Meta-learning for skin cancer detection using Deep Learning Techniques

TL;DR

This paper demonstrates that meta-learning with pre-trained models on non-medical images can significantly improve skin cancer detection accuracy, addressing distribution shift issues with limited dermoscopic data.

Contribution

It introduces a meta-learning approach that leverages non-medical image features to enhance skin cancer classification performance under limited data conditions.

Findings

Performance increased by 20 points using ImageNet features.

Features from everyday images aid in skin lesion classification.

Distribution balance impacts model accuracy.

Abstract

This study focuses on automatic skin cancer detection using a Meta-learning approach for dermoscopic images. The aim of this study is to explore the benefits of the generalization of the knowledge extracted from non-medical data in the classification performance of medical data and the impact of the distribution shift problem within limited data by using a simple class and distribution balancer algorithm. In this study, a small sample of a combined dataset from 3 different sources was used to fine-tune a ResNet model pre-trained on non-medical data. The results show an increase in performance on detecting melanoma, malignant (skin cancer), and benign moles with the prior knowledge obtained from images of everyday objects from the ImageNet dataset by 20 points. These findings suggest that features from non-medical images can be used towards the classification of skin moles and that the distribution of the data affects the performance of the model.