Prediction of Alzheimer's Disease Using Modified DNN with Optimal Feature Selection Based on Seagull Optimization | JoVE Visualize

Area of Science:

Neurology
Medical Imaging
Artificial Intelligence

Background:

Alzheimer's disease (AD) is a neurodegenerative disorder causing irreversible brain cell and tissue loss, significantly impacting memory.
Diagnosing AD is challenging due to subtle differences in brain processes and pixel intensities in medical images, hindering traditional machine learning classification.
Existing methods struggle with the complex feature representation required for accurate Alzheimer's disease detection.

Purpose of the Study:

To develop an accurate and robust method for predicting Alzheimer's disease using brain Magnetic Resonance Imaging (MRI).
To overcome the limitations of traditional machine learning in classifying Alzheimer's disease by employing advanced feature extraction and optimization techniques.
To enhance the diagnostic capabilities for Alzheimer's disease through a novel deep neural network (DNN) model.

Main Methods:

Hybrid feature extraction combining Gray Level Run Length Matrix (GLRLM), Gabor wavelet transform, and Local Energy-based Shape Histogram (LESH) for brain MRI analysis.
Image preprocessing including resizing and enhancement using the BW-net technique.
Optimal feature selection using the SEAGULL optimization algorithm, followed by training a modified Deep Neural Network (DNN) for disease prediction.

Main Results:

The proposed model achieved high performance metrics: 91% precision, 2% error, 98% accuracy, and 97% recall.
The hybrid feature extraction effectively captured shape, texture, and edge information from brain MRI scans.
The SEAGULL optimization technique successfully selected optimal features for improved DNN model training.

Conclusions:

The developed Alzheimer's disease prediction model, utilizing modified DNN with SEAGULL-optimized features, demonstrates superior performance compared to existing methods.
The integration of GLRLM, Gabor wavelet transform, and LESH offers a powerful approach for extracting discriminative features from brain MRI.
This study presents a promising AI-driven solution for accurate and efficient Alzheimer's disease diagnosis.