多模式深度学习改善了儿科低级别胶质瘤的复发风险预测。
Multimodal Deep Learning Improves Recurrence Risk Prediction in Pediatric Low-Grade Gliomas.
发表日期:2024 Aug 30
作者:
Maryamalsadat Mahootiha, Divyanshu Tak, Zezhong Ye, Anna Zapaishchykova, Jirapat Likitlersuang, Juan Carlos Climent Pardo, Aidan Boyd, Sridhar Vajapeyam, Rishi Chopra, Sanjay P Prabhu, Kevin X Liu, Hesham Elhalawani, Ali Nabavizadeh, Ariana Familiar, Sabine Mueller, Hugo J W L Aerts, Pratiti Bandopadhayay, Keith L Ligon, Daphne Haas-Kogan, Tina Y Poussaint, Hemin Ali Qadir, Ilangko Balasingham, Benjamin H Kann
来源:
NEURO-ONCOLOGY
摘要:
儿科低级别胶质瘤 (pLGG) 的术后复发风险很难通过传统的临床、放射学和基因组因素来预测。我们研究了 MRI 肿瘤特征的深度学习是否可以改善术后 pLGG 风险分层。我们使用专为 pLGG 分割而设计的预训练深度学习 (DL) 工具,从接受手术的患者的术前 T2 加权 MRI 中提取 pLGG 成像特征 (DL- MRI 特征)。患者来自两个机构:达纳法伯/波士顿儿童医院 (DF/BCH) 和儿童脑肿瘤网络 (CBTN)。我们训练了三个 DL 逻辑风险模型来预测术后无事件生存 (EFS) 概率,其中包括 1) 临床特征、2) DL-MRI 特征和 3) 多模态(临床和 DL-MRI 特征)。我们使用时间相关的一致性指数 (Ctd) 评估模型,并使用 Kaplan Meier 图和对数秩检验进行风险组分层。我们开发了一个将 pLGG 分割和 EFS 预测与最佳模型相结合的自动化流程。在分析的 396 名患者中(中位随访时间:85 个月,范围:1.5-329 个月),214 名患者(54%)接受了总切除术,110 名患者接受了总切除术。 28%)复发。与 DL-MRI 和临床模型相比,多模态模型改善了 EFS 预测(Ctd:0.85(95% CI:0.81-0.93)、0.79(95% CI:0.70-0.88)和 0.72(95% CI:0.57-0.77) ), 分别)。多模式模型改善了风险组分层(预测高风险的 3 年 EFS:31% 与低风险:92%,p<0.0001)。DL 提取影像特征,可以为 pLGG 术后复发预测提供信息。多模式 DL 改善了 pLGG 的术后风险分层,并可以指导术后决策。可能需要更大的多中心训练数据来提高模型的通用性。© 作者 2024。由牛津大学出版社代表神经肿瘤学会出版。
Postoperative recurrence risk for pediatric low-grade gliomas (pLGGs) is challenging to predict by conventional clinical, radiographic, and genomic factors. We investigated if deep learning of MRI tumor features could improve postoperative pLGG risk stratification.We used pre-trained deep learning (DL) tool designed for pLGG segmentation to extract pLGG imaging features from preoperative T2-weighted MRI from patients who underwent surgery (DL-MRI features). Patients were pooled from two institutions: Dana Farber/Boston Children's Hospital (DF/BCH) and the Children's Brain Tumor Network (CBTN). We trained three DL logistic hazard models to predict postoperative event-free survival (EFS) probabilities with 1) clinical features, 2) DL-MRI features, and 3) multimodal (clinical and DL-MRI features). We evaluated the models with a time-dependent Concordance Index (Ctd) and risk group stratification with Kaplan Meier plots and log-rank tests. We developed an automated pipeline integrating pLGG segmentation and EFS prediction with the best model.Of the 396 patients analyzed (median follow-up: 85 months, range: 1.5-329 months), 214 (54%) underwent gross total resection and 110 (28%) recurred. The multimodal model improved EFS prediction compared to the DL-MRI and clinical models (Ctd: 0.85 (95% CI: 0.81-0.93), 0.79 (95% CI: 0.70-0.88), and 0.72 (95% CI: 0.57-0.77), respectively). The multimodal model improved risk-group stratification (3-year EFS for predicted high-risk: 31% versus low-risk: 92%, p<0.0001).DL extracts imaging features that can inform postoperative recurrence prediction for pLGG. Multimodal DL improves postoperative risk stratification for pLGG and may guide postoperative decision-making. Larger, multicenter training data may be needed to improve model generalizability.© The Author(s) 2024. Published by Oxford University Press on behalf of the Society for Neuro-Oncology.