Hybrid Method of Assessing the Development of Alzheimer’s Syndrome through CNN/LSTM and PSO

Alzheimer’s Disease (AD) is a prevalent neurodegenerative disease that has a substantial influence on society. Precise estimation of Alzheimer’s disease development is essential for prompt diagnosis and customized treatment. This paper proposes a hybrid model that combines Particle Swarm Optimization (PSO), Long Short-Term Memory networks (LSTM), and Convolutional Neural Networks (CNN) to improve prediction performance. The model examines the evolution of AD using a data-focused method, making use of an extensive MRI dataset with important neuroimaging variables. AD causes mental impairment and loss of memory, which complicates medical care provided to them. Data already in the public domain are used to predict future events or outcomes for forecasting that is essential for treating the disease. Conventional approaches, like CNN and LSTM, are useful for predicting the course of an illness because they can handle temporal patterns and spatial interdependence in datasets, respectively. Using the advantages of both methodologies, a new hybrid approach is proposed in this article. In order to optimize the hybrid model variables, this work employs Particle Swarm Optimization (PSO), a highly efficient optimization approach to increase the accuracy of model forecasts. The hybrid CNN-LSTM model, with and without PSO, demonstrated accurate predictions of AD progression. Performance metrics such as F1-score, precision, accuracy, recall and ROC AUC were used to evaluate model efficacy. This research contributes to predictive analytics in healthcare, offering innovative solutions for enhancing patient outcomes in AD management.