Pre-trained convolutional networks show varying accuracy in melanoma detection
A Comparative Evaluation of Pre-trained Convolutional Neural Networks for Melanoma Detection
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Melanoma, a dangerous type of skin cancer, is hard to identify early because skin lesions can look very similar. The authors compared different pre-trained convolutional neural networks (CNNs), which are computer models trained to recognize images, to see how well they detect melanoma using two types of skin images. They found that the best CNN models achieved accuracy between 71% and 84%, but performance varied depending on the type of image used. This study helps identify which CNN models might work best for different skin image types to support doctors in diagnosing melanoma.
What this means in practice
- •For dermatology clinics: Choose the best CNN model to assist automatic melanoma detection from different types of skin images, improving diagnostic support tools.
- •For medical ai developers: Design diagnostic software that uses the best-performing pre-trained CNN for specific skin image types to improve melanoma screening accuracy.$Commercial implications: Enables creation of targeted AI-powered diagnostic tools that can be commercialized for healthcare providers.
Authors
Wagner Moreno Schmitz, Marco Antonio de Castro Barbosa, Thiago Magalhães Amaral, Dalcimar Casanova, Jefferson Tales Oliva
Abstract
Early diagnosis of melanoma is critical for improving patient survival rates. However, accurately distinguishing melanoma from other skin lesions remains a significant clinical challenge due to the high visual similarity among lesion types and variability in image acquisition conditions. Artificial intelligence, particularly machine learning, has emerged as a promising tool to support dermatological diagnosis by automating feature extraction from medical images. Among the available approaches, convolutional neural networks (CNNs) have demonstrated strong performance in image classification tasks, making them well-suited for analyzing both dermatoscopic and histopathological images, given their ability to capture hierarchical visual patterns relevant to lesion characterization. Nevertheless, despite numerous pre-trained CNN architectures having been proposed, selecting the most appropriate one for a given imaging modality remains an open challenge. In this study, we evaluate pre-trained convolutional neural networks (CNNs) for skin lesion classification using dermatoscopic and histopathological image datasets. Experiments were conducted on the HAM10000, ISIC 2018, and CR-AI4SkIN datasets, evaluating the ResNet50, VGG16, VGG19, MobileNet, and InceptionV3 architectures under the same training protocol. The experimental evaluation showed that the models achieved accuracies ranging from 71% (InceptionV3 on ISIC 2018) to 84% (ResNet50 on HAM10000) on dermatoscopic images. For histopathological images, accuracies ranged from 72% (VGG19) to 83% (ResNet50) on the CR-AI4SkIN dataset. The results demonstrate that model performance differs between dermatoscopic and histopathological image modalities, showing that architectures exhibiting similar performance on dermatoscopic images exhibit different performance on histopathological data.