Abstract:Skin cancer is one of the world’s scariest diseases, having taken the lives of thousands of people. It can be treated if diagnosed at the right time. According to WHO, every year, approximately 3 million non-melanoma and 130,000 malignant melanomas occur worldwide. Many existing technologies have demonstrated that computer-aided systems can be useful in the early identification of cancer. One of the major challenges in computer-aided diagnostic systems is accurate segmentation of the lesion and extraction of features for successful classification and detection. The study’s main goal is to recognize and segment cancerous parts of skin from the collected samples and then categorize them into separate affected and non-affected regions. The proposed model first performs the identification and separation of infected regions from the sample. This is performed by converting the RGB cell image into a greyscale colour scale. The background subtraction approach is used to track only cell structures from the image by eliminating the background, and region props are applied for segmentation from skin images. In the second phase, features from the segmented images are extracted. These features include homogeneity, contrast, energy, correlation, and some hybrid features. In the third phase, the differential analyzer approach (DAA) algorithm is used to select the significant features. In the final phase, the efficiency of the suggested optimization method is validated using different classifiers. The suggested methodology is applied to an ISIC 2018 dataset available. This result outperforms all other published papers that used the same dataset. Classification accuracy is notably higher in comparison to other approaches not following the DAA optimization algorithm. Validation of results is further extended through feature reduction ratio and still remarkable results concerning classification accuracy of 96% are achieved. The validity of the approach is examined using different classifiers including KNN, SVM, Naïve Bayes, decision tree, and random forest-based mechanism.

Computer-Aided Classification of Melanoma: A Comprehensive Survey

State-of-the-art machine learning techniques for melanoma skin cancer detection and classification: a comprehensive review

Detection and Classification of Melanoma Skin Cancer Using Image Processing Technique

An Overview of Melanoma Detection in Dermoscopy Images Using Image Processing and Machine Learning

Computational Intelligence-Based Melanoma Detection and Classification Using Dermoscopic Images

Skin Cancer Diagnostics with an All-Inclusive Smartphone Application

Melanoma Skin Cancer Detection Using Deep learning, Neural Network and Classical Machine Learning Techniques

A Computer-Aided Spectroscopic System for Early Diagnosis of Melanoma

Skincure: An Innovative Smart Phone-Based Application To Assist In Melanoma Early Detection And Prevention

Melanoma Skin Cancer Detection and Classification using Deep Learning

A Survey on Deep Learning based Computer-Aided Diagnosis Model for Skin Cancer Detection using Dermoscopic Images

A Comprehensive Evaluation Study on Risk Level Classification of Melanoma by Computer Vision on ISIC 2016-2020 Datasets

Detection for melanoma skin cancer through ACCF, BPPF, and CLF techniques with machine learning approach

Skin cancer detection and classification based on differential analyzer algorithm

Melanoma Classification Using a Novel Deep Convolutional Neural Network with Dermoscopic Images

Towards Automated Melanoma Screening: Proper Computer Vision & Reliable Results

Computer Based Melanocytic and Nevus Image Enhancement and Segmentation

AI-Powered Diagnosis of Skin Cancer: A Contemporary Review, Open Challenges and Future Research Directions

A Comparative Analysis Towards Melanoma Classification Using Transfer Learning by Analyzing Dermoscopic Images

Automated Detection of Nonmelanoma Skin Cancer Based on Deep Convolutional Neural Network

Method to Classify Skin Lesions using Dermoscopic images