Table 6 A comparative analysis of the relevant studies of MPV model using machine learning methods.
Author | Purpose | Proposed methodology | Key parameters | Model |
|---|---|---|---|---|
Ali et al.70 | MPV skin lesion detection | Utilizing ANN models for detecting MPV | \(F_{1}\)-score | VGG-16, ResNet50, and InceptionV3 models |
Situla and Sahahi71 | MPV detection | Detection of MPV by transfer learning methods | Accuracy and \(F_{1}\)-score | Xception, DenseNet |
Ahsan et al.69 | Detecting MPV | Image data collection and implementation of a deep learning-based model in detecting MPV | AUC | They propose and evaluate a VGG16 model with D curve |
Sahin et al.72 | Human MPV classification from skin lesion images | Human MPV classification from skin lesion images with deep pre-trained network | Accuracy and \(F_{1}\)-score | GoogleNet, EfficientNetb0, Nasnet- Mobile, ShuffleNet, MobileNetv2 models |
Hossain et al.73 | Lyme disease from skin lesion images | NNs with transfer learning to diagnose Lyme disease | AUC, sensitivity, accuracy and specificity | ResNet50 |
Burlina et al.74 | Automated detection of erythema migrans | Automated detection of confounding infection | AUC and accuracy | ResNet50 |
Suggested approach | Poisson random classification of MPV | Poisson noise and LMBNN | Accuracy, precision, recall \(F_{1}\) score | ’ntstool’ and ’narxne’. |
