GenFAR：基于49246例多队列MRI通过深度学习得到的脑结构通用表征
GenFAR: A generalized representation of brain structure, derived from 49,246 multi-cohort MRIs via deep learning
另 9 家 · 查看机构详情
- University of Pennsylvania（宾夕法尼亚大学）
- University of Melbourne（墨尔本大学）
- University of Wisconsin School of Medicine and Public Health（威斯康星大学医学与公共卫生学院）
- CSIRO Health and Biosecurity（联邦科学与工业研究组织健康与生物安全部）
- CSIRO（联邦科学与工业研究组织）
- University of California, San Francisco（加利福尼亚大学旧金山分校）
- Washington University in St. Louis（圣路易斯华盛顿大学）
- University of Washington（华盛顿大学）
- Wake Forest School of Medicine（维克森林医学院）
- Johns Hopkins University School of Medicine（约翰霍普金斯大学医学院）
- Indiana University（印第安纳大学）
- University of Southern California（南加利福尼亚大学）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
GenFAR是基于49246例多队列MRI的模块化深度学习框架，通过17类任务训练得到通用脑表征，可提升二级预测器的样本效率与准确性
中文摘要 AI 辅助
神经成像深度学习模型大多针对单个任务开发，限制了跨应用的知识迁移。本文提出GenFAR，一种模块化深度学习框架，可从脑MRI中学习通用的、临床相关的特征。我们在11个队列的49246名个体上训练该模块化架构，使用涵盖认知、临床、诊断、人口统计学和生物标志物的17种不同分类与回归任务，得到聚合的、聚焦的特征集，捕捉丰富的、临床和生物学相关的脑表征。我们开发了一种顺序学习方法，其中任务逐步基于先前学习的表征构建。通过对5000种任务序列的分析，我们确定最优序列长度为6个任务，并引入Donor Score指标量化每个任务对下游性能的贡献。该分析揭示了5个始终表现优异的贡献任务（年龄、AD/MCI、MMSE、高血压、高脂血症），构成了我们顺序模型的基础。我们证明了所学表征在训练集之外的各类任务中的效用，可作为专用二级预测器的基础，还表明使用该学习到的特征表征可大幅提高二级深度学习训练任务和模型的样本效率，并提升其准确性。
英文摘要
Deep learning models for neuroimaging have largely been developed for individual tasks, limiting knowledge transfer across applications. Here we introduce GenFAR, a modular deep learning framework that learns general, clinically informed features from brain MRIs. We trained this modular architecture on 49,246 individuals across 11 cohorts, using 17 diverse classification and regression tasks spanning cognition, clinical, diagnosis, demographics, and biomarkers. This yields aggregated, focused feature sets that capture rich, clinically- and biologically-relevant brain representations. We developed a sequential learning approach where tasks progressively build on previously learned representations. Through an analysis of 5,000 task sequences, we identified an optimal sequence length of six tasks and introduced a Donor Score metric to quantify each task's contribution to downstream performance. This analysis revealed five consistently strong donor tasks (Age, AD/MCI, MMSE, Hypertension, Hyperlipidemia) that formed the base of our sequential model. We demonstrated the utility of our learned representation, in various tasks beyond those included in the training set, to serve as the foundation for specialized secondary predictors. We further showed that using the learned feature representation can substantially increase the sample efficiency of secondary deep learning training tasks and models, as well as improve their accuracy.
