Multi-task Learning for Predicting Alzheimer's Disease Progression

阿尔茨海默病(AD)是最常见的痴呆症，是一种严重的神经退行性疾病。寻找能够跟踪疾病进展的生物标志物在AD的研究中受到了极大的关注。对疾病进展的准确预测将有助于临床医生和患者做出最优决策。本项目的目的是探索多任务学习方法，通过认知分数来预测AD进展，并选择预测进展的生物标志物。本项目将通过考虑时态组套索方法将预测问题描述为多任务回归问题，时态组套索方法是一种多任务、正规化的方法，用于预测随时间变化的响应变量。本项目的目标包括：1)研究时态组套索及其相关方法背后的概念，以及在AD研究中的潜在应用类型。2)建立并实现了一种基于时间组套索的AD进度模型。3)使用阿尔茨海默病神经成像倡议(ADNI)的数据对该模型进行评估。4)分析和探讨其未来潜力。

Alzheimer's disease (AD), the most common type of dementia, is a severe neurodegenerative disorder. Identifying biomarkers that can track the progress of the disease has received much attentions in AD research. An accurate prediction of disease progression would facilitate optimal decision-making for clinicians and patients.The aim of this project is to investigate multi-task learning approaches into predicting AD progression measured by the cognitive scores and selecting biomarkers predictive of the progression. This project will formulate the prediction problem as a multi-task regression problem by considering the approach of Temporal Group Lasso, which is a multi-task, regularised approach for the prediction of response variables that vary over time.This objectives of this project include: 1) to investigate the concepts behind the Temporal Group LASSO and its related methods, as well as the type of potential applications in AD research. 2) to build up and implement a Temporal Group Lasso based AD progression model. 3) to evaluate the model using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). 4) to analysis and discuss its future potential.