A HYBRID APPROACH FOR DETECTION OF ALZHEIMER'S DISEASE USING CONTINUOUS SQUASHING TECHNIQUE

Abstract

A hybrid technique for detecting Alzheimer's disease (AD) is proposed the usage of a continuous squashing approach, which mixes superior machine mastering algorithms with revolutionary function extraction strategies. The continuous squashing method enhances category performance by using way of effectively dealing with big datasets in AD analysis. Integrated with convolutional neural networks (CNNs), this method extracts discriminative capabilities from neuroimaging information in conjunction with MRI and PET scans. The hybrid version demonstrates advanced accuracy, sensitivity, and specificity compared to standard deep mastering fashions, on the identical time as addressing troubles of records imbalance and overfitting. Its ability to hit upon early tiers of AD is confirmed via move-validation, showcasing its capability for actual-worldwide scientific packages. This approach advances AD detection and helps the development of reliable, non-invasive diagnostic gear for neurodegenerative diseases.