Table 5 Classification results obtained when the models are run (5 slices with SSIM).

From: Improving early detection of Alzheimer’s disease through MRI slice selection and deep learning techniques

Models

Accuracy

Precision

Recall

F1 Score

AUC

MCC

EfficientNetB2

0.9688

0.9312

0.9110

0.9178

0.9886

0.9125

InceptionV3

0.8875

0.7533

0.6508

0.6995

0.9345

0.6912

RegNetx006

0.9889

0.9733

0.9713

0.9701

0.9981

0.9672

Basic ViT

0.9362

0.8487

0.8288

0.8390

0.9777

0.8354

EfficientNetB2 + FPN (Our Model)

0.9874

0.9734

0.9633

0.9671

0.9974

0.9610

ViT + EfficientNetB2 + FPN (Our model)

0.9517

0.8593

0.7585

0.8065

0.9078

0.8041

Back to article page