Table 4 Performance comparison of deep learning models based on precision, sensitivity, F1 score, and accuracy, evaluated using the 28-question spot-diagnosis quiz for classifying CVI & DVT, lymphedema, normal, and systemic disease conditions.

From: Deep learning-based classification of lymphedema and other lower limb edema diseases using clinical images

Architecture

Method

Lymphedema

CVI&DVT

Normal

Systemic Disease

All Class

Precision

Sensitivity

F1

Precision

Sensitivity

F1

Precision

Sensitivity

F1

Precision

Sensitivity

F1

Precision

Sensitivity

F1

Accuracy

CNN

AlexNet

0.833

0.625

0.714

0.833

0.625

0.714

0.375

0.500

0.429

0.625

0.833

0.714

0.691

0.643

0.653

0.643

GoogLeNet

0.667

0.750

0.706

1.000

0.500

0.667

0.500

0.500

0.500

0.667

1.000

1.000

0.726

0.679

0.671

0.679

ResNet50

0.750

0.750

0.750

1.000

0.375

0.546

0.250

0.333

0.286

0.667

1.000

0.800

0.696

0.607

0.603

0.607

VGG16

0.667

0.750

0.706

0.833

0.625

0.714

0.400

0.333

0.364

0.750

1.000

0.857

0.675

0.679

0.667

0.679

MobileNetV3

0.800

1.000

0.889

1.000

0.375

0.546

0.600

0.500

0.546

0.600

1.000

0.750

0.771

0.714

0.687

0.714

DenseNet169

0.750

0.750

0.750

0.857

0.750

0.800

0.600

0.500

0.546

0.750

1.000

0.857

0.749

0.750

0.743

0.750

SqueezeNet

0.857

0.750

0.800

1.000

0.500

0.667

0.556

0.833

0.667

0.750

1.000

0.857

0.810

0.750

0.746

0.750

EfficientNetV2

0.857

0.750

0.800

1.000

0.750

0.857

0.667

0.667

0.667

0.667

1.000

0.800

0.816

0.786

0.788

0.786

Transformer

ViT

0.857

0.750

0.800

0.800

0.500

0.615

0.667

0.667

0.667

0.600

1.000

0.750

0.745

0.714

0.708

0.714

TnT

0.833

0.625

0.714

0.778

0.875

0.824

0.500

0.500

0.500

0.714

0.833

0.769

0.721

0.714

0.711

0.714

Swin

0.750

0.750

0.750

1.000

0.625

0.769

0.571

0.667

0.615

0.750

1.000

0.857

0.783

0.750

0.750

0.750

CvT

0.625

0.625

0.625

0.800

0.500

0.615

0.400

0.333

0.364

0.600

1.000

0.750

0.621

0.607

0.593

0.607

PiT

0.857

0.750

0.800

1.000

0.250

0.400

0.400

0.667

0.500

0.667

1.000

0.800

0.759

0.643

0.621

0.643

CCT

0.750

0.750

0.750

0.800

0.500

0.615

0.333

0.500

0.400

0.833

0.833

0.833

0.693

0.643

0.654

0.643

MaxViT

0.667

0.750

0.706

0.667

0.500

0.571

0.400

0.333

0.364

0.750

1.000

0.857

0.627

0.643

0.627

0.643

DaViT

0.750

0.750

0.750

0.875

0.875

0.875

0.400

0.333

0.364

0.714

0.833

0.769

0.703

0.714

0.707

0.714

Back to article page