Classification of Skin Lesions

Undiagnosed skin lesions can be photographed and the images fed into a computer system that potentially differentiates malignant from benign lesions. To develop a melanoma detection system, various methods from statistics, machine learning and image analysis are applied. An image consists of millions of pixels, so reducing the enormous amount of data is an important part of image analysis. This can be done by probability density estimation and clustering. In hierarchical clustering, the dissimilarity measure has great influence on the final clustering, but there has been little focus on how to choose an adequate dissimilarity measure for density-based techniques. In this thesis, six properties for density-based dissimilarity measures are therefore proposed as a guide for the user, based on assumptions and previous knowledge about the data set. An image cannot be fed directly into a classifier because of the amount of data contained in each image; therefore a set of features is extracted from one. In melanoma detection, the colour of the lesion is of special interest. This thesis presents several approaches to colour feature extraction. (i) By clustering the pixel values and then comparing the cluster centres to pre-defined colour values, melanoma-indicative colours are detected. (ii) By estimating the probability density, and then measuring the goodness-of-fit, the variation in colours is accounted for. (iii) By the use of a dissimilarity measure, an unclassified lesion's similarity to the melanoma class or the class of benign lesions can be calculated. Different methods for feature evaluation are discussed. A thorough presentation of computer systems for melanoma detection is provided, and some of the key elements are discussed. The challenge of feature selection and classifier selection is given special attention. A computer system for melanoma detection, "Nevus Doctor", is presented. It is based on semi-automatic feature selection of both new and previously developed features, and a new hybrid classifier. The performance of the system in terms of sensitivity and specificity scores is presented and compared to that of a commercially available system for the same set of lesions. This methodology has previously been used once only, and then in a smaller study. Obstacles associated with small data sets are discussed, including cross-validation and clinical relevance. "Nevus Doctor" performed better than the commercially available system. The new colour features add value in computer-aided melanoma detection, both by improving the existing system and by introducing a new class of features. The properties for dissimilarity measures offer a new perspective on clustering and other fields where dissimilarity measures are a core element. In fair skinned populations melanoma is responsible for more than 90 % of all skin cancer deaths. Melanoma may arise at virtually any age. The diagnosis is challenging since melanoma initially may resemble a common mole. The treatment of melanoma is surgery, i.e. the cancer tissue is cut out from the skin. If surgery is performed at an early stage, the patient will almost always be cured. In contrast, late diagnosis may be associated with a fatal outcome in more than 50% of cases. Removal of a suspicious skin lesion suggestive of melanoma is easy in the majority of cases. The procedure can often be performed by general practitioners. The main challenge is to recognise a suspicious pigmented skin lesion at an early stage. Since there is a lack of dermatologists in many countries general practitioners are diagnosing a substantial amount of skin lesions. There has been increased interest in computer-aided decision support systems. A melanoma detection system may help doctors in general practice to improve their diagnostic skills and improve the outcome of these skin cancer patients. This research has shown the feasibility of a computer-aided diagnosis system. In addition there a promising results regarding the diagnosis of other skin cancer types by the same system. However, the true performance in a real clinical environment remains to be tested. A randomised controlled trial to investigate the performance of the computer-aided diagnosis system «Nevus doctor» in a general practice setting is currently underway.