Hosmi-LBP-Based Feature Extraction for Melanoma Detection Using Hybrid Deep Learning Models

Abstract

“Melanoma is a serious form of skin cancer that begins in cells known as melanocytes and more dangerous due to its spreading ability to other organs more rapidly if it is not treated at an early stage”. This paper aims to propose a Melanoma detection methodology that includes four major phases: “(i) pre-processing (ii) segmentation (iii) the proposed feature extraction and (iv) classification”. Initially, pre-processing is performed, where the input image is subjected to processing like resizing and edge smoothening. Subsequently, segmentation is carried out by the Otsu thresholding process. In the feature extraction phase, the proposed Higher-Order Standardized Moment Induced-Local Binary Patterns (HOSMI-LBP)-based features are extracted. These features are then subjected to a classification process for classifying the disease. For this, it is planned to use a hybrid classification framework, where the Convolutional Neural Network (CNN) and the Neural Network (NN) are deployed. Two-phase of classification gets processed: the extracted features are subjected to NN; the input image is directly classified using an optimized CNN framework. Finally, the classified outputs from NN and optimized CNN are averaged and the final output is considered as detected output. Particularly, the weight and initial rate of CNN is optimized using the proposed algorithm known as the Sea Lion Integrated Grey Wolf Algorithm (SLI-GWO) method that hybrid the concepts of both Sea Lion Optimization (SLnO) and Grey Wolf Optimization (GWO) algorithm. At last, the proposed work performance is computed with traditional systems in terms of various measures.

SNC_Net: Skin Cancer Detection by Integrating Handcrafted and Deep Learning-Based Features Using Dermoscopy Images

“…For data analysis, the ISBI 2016 and ISIC 2017 datasets are utilized. In order to classify the data, Kumar et al [45] mixed a CNN and neural networks in a hybrid way. A neural network is supplied with features extracted from HOSMI-LBP to aid in the process of categorization.…”

“…The main goal of previous studies [27,[46][47][48][49][50][51][52][53][54] was to distinguish melanoma from non-melanoma cases using dermoscopy images. Multiple researches based on deep learning use dermoscopy images to detect various types of skin malignancies [13][14][15][16][41][42][43][44][45][46][47]. The task of automatically classifying skin cancer in dermoscopic images is difficult because of the high levels of visual similarity and intraclass variation.…”