https://doi.org/10.1140/epjp/s13360-025-06673-7
Regular Article
Spark-based Alzheimer’s disease classification using hybrid spike Google-deep CNN-SEViT model
1
Computer Science and Engineering, Kalasalingam Academy of Research and Education, Anand Nagar, 626126, Krishnankoil, Tamil Nadu, India
2
Deparment of Computer Science and Engineering, Kalasalingam Academy of Research and Education, Anand Nagar, 626126, Krishnankoil, Tamil Nadu, India
Received:
21
June
2025
Accepted:
18
July
2025
Published online:
13
August
2025
Alzheimer's disease (AD) constitutes a neurological disorder predominantly affecting the geriatric population. It ranks as the fourth principal cause of death globally, following cardiovascular diseases, various malignancies, and cerebral hemorrhages. The pathology of Alzheimer’s leads to the degeneration of neural tissues and the demise of neurons across the cerebral cortex, thereby precipitating significant memory impairment in individuals and adversely affecting their capacity to accomplish regular activities, including writing, verbal communication, and reading. In recent decades, neuroimaging modalities have been increasingly harnessed to delineate the characteristics of AD through utilization of Machine Learning (ML) approaches, thus providing promising avenues for personalized diagnostic and prognostic assessments. More recently, Deep Learning (DL), as a nascent paradigm within the realm of ML, has witnessed substantial advancements in field of medical imaging. A more comprehensive approach is required that can accommodate big data in AD research due to the growing size of multi-modalities. This research generates a method for classification of AD using DL in spark framework. The preprocessing stage includes realignment, normalization and smoothing executed in a slave set. After preprocessing, feature extraction phase is achieved on preprocessed image in another slave set using Gold Rush Optimizer (GRO)-based VGG-16. These extracted features are then sent to master, where the Proposed SpikeGoogle- Deep Convolutional Neural Network- Squeeze-and-Excitation Vision Transformer (SpikeGoogle-DeepCNN-SEViT) model is created by combining SpikeGoogle, Deep CNN and SEViT for AD classification. The proposed SpikeGoogle-DeepCNN-SEViT model classifies the data into four classes: Mild Demented, Moderate Demented, Non-Demented, and Very Mild Demented. Experimental evidence highlights that Proposed SpikeGoogle-DeepCNN-SEViT showed higher performance compared to conventional strategies with 95.7% accuracy, 97.5% of sensitivity and 95.3% of specificity.
Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Società Italiana di Fisica and Springer-Verlag GmbH Germany, part of Springer Nature 2025
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
