,2023年分析。
- 20个中心的数据等分为训练集和测试集,3个中心的数据保留用于站点独立测试。
- 训练图神经网络MELD Graph基于表面特征识别FCD,并与现有算法比较。
- 共收集703例FCD相关癫痫患者和482例对照的34个表面MRI特征及手动病灶掩膜,57例在MRI质量控制中被排除。
主要发现
- 测试集260例(女125例[48%]、男135例[52%],平均年龄18.0岁[IQR 11.0-29.0])中,MELD Graph阳性预测值67%(敏感度70%、特异度60%),基线算法39%(敏感度67%、特异度54%)。
- 独立测试队列116例(女62例[53%]、男54例[47%],平均年龄22.5岁[IQR 13.5-27.5])中,MELD Graph阳性预测值76%(敏感度72%、特异度56%),基线算法46%(敏感度77%、特异度47%)。
- 组织病理确诊且术后1年无发作患者中,MELD Graph敏感度81.6%。
- MRI阴性FCD患者中,MELD Graph敏感度63.7%。
- 可解释报告可描述病灶位置、大小、置信度和显著特征。
关键数值与推荐
- 测试集MELD Graph阳性预测值67%(敏感度70%、特异度60%),基线算法39%(敏感度67%、特异度54%)。
- 独立测试队列MELD Graph阳性预测值76%(敏感度72%、特异度56%),基线算法46%(敏感度77%、特异度47%)。
- 组织病理确诊且术后1年无发作患者中MELD Graph敏感度81.6%,MRI阴性FCD患者中敏感度63.7%。
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
IMPORTANCE: A leading cause of surgically remediable, drug-resistant focal epilepsy is focal cortical dysplasia (FCD). FCD is challenging to visualize and often considered magnetic resonance imaging (MRI) negative. Existing automated methods for FCD detection are limited by high numbers of false-positive predictions, hampering their clinical utility. OBJECTIVE: To evaluate the efficacy and interpretability of graph neural networks in automatically detecting FCD lesions on MRI scans. DESIGN, SETTING, AND PARTICIPANTS: In this multicenter diagnostic study, retrospective MRI data were collated from 23 epilepsy centers worldwide between 2018 and 2022, as part of the Multicenter Epilepsy Lesion Detection (MELD) Project, and analyzed in 2023. Data from 20 centers were split equally into training and testing cohorts, with data from 3 centers withheld for site-independent testing. A graph neural network (MELD Graph) was trained to identify FCD on surface-based features. Network performance was compared with an existing algorithm. Feature analysis, saliencies, and confidence scores were used to interpret network predictions. In total, 34 surface-based MRI features and manual lesion masks were collated from participants, 703 patients with FCD-related epilepsy and 482 controls, and 57 participants were excluded during MRI quality control. MAIN OUTCOMES AND MEASURES: Sensitivity, specificity, and positive predictive value (PPV) of automatically identified lesions. RESULTS: In the test dataset, the MELD Graph had a sensitivity of 81.6% in histopathologically confirmed patients seizure-free 1 year after surgery and 63.7% in MRI-negative patients with FCD. The PPV of putative lesions from the 260 patients in the test dataset (125 female [48%] and 135 male [52%]; mean age, 18.0 [IQR, 11.0-29.0] years) was 67% (70% sensitivity; 60% specificity), compared with 39% (67% sensitivity; 54% specificity) using an existing baseline algorithm. In the independent test cohort (116 patients; 62 female [53%] and 54 male [47%]; mean age, 22.5 [IQR, 13.5-27.5] years), the PPV was 76% (72% sensitivity; 56% specificity), compared with 46% (77% sensitivity; 47% specificity) using the baseline algorithm. Interpretable reports characterize lesion location, size, confidence, and salient features. CONCLUSIONS AND RELEVANCE: In this study, the MELD Graph represented a state-of-the-art, openly available, and interpretable tool for FCD detection on MRI scans with significant improvements in PPV. Its clinical implementation holds promise for early diagnosis and improved management of focal epilepsy, potentially leading to better patient outcomes.
引用
- Ripart M, Spitzer H, Williams LZJ, et al. Detection of Epileptogenic Focal Cortical Dysplasia Using Graph Neural Networks: A MELD Study. JAMA Neurol. 2025. PMID: 39992650. doi:10.1001/jamaneurol.2024.5406
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。