A comparative study of features selection for skin lesion detection from dermoscopic images
- PDF / 593,938 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 68 Downloads / 260 Views
(2020) 9:4
REVIEW ARTICLE
A comparative study of features selection for skin lesion detection from dermoscopic images Rabia Javed1,2 · Mohd Shafry Mohd Rahim1 · Tanzila Saba3 · Amjad Rehman3 Received: 9 April 2019 / Revised: 29 September 2019 / Accepted: 22 November 2019 © Springer-Verlag GmbH Austria, part of Springer Nature 2019
Abstract Melanoma is rare and mainly considered as the dangerous category of skin cancer. Many researchers proposed diverse efficient techniques for melanoma detection. The main focus of this research is: (1) to discuss the traditional clinical methods for diagnosing skin cancer melanoma, and (2) to review the existing researcher’s attempts in response the critical and challenging task is features selection and extraction for skin cancer melanoma detection from dermoscopy images. This research will also be helpful to recognize the research background of skin cancer melanoma detection through image processing techniques. This cannot be done without a broad literature survey. The literature survey was performed keeping the main category as skin cancer melanoma and the survey included articles, journals, and conferences papers. To perform this study, different databases are considered. All of these databases cover medical image processing and technical literature. To conclude the review, some graphs and tables are presented which perform the comparison between existing techniques. Keywords Skin cancer · Biopsy · Dermatoscopy · ABCD rule · Melanoma Abbreviations DNA Deoxyribonucleic acid UVA Ultraviolet A UVB Ultraviolet B MRI Magnetic resonance imaging OCT Optical coherence tomography CLSM Confocal laser scanning microscopy ELM Epiluminescence microscopy ABCDE Asymmetry, border, color, diameter, and evolving CASH Color, architecture, symmetry, and homogeneity CADx Computer-aided diagnosis system SIFT Scale-invariant feature transform LBP Local binary pattern RGB Red, green, blue HSB Hue, saturation, brightness * Amjad Rehman [email protected] 1
School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor Bahru, Malaysia
2
Department of Computer Science, Lahore College for Women University, Jail Road, Lahore 54000, Pakistan
3
College of Computer and Information Sciences, Prince Sultan University, Riyadh 11586, Saudi Arabia
HSL Hue, saturation, lightness HSV Hue, saturation, value YUV Luminance, two chrominance YCbCr Luminance, chrominance CMYK Cyan, magenta, yellow, black OOP Opponent SIFT Scale invariant feature transform GLCM Gray level co-occurrence matrix GMRF Gaussian Markov random field AR Autoregressive fBm Fractional brownian motion PCA Principal component analysis WPT Wavelet packet transform KNN K-nearest neighbor ANN Artificial neural network SVM Support vector machine MLP Multilayer perceptron RABGLD Regional average binary gray level difference co-occurrence matrix SFTA Segmentation based fractal texture analysis CNN’s Convolutional neural networks DCNNs Deep convolutional neural network
Data Loading...
