Recognition of pigment network pattern in dermoscopy images based on fuzzy classification of pixels

Abstract

The main contribution of this method is the very design of the algorithm, highly innovative, which could also be used to deal with other pattern recognition problems of a similar nature. Other contributions are: 1. The good performance in discriminating between the pattern and the disturbing artefacts -which means that no prior preprocessing is required in this method- and between the pattern and other dermoscopic patterns; 2. It puts forward a new methodological approach for work of this kind, introducing the system specification as a required step prior to algorithm design and development, being this specification the basis for a required parameterisation -in the form of configurable parameters (with their value ranges) and set threshold values- of the algorithm and the subsequent conducting of the experiments.

An Enhanced Transfer Learning Based Classification for Diagnosis of Skin Cancer

“…The overall accuracy of the model was 76.5% with sensitivity of 96.3% and specificity of 89.5%. Garcia Arroyo et al [ 15 ] presented an algorithm based on machine learning and used 875 dermoscopic images. The images were collected from the Interactive Atlas of Dermoscopy dataset.…”

Machine learning (ML) techniques as effective methods for evaluating hair and skin assessments: A systematic review

“…26 A similar system was proposed and analyzed where network of pigments within dermoscopic images were classified. 32 A large database comprising of 875 images was acquired and a fuzzy classifier was applied that helped to generate a fuzzy classification model from the values of labeled pixels. These values in turn helped in the extraction of features such as color and texture and these further retrieved different categories of results including accuracy, sensitivity, specificity, and AUC.…”

“…The system achieved an accuracy of 88%, sensitivity of 90.17% and specificity of 83.44%, and an AUC of 0.912. 32 A comparison can be drawn between these two studies and an inference can be made regarding the use of fuzzy classifier as an efficient, reliable, and robust mechanism that enables classification of both skin lesion boundaries in order to diagnose Melanoma as well as detect pigment networks in dermoscopic images. 32 Another inference that can be made relates to the sample size.…”

“…32 A comparison can be drawn between these two studies and an inference can be made regarding the use of fuzzy classifier as an efficient, reliable, and robust mechanism that enables classification of both skin lesion boundaries in order to diagnose Melanoma as well as detect pigment networks in dermoscopic images. 32 Another inference that can be made relates to the sample size. In case, the sample size tends to increase with larger number of images being utilized, the computation cost and computation time along with computational complexity increases that has an adverse effect on the accuracy, sensitivity, and specificity values.…”

Development of a methodology of computational intelligence for robust composite biomarker discovery

“…This part of the thesis was mostly focused on comparing classification methods and investigating which machine learning method delivers the best results. A number of different classifiers have been utilised, with neural networks, support vector machines and discriminant analysis achieving the highest accuracy (d'Amico et al, 2004;Arroyo and Zapirain, 2014;Blanzieri et al, 2000;Celebi et al, 2009;Dreiseitl et al, 2001;Garcia-Arroyo and Garcia-Zapirain, 2018;Grana et al, 2003;Maglogiannis et al, 2001Maglogiannis et al, , 2005Maglogiannis and Kosmopoulos, 2006;Maglogiannis and Zafiropoulos, 2004;Messadi et al, 2009;Nasir et al, 2018;Rahman and Bhattacharya, 2010;Sadeghi et al, 2011;Serrano and Acha, 2009;Tanaka et al, 2004;Yuan et al, 2006;Zhang et al, 2003). Other groups were concerned mostly about the feature extraction step, using statistical methods and supervised learning to lower the feature level, to achieve better performance on classification.…”