IoT-Enabled Alzheimer Disease Detection: A Convolutional Encoder-Decoder Approach Enhanced by Alpine Skiing Optimization

Alzheimer's disease (AD) represents a significant challenge in the treatment of dementia among older adults, with early diagnosis being crucial for improving patient outcomes and extending life expectancy. This study introduces a novel Alzheimer's prediction model that leverages Internet o...

Full description

Saved in:
Bibliographic Details
Published in:2024 Second International Conference on Intelligent Cyber Physical Systems and Internet of Things (ICoICI) pp. 386 - 391
Main Authors: Evangelin Sonia, S.V., Mustare, Narendra B, Sailaja, V., Sunitha, Pamarthi
Format: Conference Proceeding
Language:English
Published: IEEE 28.08.2024
Subjects:
Accuracy
Adaptive Self-Guided Filter
Alpine skiing optimization
Alzheimer's disease
Alzheimer's prediction model
Brain modeling
Convolutional Encoders and Decoder
Feature extraction
Information filters
Internet of Things
Neuroimaging
Optimization
Predictive models
Transformers
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Alzheimer's disease (AD) represents a significant challenge in the treatment of dementia among older adults, with early diagnosis being crucial for improving patient outcomes and extending life expectancy. This study introduces a novel Alzheimer's prediction model that leverages Internet of Things (IoT) technology and a hybrid Convolutional Encoder-Decoder architecture optimized with Alpine Skiing Optimization. Utilizing the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, the proposed approach begins with Adaptive Self-Guided Filter (ASGF)-based pre-processing to refine raw brain images by removing irrelevant areas. A Convolutional Encoder-Decoder framework, combining U-Net and V-Net architectures with a Point Transformer Network (Con-EDUV-PTransNet), is employed for robust classification of AD. This hybrid method integrates region-of-interest (ROI) segmentation, feature extraction, and classification, and is optimized using Alpine Skiing Optimization to minimize the loss function. The results demonstrate superior performance with an F-Score of 99.18%, accuracy of 99.4%, precision of 99.26%, and recall of 99.87%. These metrics underscore the effectiveness of the Con-EDUV-PTransNet model in distinguishing between healthy individuals and those with AD. The experimental results affirm the model's effectiveness, offering a highly accurate and reliable method for AD diagnosis that could significantly enhance patient outcomes through improved brain imaging analysis.
DOI:10.1109/ICoICI62503.2024.10696255