MARK1 publications to scientific conferences
|1.||Perakis K., Bouras T., Kostopoulos S., Sidiropoulos K., Wayn L. & Timor H., (2014), MARK1 – A Decision Support System for the Early Detection of Malignant Melanoma. Accepted at 4th International Conference on Wireless Mobile Communication and Healthcare – “Transforming healthcare through innovations in mobile and wireless technologies” (MobiHealth), Athens, Greece, 3-5 November 2014.
Abstract. Early stage detection of melanoma (one of the most common cancers today) is of major significance for increasing chances of long term survival of affected patients. Over the last decade there have been developments in skin diagnostics, facilitated by the use of technologies such as Total Body Photography (TBP), which provide a complete record of the skin, and by the development of applications on handheld devices seeking to characterize skin lesions as part of routine self-examination. Unfortunately, these processes are rather inefficient, inaccurate, and not fully automated, missing also critical components such as the automated ability to compare between two TBP image sets in order to locate essential new and altered skin lesions. Despite some progress and because of its many flaws, the common practice today for early detection is skin self-examination. However, it is important to note that skin self-examination is usually underestimated by individuals, resulting in poor prognosis. The main objective of the present paper is to present the conceptual architecture of a platform that can address the need for early and accurate detection of skin lesion through a screening solution that will be easily accessible to the general public with the guidance, supervision and inspection of the primary care physician.
|2.||S. Kostopoulos, D. Glotsos, K. Sidiropoulos, P. Asvestas, Ch. Konstandinou, G. Xenogiannopoulos, E-K. Nikolatou, K. Perakis, Th. Bouras and D. Cavouras, MARK1 – early stage detection of melanoma through a smartphone compatible designed decision support application, International Conference on Bio-Medical Instrumentation and related Engineering and Physical Sciences (BIOMEP 2015), Athens, Greece, June 18-20, 2015.
Abstract. Malignant melanoma is the most deathful skin cancer in which melanocytes in the epidermis undergo malignant transformation. The main cause of melanoma is due to a long exposition to ultraviolet radiations, although skin type or other genetic factors can influence too. The most effective treatment is an immediate extirpation, but just when the melanoma had been detected in early phases. MARK1 project is a co-funded GREECE-ISRAEL research cooperation that aims in developing a self-examination platform/application able to adjust in commercial smartphone or handheld devices for facilitating early stage detection of melanoma. MARK1 will incorporate image processing, image analysis and pattern recognition techniques for mole identification, analysis and classification. The application will be linked with the medical specialist who will be able to advice regarding the urgency for a medical examination. In this way, MARK1 aims in early and accurate screening of melanoma by an easy to use technology accessible by general public, in order to enable fast, offline supervision of suspected moles by the health care specialist for detecting the disease at its early stages, when it is more vulnerable to available treatments.
|3.||Glotsos D., Kostopoulos S., Lalissidou S., Sidiropoulos K., Asvestas P., Konstandinou C., Xenogiannopoulos G., Nikolatou E., Perakis K., Bouras T., Cavouras D. (2015) Design of a decision support system, trained on GPU, for assisting melanoma diagnosis in dermatoscopy images. 4th International Conference on Mathematical Modeling in Physical Sciences. Mykonos, Greece, June 5-8, 2015.
Abstract. The purpose of this study was to design a decision support system for assisting the diagnosis of melanoma in dermatoscopy images. Clinical material comprised images of 44 dysplastic (clark’s nevi) and 44 malignant melanoma lesions, obtained from the dermatology database Dermnet. Initially, images were processed for hair removal and background correction using the Dull Razor algorithm. Processed images were segmented to isolate moles from surrounding background, using a combination of level sets and an automated thresholding approach. Morphological (area, size, shape) and textural features (first and second order) were calculated from each one of the segmented moles. Extracted features were fed to a pattern recognition system assembled with the Probabilistic Neural Network Classifier, which was trained to distinguish between benign and malignant cases, using the exhaustive search and the leave one out method. The system was designed on the GPU card (GeForce 580GTX) using CUDA programming framework and C++ programming language. Results showed that the designed system discriminated benign from malignant moles with 88.6 % accuracy employing morphological and textural features. The proposed system could be used for analysing moles depicted on smart phone images after appropriate training with smartphone images cases. This could assist towards early detection of melanoma cases, if suspicious moles were to be captured on smartphone by patients and be transferred to the physician together with an assessment of the mole’s nature.
|4.||Glotsos D., Kostopoulos S., Lalissidou S., Sidiropoulos K., Asvestas P., Konstandinou C., Xenogiannopoulos G., Nikolatou E., G. Sakellaropoulos, Perakis K., Bouras T., Cavouras D. (2015) Segmentation of skin cancer images using a structured sequence of image processing steps for optimizing decision support consultations in melanoma diagnosis. International Conference ‘Science and Technology’. Athens, Greece, November 5-7, 2015.
Abstract. The aim of this study is to design, develop and implement a structured sequence of image processing steps to facilitate mole segmentation in skin cancer images. The proposed segmentation scheme consists the DullRazor method, which is utilized for hair removal, the mean shift filtering algorithm, which is used for image smoothing, the Otsu’s thresholding approach, which is utilized for a preliminary estimate of the mole region boundaries and the region growing algorithm, which is applied to finalize the segmentation result. The proposed method was integrated into MARK1 platform, which is a web application and API technology used for early stage detection of melanoma. The accuracy of MARK1 platform in predicting melanoma cases converged to 85 % using the proposed segmentation framework.
MARK1 publications to scientific journals