Image based skin disease detection system

Image based skin disease detection system

The current work proposes a neural based detection method of two different skin diseases using skin imaging. Skin images of two diseases namely Basel Cell Carcinoma and Skin Angioma are utilized. SIFT feature extractor has been employed followed by a clustering phase on feature space in order to reduce the number of features suitable for neural based models. The extracted bag-of-features modified dataset is used to train metaheuristic supported hybrid Artificial Neural Networks to classify the skin images in order to detect the diseases under study. A well-known multi objective optimization technique called Non-dominated Sorting Genetic Algorithm – II is used to train the ANN (NN-NSGA-II). The proposed model is further compared with two other well-known metaheuristic based classifier namely NN-PSO (ANN trained with PSO) and NN-CS (ANN trained with Cuckoo Search) in terms of testing phase confusion matrix based performance measuring metrics such as accuracy, precision, recall and F-measure. Image based skin disease detection system

In the field of medical image analysis, skin disease analysis is one of the major do-main of interest. Skin cancer is one of the major disease and many people suffers and dies due to this image throughout the world every year. In order to support dermatologists, machine learning based methods are used for detection and classification. Image processing accelerates the process by computing different feature from the mages. Different dermoscopy methods have been proposed to improve and enhance the diagnostic performance. Dermoscopy is a method for skin imaging which is noninvasive in nature. Magnified and illuminated picture of a certain region of skin can be obtained and it helps to clarity
the spots on the skin

The ANN is one of the most used modeling approaches [2-3]. It achieves accurate classification even with very small dataset. It can handle imprecise relationships during its training stage. The ANN structure is consists of interconnected computational neurons, which involved in the mathematical mapping through the learning process, which attempt to adjust the weight value. Initially, the training phase is started by a part of the dataset to classify its inputs along with its class label to create the classification model. Afterward, the validation phase is performed to confirm the effectiveness of the trained model using another dataset. Finally, the evaluation phase is used to test the classification model accuracy using another set of test data. In general, the artificial neuron uses the input signal (x) and their equivalent weights (w) to form the input (N J ). This input is then surpassed to a linear threshold filter till it exceeds the output signal (y) to another neuron. If N j exceeds the threshold of that neuron, the neuron is inspired.

Feature extraction refers to the process of extracting some meaningful information from some initial set of data. It should be informative and non-redundant. It helps us to reduce the volume of resources needed to describe an object. An analysis of different feature extraction methods and their applications are given in [13-15]. One of the major problems related with data analysis is the number of variables involved. Good amount of memory as well as time is needed to analyze a large number of variables. We also need high computation power to perform such kind of analysis. In some cases it may overfit a classifier and reduces its efficiency. Feature extraction helps us to bypass these issues by forming some combinations of these variables in such a way that it can convey come information. In this paper, Scale Invariant Feature Transform (SIFT) has been used to detect interest points and find the corresponding descriptors. SIFT has been proved as a robust method and effective in different images

Skin cancer is one of the global threats and needs serious attention. Automated detection is extensively needed not only to reduce the work load; it is also required for early stage detection. The proposed method seems to be effective and can be employed in real world applications. Moreover, SIFT based bag-of-feature method can be employed in different modality of bio-medical image analysis as well as other domains like satellite imaging etc. Other metaheuristc methods can be integrated to train the artificial neural networ