A novel deep learning technique for multi classify Alzheimer disease: hyperparameter optimization technique.

Area of Science:

• Neurology
• Artificial Intelligence
• Medical Imaging

Background:

• Alzheimer's disease (AD) is a progressive brain disorder impacting memory and cognition.
• Early detection and classification of AD phases are critical for effective therapeutic interventions.
• Existing methods face challenges with limited data and computational resources.

Purpose of the Study:

• To develop a novel deep learning method for accurate Alzheimer's disease phase classification.
• To address limitations of data scarcity and computational demands in AD detection.
• To improve the efficiency and effectiveness of classifying different stages of Alzheimer's disease.

Main Methods:

• A novel deep learning approach utilizing a ResNet152V2 model.
• A newly proposed hyperparameter optimization technique to identify optimal model parameters.
• Comparison against state-of-the-art transfer learning and classical models.

Main Results:

• The proposed deep learning method demonstrated superior performance in classifying AD phases.
• The model achieved high scores in recall, precision, F1 score, and accuracy.
• The approach proved more effective and efficient than existing methods.

Conclusions:

• The novel deep learning method offers a promising solution for accurate Alzheimer's disease phase classification.
• This advancement supports earlier diagnosis, enabling timely disease management strategies.
• The optimized ResNet152V2 model effectively overcomes data and resource limitations.