Alzheimer’s disease (AD) is a dementia type, which is highly faced by elderly persons. In AD, cells of the brain that are accountable for forming memories, as well as cognitive decisions, are affected, which forms shrinkage of the entire gray matter in the human brain. As the patients of with AD are increasing across the globe, researchers are attempting to devise an accurate method to diagnose the disease utilizing images of the brain. Here, Manta Crow Light Spectrum Optimization-based SpinalNet (MCLSO-SpinalNet) is designed for the classification of AD. The adaptive median filter is utilized for pre-processing, wherein Region of Interest (ROI) extraction is also performed. Thereafter, the segmentation process is performed utilizing K-Net and training of K-Net is done by Light Spectrum Optimizer (LSO). After that, features are extracted from the feature extraction stage. From extracted features, the detection of AD is accomplished by employing a Deep Neuro Fuzzy Network (DNFN) that is tuned by MCLSO. However, MCLSO is an incorporation of Manta Crow Search Optimization (MCSO) and LSO. Additionally, MCSO is designed by integrating Manta-Ray Foraging Optimization (MRFO) with the Crow Search Algorithm (CSA). At last, AD classification is executed with SpinalNet, which is also trained by MCLSO. Furthermore, MCLSO-SpinalNet obtained high accuracy, sensitivity, F-measure, and specificity of about 92.6%, 93.6%, 91.7%, and 93.3%.

阿尔茨海默病（AD）是一种老年痴呆症，是老年人最容易罹患的疾病。在 AD 中，负责形成记忆和认知决策的大脑细胞受到影响，从而导致人脑整个灰质萎缩。随着全球 AD 患者数量的增加，研究人员正在尝试设计一种利用大脑图像诊断该疾病的准确方法。这里，基于蝠鲼光谱优化的 SpinalNet (MCLSO-SpinalNet) 是为 AD 分类而设计的。利用自适应中值滤波器进行预处理，其中还执行感兴趣区域(ROI)提取。此后，利用 K-Net 执行分割过程，并通过光谱优化器 (LSO) 完成 K-Net 的训练。之后，从特征提取阶段提取特征。根据提取的特征，AD 的检测是通过采用由 MCLSO 调整的深度神经模糊网络 (DNFN) 来完成的。然而，MCLSO 是 Manta Crow 搜索优化 (MCSO) 和 LSO 的结合。此外，MCSO 的设计是将蝠鲼觅食优化 (MRFO) 与乌鸦搜索算法 (CSA) 相集成。最后，使用 SpinalNet 执行 AD 分类，该网络也是由 MCLSO 训练的。此外，MCLSO-SpinalNet 获得了约 92.6%、93.6%、91.7% 和 93.3% 的高精度、灵敏度、F 测量和特异性。