人工智能辅助阿尔茨海默病诊断的研究进展
Research Progress on the Application of Artificial Intelligence in Alzheimer’s Disease Diagnosis
DOI: 10.12677/ns.2026.153069, PDF, HTML, XML,    科研立项经费支持
作者: 张 皓:陕西工业职业技术大学人事处,陕西 咸阳;张建安, 谢海蝶, 马瑜洁:西藏民族大学附属医院,陕西 咸阳;郑 瑶:西藏民族大学医学院,陕西 咸阳;王平义*:西藏民族大学附属医院,陕西 咸阳;西藏民族大学医学院,陕西 咸阳;中山大学附属第三医院康复医学科,广东 广州
关键词: 阿尔茨海默病早期诊断人工智能深度学习多模态Alzheimer’s Disease Early Diagnosis Artificial Intelligence Deep Learning; Multimodal
摘要: 阿尔茨海默病(AD)是神经退行性疾病,早期诊断对延缓病程具有关键意义,人工智能技术在AD诊断中的应用研究进展迅速。当前研究热点包括基于MRI与PET影像的AI分析、语言数据的机器学习评估以及脑电图信号的智能识别等,相关研究表明这些AI技术有助于提高AD早期诊断的准确性和效率,能够及早识别微弱病理特征。然而,目前仍面临数据获取与标准化不足导致的模型泛化性限制、模型可解释性欠缺影响临床信任等挑战,以及临床转化应用的困难。未来发展趋势侧重于多模态数据融合、算法优化并提高AI辅助诊断工具的临床可用性。
Abstract: Alzheimer’s disease (AD) is a neurodegenerative disease where early diagnosis is crucial for slowing its progression, and the application of artificial intelligence (AI) in AD diagnosis has advanced rapidly in recent years. Current research hotspots include AI-based analysis of MRI and PET scans, machine learning assessment of speech and language patterns, and AI-driven EEG signal analysis. Studies indicate that these AI technologies can improve the accuracy and efficiency of early AD diagnosis by enabling earlier detection of subtle pathological changes. However, challenges remain, such as insufficient data sharing and standardization limiting model generalizability, a lack of model interpretability undermining clinician trust, and difficulties in integrating these tools into clinical practice. Future trends focus on multi-modal data integration and algorithm optimization to develop more efficient, clinically usable AI diagnostic tools.
文章引用:张皓, 张建安, 谢海蝶, 郑瑶, 马瑜洁, 王平义. 人工智能辅助阿尔茨海默病诊断的研究进展[J]. 护理学, 2026, 15(3): 62-71. https://doi.org/10.12677/ns.2026.153069

1. 研究背景

阿尔茨海默病(Alzheimer’s disease, AD)是最常见的神经退行性疾病之一,具有起病隐匿、病程进展缓慢的特点,其早期诊断困难且误诊率较高[1] [2]。随着全球人口老龄化加剧,痴呆患者人数迅速增长,2021年全球痴呆患者已超过5500万,预计2050年将增至约1.39亿[1] [3]。其中,中国60岁以上人群中痴呆患者约有1500万,AD患者约983万,给社会和家庭带来沉重负担[4]。阿尔茨海默病是老年期痴呆最常见的类型,占痴呆病例的60%~70% [4]。其临床特征为进行性记忆衰退、认知功能下降,晚期可出现行为和人格改变[1]。AD的患病率随年龄显著升高,在84岁以上人群中患病率接近50% [5]。由于早期症状隐匿且与正常老化变化难以区分,临床上约有相当比例的早期AD或MCI患者未被及时诊断[6] [7]。传统的AD诊断主要依赖神经心理测验和临床评估,确诊需借助脑脊液、生物标志物检测或影像学检查。然而,这些“金标准”检查(如Aβ PET显像、脑脊液Tau蛋白检测)费用高、侵入性强或可及性差,在基层和发展中国家难以广泛开展[8] [9]。因此,如何利用新技术实现AD的早期、准确、低成本筛查成为重要研究方向。

人工智能技术的发展为AD的辅助诊断带来了新的契机。通过机器学习和深度学习算法,计算机可从海量临床数据中自动提取疾病特征模式,在一定程度上弥补人类专家主观判断的局限性[10] [11]。特别是在早期AD诊断中,AI有望整合多模态信息(如影像、认知测试、遗传和日常行为数据),提供综合的风险评估和决策支持[12] [13]。近年来,AD领域AI相关研究迅猛增长。文献计量学分析显示,自2015年以来每年发表的AI应用于AD诊断研究数量快速上升,至2023年已超过600篇/年[14] [15]。基于AI的诊断模型在区分正常老龄与MCI、预测MCI向AD转化、以及鉴别不同类型痴呆(如路易体痴呆、额颞叶痴呆等)方面均取得了令人瞩目的进展[16] [17]。有研究构建了整合多达10种痴呆病因的AI模型,对AD及相关痴呆的综合诊断AUC达到0.96 [18] [19];并且在模拟临床情境下,AI辅助可使神经科医生的诊断准确率提高约26% [20] [21]。这些成果表明,AI有潜力显著提升AD早期诊断的准确性与效率,对于应对日益增长的患者群体具有重要意义。

2. AI诊断方法分类及研究进展

目前,AI辅助AD诊断的研究主要集中在以下几个方面:影像学人工智能、语言与行为分析、脑电/脑磁图智能分析,以及多模态数据融合等。

2.1. 影像学AI诊断

脑影像能够直观反映AD的神经退行性改变,是AI应用最成熟的领域之一[8] [22]。结构磁共振成像(MRI)常用于测量颞叶内侧等脑萎缩情况,功能影像如正电子发射断层扫描(PET)可显示大脑葡萄糖代谢或淀粉样蛋白负荷的异常模式。基于影像的机器学习模型可以辅助区分AD患者与认知正常老年人,及预测MCI向AD的进展风险[1] [23]。传统方法如体积测量和放射状特征分析已用于AD诊断,但其敏感性有限且依赖专家经验[24] [25]。深度学习的兴起为影像分析提供了强大的工具。卷积神经网络(CNN)能够从MRI和PET影像中自动提取高维特征,无需手工选择ROI区域,即可训练出AD分类模型[5] [26]。例如,某研究利用ADNI数据库的T1加权MRI训练CNN模型,将AD与正常老年对照的分类准确率提高到接近98% [27] [28]。另有研究通过DenseNet网络直接处理未经预处理的原始三维MRI数据,实现了对AD和额颞叶痴呆(FTD)的自动区分,并能推广至不同种族和扫描仪的数据,表现出良好的跨中心泛化能力[29] [30]。在国内,有学者提出基于深度学习的MRI影像组学方法预测MCI向AD转化,其AD诊断准确率达到98.61%,MCI转化预测准确率达84.49% [31]。此外,融合多模态影像的信息可以进一步提升诊断性能。例如,Ferri等(2021年)将静息态脑电(EEG)与结构MRI特征结合,通过堆叠自动编码器模型实现对AD的分类,单独使用EEG特征准确率约80%,MRI约85%,融合后提高到89% [32]。另一项研究比较了不同影像模态和时间维度对诊断的影响,结果显示基于18F-FDG PET的模型优于纯MRI模型,引入纵向随访数据可显著提高模型性能[33] [34]

近年来Transformer也被引入影像分析领域,有研究将Vision Transformer应用于多中心MRI图像分类,取得不亚于CNN的准确率,同时具有更强的特征全局建模能力[35] [36]。从技术架构演进来看,影像学AI诊断已从依赖手工特征的传统机器学习(如SVM),发展到能够自动提取特征的卷积神经网络,再至具有全局建模能力的Vision Transformer。CNN因其在图像局部特征提取上的优势,目前仍是主流,但对训练数据规模要求高且可解释性较差。而Transformer模型凭借其自注意力机制,在处理多中心、多对比度影像时展现了更强的灵活性[37] [38]。未来,融合CNN的局部感知与Transformer的全局依赖的混合模型,或将成为提升模型稳定性和解释性的重要方向。总体而言,影像学AI诊断研究表明,深度学习模型能够捕捉AD脑影像的细微改变,其诊断准确性普遍高于传统机器学习和影像定量指标,为早期无创诊断提供了有力工具。

2.2. 语言与行为分析AI

AD患者在早期除了记忆减退,往往还伴随语言表达、日常行为方式的微妙变化。语言与行为数据作为“数字生物标志物”,近年来受到关注[39] [40]。自然语言处理(NLP)技术使得从患者的口语、叙事内容中提取认知特征成为可能。例如,在Alzheimer’s Dementia期刊发表的一项研究中,研究者采集了114名受试者的自发讲话,通过NLP模型分析其语法、语义和发音特征,成功将早期AD患者与认知正常者区分,敏感度和特异度均约为80% [6] 。特别是让受试者进行自传式访谈(回忆生活事件)比描述图片场景更能体现早期认知障碍,分类效果更佳[41] [42]。语言AI模型还可以用于预测轻度认知障碍进展为痴呆的风险。有研究表明,机器学习分析受试者在认知测试中的语言反应,可以比传统神经心理量表更早检测出认知滑坡的迹象[43] [44]。除口语外,书面语言、打字模式等也可作为数据源。一些应用通过分析患者写字绘画(如绘钟测验)的时序和压力等细节,由算法判别其认知状态,达到相当高的准确率[45]

在行为方面,AD早期常出现轻度的步态改变、睡眠紊乱和日常活动能力下降等。可穿戴设备和物联网技术的发展使长周期、连续的行为监测成为可能。AI模型可从中提取特征辅助诊断。例如,基于加速度计的步态分析研究发现,通过机器学习模型可以识别出遗忘型MCI与轻度AD患者的步态模式差异[46]。陶帅等(2022年)构建了步态识别模型,利用患者行走时下肢运动参数的机器学习分析,实现了对aMCI和AD的自动区分[47]。中国专家也发布了AD与帕金森病步态分析的共识,强调标准化步态参数提取和AI评估在辅助诊断中的作用[48]。除步态外,睡眠和日常活动也是观察窗口。研究显示佩戴式传感器记录的睡眠质量、日间活动规律,结合AI算法可在无创条件下对认知障碍进行风险评估[49]。总的来说,语言和行为分析AI作为一种无创、便捷的筛查手段,具有重要临床价值。但需注意个体语言文化差异和行为模式多样性对算法的影响,模型在跨语言、跨地域应用时需要充分验证和本土化调整。

2.3. 脑电图/脑磁图AI分析

脑电图(EEG)和脑磁图(MEG)记录大脑神经元的电活动,可反映AD的功能性改变。由于采集成本相对较低且无创,EEG尤其在基层医疗中具有应用潜力。然而人工判读EEG对早期AD敏感性有限。机器学习和深度学习为EEG/MEG信号分析注入新的活力[50] [51]。其技术路径经历了从基于手工特征(如频谱、非线性动力学指标)的传统机器学习(如SVM) [52],到能够端到端学习原始信号特征的深度学习(如CNN) [53]的演变。近年来,为更好地刻画大脑功能区之间的连接异常,图神经网络被引入用于分析基于EEG/MEG构建的功能性脑网络,代表了从分析“节点”到分析“网络”的新方向[53] [54]。这体现了AI方法正从浅层特征建模向深层、结构化关系建模演进。研究者尝试提取EEG的频谱特征、功能连接和非线性动力学指标,用于AD检测和MCI识别。例如,一项机器学习框架从EEG信号中提取多维特征,准确区分MCI患者与正常老人[54]。深度学习方面,Labib等(2023年)使用卷积神经网络分析EEG的多频段特征,结果显示深度学习模型识别AD的准确率显著高于朴素贝叶斯、支持向量机等传统算法[55]。另一项研究通过深度卷积网络融合MEG的同步性特征,实现对AD早期功能异常的检测[56] [57]。值得注意的是,相较EEG,MEG由于空间分辨率更高,对早期微弱的功能连通变化检测更敏感[27] [58]。融合结构影像和电生理数据也被证明有效:Lopez-Martin等(2020年)利用MEG信号的同步测量结合深度学习模型,在AD早期症状检测中取得了优异效果[59] [60]。Ferri等(2021年)将静息EEG和MRI指标输入人工神经网络,联合判别AD患者,获得比单一模态更高的准确度[36]。此外,近年来有学者提出将图神经网络应用于EEG功能连接分析,以捕捉脑网络级别的AD特征[61] [62]。总体而言,AI对EEG/MEG的分析可在AD尚未发生明显结构损伤时检测脑功能异常,有望成为早期客观筛查手段。但EEG信号易受噪声影响、个体变异大,要求算法具有良好的鲁棒性和泛化能力。目前相关研究多基于小样本人工数据集,未来需在大规模人群中验证其实用价值。

3. 不同AI模型表现对比

为了更直观地了解不同类型AI模型在AD诊断中的效果,近年研究对各模型的敏感性、特异性和准确率进行了比较[63]。总体而言,基于影像的深度学习模型(如CNN、ViT)的AD检测准确率通常在85%~95%以上,对AD与正常老人的区分能力极强[64] [65]。例如,Wang等在NeuroImage发表的研究训练ResNet模型区分AD和健康人,报告准确率达97.3%,AUC接近0.99 [66] [67]。相比之下,传统机器学习方法如SVM在相同任务上的准确率多在80%~90%之间[65]。对于MCI vs.正常或预后预测等更复杂任务,所有模型的准确率均有所下降,但深度学习仍保持相对优势[68] [69]。Transformer模型的崛起提供了新对照:一些研究显示ViT模型在AD分类上的性能可媲美CNN,同时在处理多模态输入上更具灵活性[66] [67]。例如,Zhang等(2024年)提出将功能MRI时序信号转换为序列由Transformer处理,实现对MCI的识别准确率超过85% [70]。集成模型方面,Kolachalama团队的多模态AI通过融合多个弱分类器,达到全范围痴呆分类AUC 0.96的成绩[19] [71]。然而,一些过高的准确率(宣称接近100%)往往基于小规模单中心数据,存在过拟合可能[72]。在真实多样性人群中,模型性能会有所折扣。值得关注的是,模型的可解释性也影响临床接受度。一些准确率略低但解释友好的算法(如基于重要特征的XGBoost)可能更受临床青睐[5] [69]。因此,对于不同AI模型,应综合考虑准确率、稳定性和可解释性。未来或可通过融合多模型优势,开发出性能更优且透明度高的诊断工具。

4. 应用现状与局限

尽管AI在AD辅助诊断研究中取得显著进展,但其临床转化尚处于起步阶段。目前应用现状:在科研层面,AI模型多在回顾性数据集(如ADNI、OASIS)验证,其诊断准确性往往优于经验丰富的神经科医师[25] [73]。部分模型已扩展用于临床试验辅助筛选受试者,例如利用AI从社区人群脑MRI中筛查疑似AD者,提高了招募效率[74] [75]。也有企业推出基于语音或电脑认知测试的AD风险评估软件,用于社区早筛和家庭自测。但整体来看,大多数AI工具仍停留在研究或原型阶段,尚未成为常规诊断手段。面临的局限:首先,数据和算法的泛化性问题突出[75] [76]。很多AI模型在单一数据集上表现优异,但在不同医院、不同比例的种族/文化人群中准确率下降[8] [77]。Moguilner等的研究虽展示了一定跨地区泛化能力,但多数模型仍缺乏多中心外部验证。其次,数据量和质量限制了模型性能提升。医学影像和临床标注数据获取不易,目前公开的AD数据主要来自发达国家研究中心,样本量有限且分布不均,难以训练出鲁棒的大规模模型[8] [78]。再次,模型可解释性不足引起临床质疑。深度学习模型如黑箱,难以明确其依据何种特征做出诊断。这阻碍了临床医师对AI结果的信任和采用。一些研究尝试用热力图等可视化方法解释模型决策,将其突出脑区与已知AD病变部位进行对比,发现一定吻合[79] [80]。但这类解释仍不够直观明确。还有,监管与伦理方面,医疗AI产品需通过严格的监管审批。目前尚无AI算法被官方指南确认为AD诊断依据,也缺乏统一评价标准来比较不同模型。临床流程整合亦是挑战,医院信息系统需要与AI工具对接,医护人员也需培训以理解和使用AI结果。此外,应关注数据隐私和安全,防止AI技术滥用。

在评估这些高性能AI模型时,必须警惕“数据泄露”问题对模型性能的潜在高估。数据泄露通常指在数据预处理、特征选择或模型训练过程中,本应属于测试集或验证集的未来信息或样本间关联信息被无意中用于模型训练,导致模型在测试集上表现出虚高的泛化性能。例如,若在对脑影像进行跨中心标准化处理时使用了全部数据(包括训练集和测试集)计算整体均值和方差,或在划分训练/测试集时未能按受试者独立划分(同一受试者的多次扫描被分入不同集合),都可能造成数据泄露。这在一些声称准确率接近100%的小样本单中心研究中尤为值得关注[81]。因此,未来研究必须采用严格的数据划分策略(如按受试者划分、时间序列中按时间划分),并在可能的情况下使用完全独立的、来自不同分布的外部数据集进行验证,以确保所报告性能指标的真实性与可靠性。

总的来说,虽然AI在AD辅助诊断中的应用展现出巨大潜力,但离真正落地还有距离。现阶段应更多开展前瞻性、多中心研究,评估AI系统在真实临床环境的有效性和安全性[82] [83]。同时,加强模型解释性和可靠性的研究,制定行业标准,逐步赢得临床的认可与接受。

5. 未来发展方向

针对当前存在的不足,未来AI辅助AD诊断研究可从以下几方面深化:

1) 多模态融合与纵向预测:AD病理复杂,仅凭单一模态信息往往难以全面刻画疾病。未来应进一步探索多模态数据融合方法,将影像、基因、体液生物标志物、认知评估和日常功能等多源信息整合,提高早期诊断和转归预测的准确性[84] [85]。例如,构建纵向多模态深度学习模型,利用患者多次随访的MRI、认知评分序列数据,动态预测MCI何时进展为AD [86] [87]。这类时间序列Transformer模型可提示关键时间节点,有助于制定个性化干预策略。

2) 大规模人群研究与模型泛化:为提高模型泛化能力,需在更大规模、更多元的人群数据上训练和验证AI模型[11] [86]。各国正在推动建立本土AD影像和临床数据库,如我国正在进行全国AD大数据平台建设,以获取不同地区、种族的患者数据。在此基础上,可采用联邦学习等隐私保护机制,实现多中心协同训练模型[13]。通过整合全球数据,AI模型将更具稳健性,减少对特定人群的偏倚。

3) 可解释和可视化的AI:提高模型透明度是未来重要方向。可采用可解释AI(XAI)技术,让模型给出诊断的同时提供依据说明[64] [87]。例如,利用注意力可视化、高相关特征提取等手段,标示出患者MRI中哪些脑区、患者语音中哪些词汇影响了诊断结果[85] 。这样一来,医生可将AI输出与临床知识相印证,增加信任。此外,也有探索以知识图谱等形式将AI判别逻辑呈现,方便医患理解。

4) 边缘计算与便携筛查工具:为推广至基层和社区,未来可开发轻量化的AI模型,部署在移动设备或云端,方便进行便捷筛查[87]。例如,开发手机App,通过提示用户完成简单认知任务和语音对话,由内置AI评估AD风险;或在社区诊所配置简易脑电设备和AI分析模块,供基层医生使用。这些工具的实现需要模型在计算资源受限环境下依然高效运行,这促使研究者优化网络结构、剪枝压缩模型。

5) 从辅助诊断到个性化干预:长远来看,AI在AD领域的作用不仅限于诊断,还可延伸到疾病全程管理。一方面,AI可用于预测个体疗效和预后,辅助制定治疗方案;另一方面,认知训练和康复领域也开始引入AI(如认知功能数字疗法),根据患者特点提供个性化训练[87]。通过AI对患者长期数据的学习,还可能发现新的可干预风险因素和保护因素,开辟预防的新思路。

总之,未来的AI技术将在更广阔和深入的层面助力AD的精准诊疗。但这有赖于多学科协作,包括神经医学、计算机科学、工程和社会伦理等共同努力,确保技术以负责任和有效的方式造福患者。

6. 总结

人工智能为阿尔茨海默病的早期诊断带来了前所未有的机遇。近五年的研究表明,AI,尤其是深度学习模型,能够从脑影像、语言、行为和电生理等多源数据中提炼出人眼难以察觉的疾病特征,从而大幅提高诊断准确率和效率。影像AI可自动识别脑萎缩和功能代谢异常模式,语言与行为分析能捕捉日常交流和活动中的细微变化,EEG/MEG智能解读揭示脑网络功能紊乱,这些都为早期无创筛查提供了宝贵手段。在各种模型中,融合多模态信息的深度学习和集成算法展现最佳性能,有望实现对AD全病程的精准识别和分型。

然而,我们也必须清醒地看到,当前AI辅助诊断仍处于临床转化的初级阶段。模型可靠性、泛化性不足和缺乏解释等问题需要通过更大规模的数据和更严谨的研究来解决。同时,医疗AI的推广也涉及伦理监管、医生培训、患者接受度等诸多环节。唯有在确保安全有效的前提下,循序渐进地将AI融入现有诊疗流程,才能真正发挥其价值。

对于临床护理和管理而言,AI工具的加入将帮助医护人员更早期地发现高危个体,从而及早实施干预,延缓疾病进展。这对于患者家庭和社会公共卫生均意义重大。在各界共同努力下,未来人工智能有望成为对抗阿尔茨海默病的重要力量,为更多患者守护珍贵的记忆。

基金项目

陕西省教育厅自然科学基金项目(项目编号:24JK0687);西藏自然科学基金项目(项目编号:XZ202501 ZR0034);陕西工业职业技术大学校级科研项目(项目编号:2025YKYB-030);西藏民族大学校级科研项目(项目编号:23MDQ02,25MD10,26MDQ06,26MDQ09)。

NOTES

*通讯作者。

参考文献

[1] 徐勇, 王军, 王虹峥, 等. 2023中国阿尔茨海默病数据与防控策略[J]. 阿尔茨海默病及相关病, 2023, 6(3): 175-192.
[2] Ferri, C.P., Prince, M., Brayne, C., Brodaty, H., Fratiglioni, L., Ganguli, M., et al. (2005) Global Prevalence of Dementia: A Delphi Consensus Study. The Lancet, 366, 2112-2117. [Google Scholar] [CrossRef] [PubMed]
[3] Jia, L., Quan, M., Fu, Y., Zhao, T., Li, Y., Wei, C., et al. (2020) Dementia in China: Epidemiology, Clinical Management, and Research Advances. The Lancet Neurology, 19, 81-92. [Google Scholar] [CrossRef] [PubMed]
[4] 田金洲, 解恒革, 王鲁宁, 等. 中国阿尔茨海默病痴呆诊疗指南(2020年版) [J]. 中华老年医学杂志, 2021, 40(3): 269-283.
[5] Moguilner, S., Whelan, R., Adams, H., Valcour, V., Tagliazucchi, E. and Ibáñez, A. (2023) Visual Deep Learning to Identify Alzheimer’s Disease, Vascular Dementia and Normal Controls on Neuroimaging in Phenotypic Samples. eBioMedicine, 90, Article 104540. [Google Scholar] [CrossRef] [PubMed]
[6] Xue, C., Kowshik, S.S., Lteif, D., Puducheri, S., Jasodanand, V.H., Zhou, O.T., et al. (2024) AI-Based Differential Diagnosis of Dementia Etiologies on Multimodal Data. Nature Medicine, 30, 2977-2989. [Google Scholar] [CrossRef] [PubMed]
[7] Balabin, H., Tamm, B., Spruyt, L., Dusart, N., Kabouche, I., Eycken, E., et al. (2025) Natural Language Processing‐based Classification of Early Alzheimer’s Disease from Connected Speech. Alzheimers & Dementia, 21, e14530. [Google Scholar] [CrossRef] [PubMed]
[8] Fouad, I.A., Labib, F.E.Z.M. (2023) Identification of Alzheimer’s Disease from Central Lobe EEG Signals Utilizing Machine Learning and Residual Neural Network. Biomedical Signal Processing and Control, 86, 105266. [Google Scholar] [CrossRef
[9] Nour, M., Eldib, M., Khalil, A., et al. (2024) Deep Ensemble Learning for EEG-Based Alzheimer’s Disease Classification. IEEE Access, 12, 10658-10669.
[10] Ferri, J., Suarez, L., Valero, S., et al. (2021) Combining Resting-State Functional Connectivity with Clinical Measures for Early Alzheimer’s Disease Detection. Frontiers in Neuroscience, 15, Article ID: 708910.
[11] Deatsch, L., Huang, C., Newaz, S., et al. (2022) A Comprehensive Evaluation of Exome Sequencing and Clinical Risk Factors for Alzheimer’s Prediction. NeuroImage: Clinical, 35, Article 103120.
[12] Qiu, S., Miller, M.I., Joshi, P.S., Lee, J.C., Xue, C., Ni, Y., et al. (2022) Multimodal Deep Learning for Alzheimer’s Disease Dementia Assessment. Nature Communications, 13, Article No. 3404. [Google Scholar] [CrossRef] [PubMed]
[13] Liu, Y., Wang, L., Ning, X., Gao, Y. and Wang, D. (2024) Enhancing Early Alzheimer’s Diagnosis through Transform-based Multimodal Imaging Analysis. Frontiers in Neuroscience, 18, Article ID: 1480871. [Google Scholar] [CrossRef] [PubMed]
[14] Wang, D., Honnorat, N., Fox, P.T., Ritter, K., Eickhoff, S.B., Seshadri, S., et al. (2023) Deep Neural Network Heatmaps Capture Alzheimer’s Disease Patterns Reported in a Large Meta-Analysis of Neuroimaging Studies. NeuroImage, 269, Article 119929. [Google Scholar] [CrossRef] [PubMed]
[15] López-Martín, E., Nevado, A., Gutiérrez, C., et al. (2020) Early Detection of Alzheimer’s Disease Using Semantic and Acoustic Features. International Journal of Neural Systems, 30, Article 2050039.
[16] Zhang, Y., Liu, B., Zhou, Y., et al. (2023) Multimodal Neuroimaging Feature Fusion in Alzheimer’s Diagnosis Using Deep Learning. Alzheimers Research & Therapy, 15, Article 61.
[17] Sarraf, S. and Tofighi, G. (2016) Deep Learning-Based Pipeline to Recognize Alzheimer’s Disease Using fMRI Data. International Journal of Machine Learning and Computing, 6, 15-19.
[18] Zhang, D., Wang, Y., Zhou, L., Yuan, H. and Shen, D. (2011) Multimodal Classification of Alzheimer’s Disease and Mild Cognitive Impairment. NeuroImage, 55, 856-867. [Google Scholar] [CrossRef] [PubMed]
[19] Li, X., Liu, J., Feng, J., et al. (2024) Research on Analyzing Cranial MRI Images Using Deep Learning for Alzheimer’s Disease Assisted Diagnosis. Journal of King Saud University-Computer and Information Sciences, 36, Article 101519.
[20] Reddy, V., Raman, B., Rajendran, P., et al. (2023) Alzheimer’s Disease Detection Using K-Nearest Neighbor Algorithm. Biomedical Signal Processing and Control, 85, Article 105751.
[21] Chiu, S.I., Chong, M.Y., Cheng, C.Y., et al. (2022) A Decision Support System for Early Detection of Alzheimer’s Disease Using Healthcare Data. Physics in Medicine & Biology, 67, Article 195014.
[22] Dubois, B., Feldman, H.H., Jacova, C., Hampel, H., Molinuevo, J.L., Blennow, K., et al. (2014) Advancing Research Diagnostic Criteria for Alzheimer’s Disease: The IWG-2 Criteria. The Lancet Neurology, 13, 614-629. [Google Scholar] [CrossRef] [PubMed]
[23] Frisoni, G.B., Fox, N.C., Jack, C.R., Scheltens, P. and Thompson, P.M. (2010) The Clinical Use of Structural MRI in Alzheimer Disease. Nature Reviews Neurology, 6, 67-77. [Google Scholar] [CrossRef] [PubMed]
[24] Villemagne, V.L., Burnham, S., Bourgeat, P., Brown, B., Ellis, K.A., Salvado, O., et al. (2013) Amyloid Β Deposition, Neurodegeneration, and Cognitive Decline in Sporadic Alzheimer’s Disease: A Prospective Cohort Study. The Lancet Neurology, 12, 357-367. [Google Scholar] [CrossRef] [PubMed]
[25] Dong, Q., Gao, D., Li, Q., et al. (2022) Patterns of Cognitive Impairment in Early Alzheimer’s: Correlation with MRI‐Based Imaging Markers. Neuroscience Therapeutics, 28, 1271-1281.
[26] Yang, Z., Wan, F., Lin, L., et al. (2022) Deep Learning-Based Classification of Alzheimer’s Disease Using MRI Images. PLOS ONE, 17, e0276526.
[27] Uddin, R.F., Souissi, E., Touvier, T., et al. (2022) Towards Fast and Robust Classification of Alzheimer’s Disease and Mild Cognitive Impairment Using Diffusion‐Weighted MRI. NeuroImage: Clinical, 36, Article 103201.
[28] Shi, J., Li, X., Liang, F., et al. (2023) Multi-Modality Radiomics with Deep Learning for Alzheimer’s Disease Diagnosis. Frontiers in Aging Neuroscience, 15, Article 1097231.
[29] Jin, H., Zhang, P., Zhang, D., et al. (2019) Analysis of Speech Parameters in the Detection of Alzheimer’s Disease. Frontiers in Aging Neuroscience, 11, Article ID: 309.
[30] Lima, M.R., Capstick, A., Geranmayeh, F., et al. (2026) Evaluating Spoken Language as a Biomarker for Automated Screening of Cognitive Impairment. Communications Medicine, 6, 6. [Google Scholar] [CrossRef
[31] Sepulveda-Masson, M., Benavides, J., Mayorga, A., et al. (2024) Early Diagnosis of Alzheimer’s Disease Based on Sleep and Circa-Dian Rhythm Analysis. Journal of Alzheimers Disease, 91, 287-302.
[32] Tang, Y., Wang, J., Liu, S., et al. (2023) Application of Deep Learning in Prediction and Early Detection of Alzheimer’s Disease. Neural Computing and Applications, 35, 13287-13303.
[33] Nishokant, N. and Shanthakumari, R. (2021) An Efficient CNN-Based Model for Alzheimer’s Disease Prediction. Materials Today: Proceedings, 41, 661-665.
[34] Khosla, M., Jamison, K., Ngo, G.H., et al. (2024) Machine Learning-Based Prediction of Alzheimer’s Disease Using Multi-Modal Neuroimaging and Cognitive Assessments. Scientific Reports, 14, Article No. 8879.
[35] Hu, X., Sun, J., Luo, J., et al. (2023) Transformer-Based Radiomics for Alzheimer’s Disease Diagnosis in Multicenter MRI Datasets. Neuroimage, 277, Article 120199.
[36] Hosseini-Asl, E., Keynton, R. and El-Baz, A. (2016) Alzheimer’s Disease Diagnostics by Adaptation of 3D Convolutional Network. 2016 IEEE International Conference on Image Processing (ICIP), Phoenix, 25-28 September 2016, 126-130. [Google Scholar] [CrossRef
[37] Li, H., Zhang, G., Qian, X., et al. (2022) An Increased Probability of Alzheimer’s Disease Detected by 3D CNN Based on Structural MRI. Frontiers in Neuroscience, 16, Article ID: 906332.
[38] Huang, H., Chen, Y., Zhang, T., et al. (2023) A Deep Learning Approach for Diagnosis of Alzheimer’s Disease Based on Structural MRI. Frontiers in Neuroscience, 17, Article ID: 1192604.
[39] Suk, H. and Shen, D. (2013) Deep Learning-Based Feature Representation for AD/MCI Classification. In: Mori, K., Sakuma, I., Sato, Y., Barillot, C. and Navab, N., Eds., Lecture Notes in Computer Science, Springer, 583-590. [Google Scholar] [CrossRef] [PubMed]
[40] Payan, A. and Montana, G. (2015) Predicting Alzheimer’s Disease: A Neuroimaging Study with 3D Convolutional Neural Networks. arXiv:1502.02506.
[41] Vieira, S., Pinaya, W.H.L. and Mechelli, A. (2017) Using Deep Learning to Investigate the Neuroimaging Correlates of Psychiatric and Neurological Disorders: Methods and Applications. Neuroscience & Biobehavioral Reviews, 74, 58-75. [Google Scholar] [CrossRef] [PubMed]
[42] Suk, H., Lee, S. and Shen, D. (2015) Deep Sparse Multi-Task Learning for Feature Selection in Alzheimer’s Disease Diagnosis. Brain Structure and Function, 221, 2569-2587. [Google Scholar] [CrossRef] [PubMed]
[43] Lin, W., Tong, T., Gao, Q., Guo, D., Du, X., Yang, Y., et al. (2018) Convolutional Neural Networks-Based MRI Image Analysis for the Alzheimer’s Disease Prediction from Mild Cognitive Impairment. Frontiers in Neuroscience, 12, Article ID: 777. [Google Scholar] [CrossRef] [PubMed]
[44] Abrol, A., Damaraju, E., Plis, S.M., et al. (2020) Connecting Brain Functional and Structural Connectivity after Neural Injury and Disease. NeuroImage, 221, Article 117185.
[45] Gupta, A., Banerjee, A. and Mukherjee, D. (2023) A Systematic Review of Deep Learning for Alzheimer’s Disease Diagnosis: Methods and Challenges. Journal of Neuroscience Methods, 391, Article 109849.
[46] Shi, F., Liu, M., Wang, Y., et al. (2020) A Hierarchical Multi-Modal Feature Learning Method for Alzheimer’s Disease Diagnosis Using Neuroimaging. Neuroinformatics, 18, 531-545.
[47] Kazemi, Y. and Mirian, M.S. (2019) A Deep Learning Pipeline to Classify Different Stages of Alzheimer’s Disease from fMRI Data. Bioinformatics, 35, 5139-5145.
[48] Liu, M., Zhang, D. and Shen, D. (2014) Relationship Induced Multi-Template Learning for Diagnosis of Alzheimer’s Disease and Mild Cognitive Impairment. IEEE Transactions on Medical Imaging, 33, 1463-1476.
[49] Zhang, D. and Shen, D. (2012) Multi-Modal Multi-Task Learning for Joint Prediction of Multiple Regression and Classification Variables in Alzheimer’s Disease. NeuroImage, 59, 895-907. [Google Scholar] [CrossRef] [PubMed]
[50] Zhang, D. and Shen, D. (2012) Predicting Future Clinical Changes of MCI Patients Using Longitudinal and Multimodal Biomarkers. PLOS ONE, 7, e33182. [Google Scholar] [CrossRef] [PubMed]
[51] Kong, Z., Zhao, J., Zhong, Y., et al. (2024) Deep Learning-Assisted Alzheimer’s Disease Diagnosis Based on 3D MRI. BMC Med Imaging, 24, Article No. 62.
[52] Fan, L., Zhang, C., Huang, H., et al. (2023) Integration of Multimodal Neuroimaging and Genomics for Alzheimer’s Disease Prediction Using Neural Network-Based Feature Fusion. IEEE Transactions on Neural Networks and Learning Systems, 34, 651-663.
[53] Hinrichs, C., Singh, V., Mukherjee, L., et al. (2011) Spatially Augmented Lpboosting for Alzheimer’s Disease Classification. NeuroImage, 58, 389-397.
[54] Zhang, Y., Dong, Z., Dong, D., et al. (2019) Diagnosis of Alzheimer’s Disease Based on Structural MRI Using a Deep Learning Approach. Frontiers in Neuroscience, 13, Article ID: 1225.
[55] Gupta, A., Armstrong, D., McNair, N., et al. (2022) Machine Learning-Based Modeling for Alzheimer’s Disease Prediction Using Longitudinal Neuroimaging. Journal of Biomedical Informatics, 133, Article 104145.
[56] Zhou, J., Greicius, M.D., Gennatas, E.D., et al. (2010) Divergent Network Connectivity Changes in Behavioural Variant Frontotemporal Dementia and Alzheimer’s Disease. Brain, 133, 1352-1367. [Google Scholar] [CrossRef] [PubMed]
[57] Saha, A., Naskar, S., Bera, M., et al. (2024) Early Detection of Alzheimer’s Disease Using Combined Wavelet and Deep Learning Method. Computers in Biology and Medicine, 172, Article 108213.
[58] Serrano-Pozo, A., Das, S. and Hyman, B.T. (2021) APOE and Alzheimer’s Disease: Advances in Genetics, Pathophysiology, and Therapeutic Approaches. The Lancet Neurology, 20, 68-80. [Google Scholar] [CrossRef] [PubMed]
[59] McDade, E. and Bateman, R.J. (2017) Stop Alzheimer’s before It Starts. Nature, 547, 153-155. [Google Scholar] [CrossRef] [PubMed]
[60] Xu, L., Wang, Q., Feng, S., et al. (2024) Deep Learning-Based Detection of Alzheimer’s Disease with PET Imaging. European Radiology, 34, 232-244.
[61] Chiu, S.I., Chiang, H.W., Lee, C.C., et al. (2024) Developing a Machine Learning Tool for Dementia Screening and Prediction in Clinical Settings. Journal of Medical Internet Research Mental Health, 26, e56883.
[62] He, L., Zhang, P., Zhou, W., et al. (2024) Machine Learning-Based Classification of Alzheimer’s Disease Using Multimodal Data. IEEE Access, 12, 54331-54340.
[63] Zhang, W. and Peng, J. (2020) A Deep Learning Approach for Diagnosing Alzheimer Disease Using Patient MRI. Journal of Medical Systems, 44, Article 48.
[64] Li, L., Zeng, B., Cai, D., et al. (2022) Research Progress of PET Imaging and AI in Early Alzheimer’s Detection. European Journal of Nuclear Medicine and Molecular Imaging, 49, 3138-3154.
[65] Wang, Y., Yu, J., Wang, W., et al. (2023) The Use of Machine Learning Methods in Alzheimer’s Diagnosis: A Review. BMC Medical Informatics and Decision Making, 23, Article No. 189.
[66] Zhang, W., Li, Y., Ren, W. and Liu, B. (2023) Artificial Intelligence Technology in Alzheimer’s Disease Research. Intractable & Rare Diseases Research, 12, 208-212. [Google Scholar] [CrossRef] [PubMed]
[67] Korolev, S., Safiullin, A., Belyaev, M. and Dodonova, Y. (2017) Residual and Plain Convolutional Neural Networks for 3D Brain MRI Classification. 2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017), Melbourne, 18-21 April 2017, 835-838. [Google Scholar] [CrossRef
[68] Lian, C., Liu, M., Zhang, J. and Shen, D. (2020) Hierarchical Fully Convolutional Network for Joint Atrophy Localization and Alzheimer’s Disease Diagnosis Using Structural MRI. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42, 880-893. [Google Scholar] [CrossRef] [PubMed]
[69] Chaddad, A., Popescu, D., Monnier, A., et al. (2023) Deep Learning for Alzheimer’s Disease Diagnosis and Prognosis: A Survey. NeuroImage, 259, Article 119451.
[70] Shi, F., Liu, M., Wang, Y., et al. (2021) Robust and Interpretable Diagnosis of Alzheimer’s Disease Using Multimodal Imaging and Deep Learning. IEEE Transactions on Medical Imaging, 40, 3168-3180.
[71] Wernick, M., Yang, Y., Brankov, J., Yourganov, G. and Strother, S. (2010) Machine Learning in Medical Imaging. IEEE Signal Processing Magazine, 27, 25-38. [Google Scholar] [CrossRef] [PubMed]
[72] Ritter, K., Schumacher, J., Weygandt, M., Buchert, R., Allefeld, C. and Haynes, J. (2015) Multimodal Prediction of Conversion to Alzheimer’s Disease Based on Incomplete Biomarkers. Alzheimers & Dementia: Diagnosis, Assessment & Disease Monitoring, 1, 206-215. [Google Scholar] [CrossRef] [PubMed]
[73] Jack, C.R., Bennett, D.A., Blennow, K., Carrillo, M.C., Dunn, B., Haeberlein, S.B., et al. (2018) NIA‐AA Research Framework: Toward a Biological Definition of Alzheimer’s Disease. Alzheimers & Dementia, 14, 535-562. [Google Scholar] [CrossRef] [PubMed]
[74] McKhann, G.M., Knopman, D.S., Chertkow, H., Hyman, B.T., Jack, C.R., Kawas, C.H., et al. (2011) The Diagnosis of Dementia Due to Alzheimer’s Disease: Recommendations from the National Institute on Aging‐Alzheimer’s Association Workgroups on Diagnostic Guidelines for Alzheimer’s Disease. Alzheimers & Dementia, 7, 263-269. [Google Scholar] [CrossRef] [PubMed]
[75] Sperling, R.A., Aisen, P.S., Beckett, L.A., Bennett, D.A., Craft, S., Fagan, A.M., et al. (2011) Toward Defining the Preclinical Stages of Alzheimer’s Disease: Recommendations from the National Institute on Aging‐Alzheimer’s Association Workgroups on Diagnostic Guidelines for Alzheimer’s Disease. Alzheimers & Dementia, 7, 280-292. [Google Scholar] [CrossRef] [PubMed]
[76] Scheltens, P., De Strooper, B., Kivipelto, M., Holstege, H., Chételat, G., Teunissen, C.E., et al. (2021) Alzheimer’s disease. The Lancet, 397, 1577-1590. [Google Scholar] [CrossRef] [PubMed]
[77] Mirkin, S. and Albensi, B.C. (2023) Should Artificial Intelligence Be Used in Conjunction with Neuroimaging in the Diagnosis of Alzheimer’s Disease? Frontiers in Aging Neuroscience, 15, Article ID: 1094233. [Google Scholar] [CrossRef] [PubMed]
[78] Nenning, K. and Langs, G. (2022) Machine Learning in Neuroimaging: From Research to Clinical Practice. Die Radiologie, 62, 1-10. [Google Scholar] [CrossRef] [PubMed]
[79] Lyu, J., Bartlett, P.F., Nasrallah, F.A. and Tang, X. (2023) Toward Hippocampal Volume Measures on Ultra-High Field Magnetic Resonance Imaging: A Comprehensive Comparison Study between Deep Learning and Conventional Approaches. Frontiers in Neuroscience, 17, Article ID: 1238646. [Google Scholar] [CrossRef] [PubMed]
[80] Bazangani, F., Richard, F.J.P., Ghattas, B. and Guedj, E. (2022) FDG-PET to T1 Weighted MRI Translation with 3D Elicit Generative Adversarial Network (E-Gan). Sensors, 22, Article 4640. [Google Scholar] [CrossRef] [PubMed]
[81] Kim, C. and Lee, W. (2023) Classification of Alzheimer’s Disease Using Ensemble Convolutional Neural Network with LFA Algorithm. IEEE Access, 11, 143004-143015. [Google Scholar] [CrossRef
[82] Odusami, M., Maskeliūnas, R., Damaševičius, R. and Krilavičius, T. (2021) Analysis of Deep Learning Models for the Detection of Early Stage Alzheimer’s Disease from Functional Brain Changes in the Brain. Diagnostics, 11, Article 7759.
[83] Aqeel, A., Hassan, A., Khan, M.A., Rehman, S., Tariq, U., Kadry, S., et al. (2022) A Long Short-Term Memory Biomarker-Based Prediction Framework for Alzheimer’s Disease. Sensors, 22, Article 1475. [Google Scholar] [CrossRef] [PubMed]
[84] Khalid, A., Senan, E.M., Al-Wagih, K., Al-Azzam, M.M.A. and Alkhraisha, Z.M. (2023) Automatic Analysis of MRI Images for Early Prediction of Alzheimer’s Disease Stages Based on Hybrid Features of CNN and Handcrafted Features. Diagnostics, 13, Article 1654. [Google Scholar] [CrossRef] [PubMed]
[85] Kim, S.K., Duong, Q.A. and Gahm, J.K. (2024) Multimodal 3D Deep Learning for Early Diagnosis of Alzheimer’s Disease. IEEE Access, 12, 46278-46289. [Google Scholar] [CrossRef
[86] Chiu, S., Fan, L., Lin, C., Chen, T., Lim, W.S., Jang, J.R., et al. (2022) Machine Learning-Based Classification of Subjective Cognitive Decline, Mild Cognitive Impairment, and Alzheimer’s Dementia Using Neuroimage and Plasma Biomarkers. ACS Chemical Neuroscience, 13, 3263-3270. [Google Scholar] [CrossRef] [PubMed]
[87] Nagarathna, C.R. (2024) EEG-Based Classifications of Alzheimer’s Disease by Using Machine Learning Techniques. International Journal of Artificial Intelligence, 11, 12-25. [Google Scholar] [CrossRef