Alzheimer's disease (AD) is an irreversible neurodegenerative brain disorder and the most common cause of dementia with significant social and economic impact. Despite the severity and prevalence of AD, early diagnosis of AD remains quite challenging. One of the important markers of neuropathology in AD patients is the aggregation of Aβ1–42 monomers into oligomers, which then become fibrils and eventually form beta-amyloid plaques. Herein, a highly sensitive surface-enhanced Raman spectroscopy (SERS) method based on graphene oxide/gold nanoparticles (GO/Au NPs) for the detection of Aβ 1–42 is proposed. Specifically, Au NPs were deposited on the GO surface by in situ reduction to form high-density hot spots for SERS. The limits of detection are 0.0232 ng mL^-1 and 0.0192 ng mL^-1 for Aβ1–42 monomer and fibrils. In addition, support vector machine (SVM) and one-dimensional convolutional neural network (1DCNN) algorithm were used to identify samples with different fibrils degrees. SERS is expected to be used for label-free diagnosis and early detection of AD, which has exciting potential for biomedical detection.

阿尔茨海默病（AD）是一种不可逆的神经退行性脑部疾病，也是痴呆症的最常见原因，具有重大的社会和经济影响。尽管 AD 很严重且普遍存在，但 AD 的早期诊断仍然相当具有挑战性。AD 患者神经病理学的重要标志之一是 Aβ1-42 单体聚集成寡聚体，然后形成原纤维并最终形成 β-淀粉样斑块。在此，提出了一种基于氧化石墨烯/金纳米粒子（GO/Au NPs）的高灵敏度表面增强拉曼光谱（SERS）方法，用于检测Aβ 1-42。具体来说，通过原位还原将Au NPs沉积在GO表面上，形成SERS的高密度热点。Aβ1–42 单体和原纤维的检测限为 0.0232 ng mL ^-1和 0.0192 ng mL ^{-1 。}此外，还利用支持向量机（SVM）和一维卷积神经网络（1DCNN）算法来识别不同原纤维程度的样本。SERS有望用于AD的无标记诊断和早期检测，这在生物医学检测方面具有令人兴奋的潜力。