Improving Diagnostic Accuracy of Pigmented Skin Lesions With CNNs: an Application on the DermaMNIST Dataset

TL;DR

This paper evaluates CNN models like ResNet-50 and EfficientNetV2L on the DermaMNIST dataset to improve pigmented skin lesion classification accuracy, demonstrating the potential of deep learning in biomedical diagnostics.

Contribution

It introduces an assessment of advanced CNN architectures with transfer learning on DermaMNIST, achieving state-of-the-art or comparable results in pigmented skin lesion classification.

Findings

One configuration outperforms existing methods.

CNNs significantly improve diagnostic accuracy.

Transfer learning enhances model performance.

Abstract

Pigmented skin lesions represent localized areas of increased melanin and can indicate serious conditions like melanoma, a major contributor to skin cancer mortality. The MedMNIST v2 dataset, inspired by MNIST, was recently introduced to advance research in biomedical imaging and includes DermaMNIST, a dataset for classifying pigmented lesions based on the HAM10000 dataset. This study assesses ResNet-50 and EfficientNetV2L models for multi-class classification using DermaMNIST, employing transfer learning and various layer configurations. One configuration achieves results that match or surpass existing methods. This study suggests that convolutional neural networks (CNNs) can drive progress in biomedical image analysis, significantly enhancing diagnostic accuracy.