Page 149 - 《南京医科大学学报(自然科学版)》2026年第3期
P. 149

第46卷第3期                           南京医科大学学报(自然科学版)
                  2026年3月                   Journal of Nanjing Medical University(Natural Sciences)     ·457 ·


               ·综 述·

                无创生物标志物与人工智能多模态整合技术在阿尔茨海默病

                早期诊断与筛查中的研究进展



                王安倩 ,李心竹 ,周辰宇          3,4*
                      1
                              2
                南京医科大学第一临床医学院,医学影像学院,医政学院,数智技术与健康治理实验室,江苏                                 南京 211166
                1                        2           3        4


               [摘   要] 阿尔茨海默病(Alzheimer’s disease,AD)已成为21世纪全球公共卫生领域的重大挑战。目前临床诊断主要依赖脑
                脊液生物标志物检测和淀粉样蛋白PET成像等技术,存在明显的局限性,如脑脊液检测具有侵入性,PET成像则具有成本高昂
                和放射性暴露等问题。虽然神经心理学量表在临床实践中应用广泛,但主观性强、特异性不足等缺点导致其对早期AD的诊
                断敏感性显著不足。针对这一现状,文章系统综述了AD无创诊断技术的最新研究进展,重点探讨了外周体液(如血液、唾液
                等)生物标志物检测技术的突破,以及人工智能驱动的多模态数据融合策略在AD早期识别和精准诊断中的应用前景与创新
                价值。
               [关键词] 阿尔茨海默病;生物标志物;人工智能;多模态整合
               [中图分类号] TP18;R749.16              [文献标志码] A                      [文章编号] 1007⁃4368(2026)03⁃457⁃10
                doi:10.7655/NYDXBNSN250832


                Research progress of non ⁃ invasive biomarkers and artificial intelligence multimodal
                integration technology in the early diagnosis and screening of Alzheimer’s disease

                            1
                                      2
                WANG Anqian ,LI Xinzhu ,ZHOU Chenyu 3,4*
                                                                           3
                1 The First Clinical Medical College,School of Medical Imaging,School of Health Policy & Management,
                                                 2
                4 Laboratory for Digital Intelligence & Health Governance,Nanjing Medical University,Nanjing 211166,China
               [Abstract] Alzheimer’s disease(AD)has emerged as a major global public health challenge in the 21st century. Current clinical
                diagnosis primarily relies on techniques such as cerebrospinal fluid biomarker testing and amyloid PET imaging,yet these methods
                have exhibited significant limitations:cerebrospinal fluid testing is invasive,and PET imaging involves high costs and radiation
                exposure. Although neuropsychological scales are widely used in clinical practice,their strong subjectivity and lack of specificity
                considerably reduce diagnostic sensitivity,particularly in early⁃stage AD. In response to this situation,this article provides a systematic
                review of the latest advances in non⁃invasive diagnostic technologies for AD,with a focus on breakthroughs in peripheral biofluids(e.g.,
                blood,saliva)biomarker detection techniques,as well as the application prospects and innovative value of artificial intelligence⁃driven
                multimodal data integration strategies in the early identification and precise diagnosis of AD.
               [Key words] Alzheimer’s disease;biomarkers;artificial intelligence;multimodal integration
                                                                          [J Nanjing Med Univ,2026,46(03):457⁃465,474]





                    阿尔茨海默病(Alzheimer’s disease,AD)作为最             常见的神经退行性疾病,其疾病负担正随着全球老龄
                                                                  化加剧而持续加重。流行病学预测显示,到2050年
                                                                  全球AD患者将达1.528亿例 。中国作为老龄化速
                                                                                           [1]
               [基金项目] 江苏省高校哲学社会科学研究一般项目
                                                                  度最快的国家之一,AD 防控形势尤为严峻,但现有
               (2024SJYB0239)
                                                                                                             [2]
                                                                  诊断技术的局限性严重制约了早期干预效果 。
                ∗
                通信作者(Corresponding author),E⁃mail:zhouchenyu@njmu.
                edu.cn(ORCID:0009⁃0009⁃3930⁃3993)                 AD病理进程如β⁃淀粉样蛋白(amyloid β,Aβ)沉积,
   144   145   146   147   148   149   150   151   152   153   154