Lu Pinya, Chen Mingfeng, Chen Lili, Lin Fan, Yang Hongqin, Wang Yuhua, Ding Xuemei
Fujian Provincial Engineering Research Center for Public Service Big Data Mining and Application, Fujian Provincial University Engineering Research Center for Big Data Analysis and Application, Key Laboratory of OptoElectronic Science and Technology for Medicine of Ministry of Education, Fujian Provincial Key Laboratory for Photonics Technology, Fujian Normal University, Fuzhou, China.
Department of Neurology, Fujian Provincial Hospital, Fuzhou University Affiliated Provincial Hospital, Fuzhou, China.
J Alzheimers Dis. 2025 May 25;106(2):13872877251343241. doi: 10.1177/13872877251343241.
BackgroundAlzheimer's disease (AD), marked by progressive memory loss and cognitive decline, poses diagnostic challenges due to its multifactorial nature. Therefore, researchers are increasingly leveraging artificial intelligence and data-driven approaches to develop computerized clinical decision support systems (CCDSS), aiming to enhance early detection, improve treatment, and slow disease progression.ObjectiveThis study seeks to conduct a systematic review of the most recently developed AD-CCDSS, delving into their progress and the challenges to guide future development and implementation of CCDSS for AD-related decision-making and intervention strategies.MethodsWe follow the PRISMA 2020 guideline to search for articles published within the past seven years across PubMed, ScienceDirect, IEEE Xplore Digital Library, Web of Science, and Scopus, with Google Scholar as a supplementary source. Key components are then extracted from the selected studies for qualitative analysis, including data modalities, computational modeling approaches, system explainability and interpretability, research priorities, and graphical user interfaces designed for non-technical stakeholders.ResultsAfter searching and removing duplicates, we meticulously selected 55 studies. After reviewing key components of CCDSS, we highlight advancements and potential clinical applications, demonstrating their promise in enhancing decision support. However, despite growing attention to explainability in AD-CCDSS, its clinical applicability remains limited. Moreover, challenges such as multi-center system interoperability and data security remain underexplored, hindering real-world implementation.ConclusionsThis study analyzes recent translational AD-CCDSS, identifying key challenges in advancing CCDSS for clinical applications. It offers insights for researchers to enhance CCDSS development and facilitate their integration into clinical practice.
背景
阿尔茨海默病(AD)以进行性记忆丧失和认知衰退为特征,因其多因素性质而带来诊断挑战。因此,研究人员越来越多地利用人工智能和数据驱动方法来开发计算机化临床决策支持系统(CCDSS),旨在加强早期检测、改善治疗并减缓疾病进展。
目的
本研究旨在对最近开发的AD - CCDSS进行系统评价,深入探讨其进展和挑战,以指导未来用于AD相关决策和干预策略的CCDSS的开发与实施。
方法
我们遵循PRISMA 2020指南,在PubMed、ScienceDirect、IEEE Xplore数字图书馆、Web of Science和Scopus上搜索过去七年发表的文章,并以谷歌学术作为补充来源。然后从选定研究中提取关键组件进行定性分析,包括数据模式、计算建模方法、系统可解释性和可理解性、研究重点以及为非技术利益相关者设计的图形用户界面。
结果
在搜索并去除重复项后,我们精心挑选了55项研究。在审查CCDSS的关键组件后,我们突出了进展和潜在临床应用,证明了它们在增强决策支持方面的前景。然而,尽管AD - CCDSS中对可解释性的关注日益增加,但其临床适用性仍然有限。此外,多中心系统互操作性和数据安全等挑战仍未得到充分探索,阻碍了实际应用。
结论
本研究分析了近期转化型AD - CCDSS,确定了推进CCDSS用于临床应用的关键挑战。它为研究人员增强CCDSS开发并促进其融入临床实践提供了见解。