Digital Health Hub, Simon Fraser University, 4190 Galleria 4, 250 - 13450 102 Ave, Surrey, BC, V3T 0A3, Canada.
School of Interactive Arts and Technology, Simon Fraser University, 250 - 13450 102 Ave, Surrey, BC, V3T 0A3, Canada.
Syst Rev. 2020 Apr 2;9(1):71. doi: 10.1186/s13643-020-01332-7.
Alzheimer's disease (AD) is a neurodegenerative disorder commonly associated with deficits of cognition and changes in behavior. Mild cognitive impairment (MCI) is the prodromal stage of AD that is defined by slight cognitive decline. Not all with MCI progress to AD dementia. Thus, the accurate prediction of progression to Alzheimer's, particularly in the stage of MCI could potentially offer developing treatments to delay or prevent the transition process. The objective of the present study is to investigate the most recent neuroimaging procedures in the domain of prediction of transition from MCI to AD dementia for clinical applications and to systematically discuss the machine learning techniques used for the prediction of MCI conversion.
Electronic databases including PubMed, SCOPUS, and Web of Science will be searched from January 1, 2017, to the date of search commencement to provide a rapid review of the most recent studies that have investigated the prediction of conversion from MCI to Alzheimer's using neuroimaging modalities in randomized trial or observational studies. Two reviewers will screen full texts of included papers using predefined eligibility criteria. Studies will be included if addressed research on AD dementia and MCI, explained the results in a way that would be able to report the performance measures such as the accuracy, sensitivity, and specificity. Only studies addressed Alzheimer's type of dementia and its early-stage MCI using neuroimaging modalities will be included. We will exclude other forms of dementia such as vascular dementia, frontotemporal dementia, and Parkinson's disease. The risk of bias in individual studies will be appraised using an appropriate tool. If feasible, we will conduct a random effects meta-analysis. Sensitivity analyses will be conducted to explore the potential sources of heterogeneity.
The information gathered in our study will establish the extent of the evidence underlying the prediction of conversion to AD dementia from its early stage and will provide a rigorous and updated synthesis of neuroimaging modalities allied with the data analysis techniques used to measure the brain changes during the conversion process.
PROSPERO,CRD42019133402.
阿尔茨海默病（AD）是一种神经退行性疾病，通常与认知功能障碍和行为改变有关。轻度认知障碍（MCI）是 AD 的前驱阶段，定义为轻微的认知下降。并非所有 MCI 都会进展为 AD 痴呆。因此，准确预测向 AD 的进展，特别是在 MCI 阶段，可能为开发治疗方法提供机会，以延缓或阻止过渡过程。本研究的目的是调查在从 MCI 到 AD 痴呆的过渡预测领域中的最新神经影像学程序，并系统地讨论用于预测 MCI 转换的机器学习技术。
将从 2017 年 1 月 1 日至搜索开始日期搜索电子数据库，包括 PubMed、SCOPUS 和 Web of Science，以快速审查使用随机试验或观察性研究中的神经影像学模式预测 MCI 向 AD 转换的最新研究。两位审查员将使用预定义的纳入标准筛选纳入论文的全文。如果研究涉及 AD 痴呆和 MCI，解释了能够报告性能指标（如准确性、敏感性和特异性）的结果的研究将被纳入。仅将使用神经影像学模式解决 AD 痴呆及其早期 MCI 的研究纳入。我们将排除其他形式的痴呆，如血管性痴呆、额颞叶痴呆和帕金森病。将使用适当的工具评估个别研究的偏倚风险。如果可行，我们将进行随机效应荟萃分析。将进行敏感性分析以探索异质性的潜在来源。
我们研究中收集的信息将确定从早期阶段预测向 AD 痴呆转化的证据程度，并为预测向 AD 痴呆转化的神经影像学模式提供严格和最新的综合分析，并提供用于测量转化过程中大脑变化的数据分析技术。
PROSPERO，CRD42019133402。
