An Efficient Internet of Things (IoT)-Enabled Skin Lesion Detection Model using Hybrid Feature Extraction with Extreme Machine Learning Model

Abstract

At present times, Internet of Things (IoT)-based healthcare diagnosis models become more popular and applicable in diverse scenarios. Skin lesion segmentation acts as an important part in the early identification of skin cancer using automated diagnosis models. The automated detection and classification of skin lesion is a critical task because of the constraints such as artefacts, unsure boundaries, and different shapes of the lesion images. This study introduces a new IoT based automated skin lesion detection model. The proposed model involves a series of processes, namely data acquisition using IoT devices, bilateral filtering-based preprocessing, K-means clustering-based segmentation, hybrid feature extraction, and extreme learning machine (ELM)-based classification. The HF-ELM model determines the identification of lesions exist in the dermoscopic images. The HF-ELM model undergo simulation using skin image dataset, and the simulation results indicated the effective performance of the presented model over the other methods.