Automatic segmentation of dermoscopy images using self-generating neural networks seeded by genetic algorithm

• Published in: Pattern recognition Vol. 46; no. 3; pp. 1012 - 1019
• Main Authors: Xie, Fengying, Bovik, Alan C.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.03.2013; Elsevier
• Subjects: Applied sciences; Artificial intelligence; Automatic segmentation; Clusters; Computer science; control theory; systems; Connectionism. Neural networks; Dermoscopy images; Exact sciences and technology; Generic algorithms; Genetic algorithms; Image clustering; Image processing; Information theory; Information, signal and communications theory; Neural networks; Samples; Segmentation; Self-generating neural network; Signal and communications theory; Signal processing; Signal representation. Spectral analysis; Signal, noise; Statistical analysis; Statistical methods; Telecommunications and information theory; Self-generating neural network; Image clustering; Automatic segmentation; Generic algorithms; Dermoscopy images; Automatic classification; Image processing; Background; Threshold detection; K means algorithm; Neural network; Signal classification; Optimization; Learning; Image segmentation; Accuracy; Genetic algorithm; Imaging; Information processing; Data fusion
• ISSN: 0031-3203; 1873-5142
• DOI: 10.1016/j.patcog.2012.08.012

A novel dermoscopy image segmentation algorithm is proposed using a combination of a self-generating neural network (SGNN) and the genetic algorithm (GA). Optimal samples are selected as seeds using GA; taking these seeds as initial neuron trees, a self-generating neural forest (SGNF) is generated by training the rest of the samples using SGNN. Next the number of clusters is determined by optimizing the SD index of cluster validity, and clustering is completed by treating each neuron tree as a cluster. Since SGNN often delivers inconsistent cluster partitions owing to sensitivity relative to the input order of the training samples, GA is combined with SGNN to optimize and stabilize the clustering result. In the post-processing phase, the clusters are merged into lesion and background skin, yielding the segmented dermoscopy image. A series of experiments on the proposed model and the other automatic segmentation methods (including Otsu's thresholding method, k-means, fuzzy c-means (FCM) and statistical region merging (SRM)) reveals that the optimized model delivers better accuracy and segmentation results. ► Self-generating neural network is improved through generalizing SGNT to SGNF. ► GA is combined with SGNN to optimize and stabilize the clustering result. ► The SD validity index is used to automatically determine the number of clusters. ► The post-processing is carried on the clustering regions to segment image accurately.