Core Concepts
Using deep-learning models and ensemble techniques, accurate early detection of Alzheimer's disease is achievable with high precision and recall rates.
Abstract
Alzheimer's disease progression stages: CN, pMCI, sMCI, AD.
Deep-learning models VGG16, AlexNet, custom CNN used for classification.
Ensemble method improves overall accuracy to 93.13% with an AUC of 94.4%.
Image processing steps include co-registration, normalization, segmentation.
Metrics like accuracy, precision, recall, and AUC are crucial for evaluation.
Results show Custom CNN outperforms other models in both AD/CN and pMCI/sMCI classifications.
Stats
The results show that using deep-learning models gives an overall average accuracy of 93.13% and an AUC of 94.4%.
Quotes
"The majority voting method is used to combine the models' results."