Classification of skin hyper-pigmentation lesions with multi-spectral images
نویسندگان
چکیده
According to clinical protocols, skin diseases are quanti ed by dermatologists throughout a treatment period, and then a statistical test on these measures allows to evaluate a treatment e cacy. The rst step of this process it to classify pathological interest areas. This task is challenging due to the high variability of the images in one clinical data set. In this report, we rst review algorithms that exist in the literature and adapt them to our problem. Then we choose the more appropriate algorithm to design a classi cation strategy. Thereby, we propose to use data reduction combined with SVM to do a rst classi cation of the disease. Then we associate the obtained classi cation map with a segmentation map in an interactive classi cation tool in order to compromise between operator dependency and algorithm robustness. Key-words: skin, hyper-pigmentation, SVM, data reduction, multi-spectral images ∗ Morphème / Ayin † Morphème ‡ Galderma R&D § Ayin ha l-0 07 45 36 7, v er si on 1 25 O ct 2 01 2 Classi cation de lésions d'hyper-pigmentation cutanée à partir d'images multi-spectrales. Résumé : Lors des protocoles cliniques actuels, les maladies de peau sont quanti ées par les dermatologues tout au long d'une période de traitement. Puis un test statistique sur ces diagnostiques permet d'évaluer l'e cacité d'un traitement. A n d'automatiser un tel processus à l'aide de l'imagerie spectrale, la première étape est d'extraire les zones pathologiques d'intérêt. Cette tache est di cile en raison de la grande variabilité des images dans un ensemble de données cliniques. Dans ce rapport, nous examinons d'abord les algorithmes qui existent dans la littérature et qui peuvent être adaptés à notre problème. Puis, nous choisissons l'algorithme le plus approprié a n de concevoir une stratégie de classi cation. Ainsi, nous proposons d'utiliser la réduction de données combinée avec un séparateur à vaste marge (SVM) pour faire une première classi cation de la pathologie. Ensuite, nous associons la carte de classi cation obtenue avec une carte de segmentation grâce à un outil de classi cation interactive a n de trouver un compromis entre la dépendance à un opérateur et la robustesse de l'algorithme. Mots-clés : peau, hyper-pigmentation, SVM, réduction de dimension, imagerie multi-spectrale ha l-0 07 45 36 7, v er si on 1 25 O ct 2 01 2 Classi cation of skin hyper-pigmentation lesions 3
منابع مشابه
3D Classification of Urban Features Based on Integration of Structural and Spectral Information from UAV Imagery
Three-dimensional classification of urban features is one of the important tools for urban management and the basis of many analyzes in photogrammetry and remote sensing. Therefore, it is applied in many applications such as planning, urban management and disaster management. In this study, dense point clouds extracted from dense image matching is applied for classification in urban areas. Appl...
متن کاملHyper Pigmentation of the Skin following Chloroquine Treatment Case Series Report
In this article, we are collectively presenting case reports of 15 patients who developed pigmentary changes of the skin and mucus membrane during the course of chloroquine therapy for connective tissue disorders. These female patients developed hyper pigmentation of the skin, largely on the exposed parts of the body. The pigmentary changes varied from brownish/ grayish/ bluish-black in color, ...
متن کاملSegmentation and characterization of skin tumors images used for aided diagnosis of melanoma
In this paper, a methodological approach devoted to the segmentation and the characterization of tumors skin lesions is presented. Melanoma is the most malignant skin tumor, growing in melanocytes cells that are responsible of pigmentation. Nowadays, this type of cancer is increasing rapidly. Besides, its related mortality rate increases by more modest and inversely proportional to the thicknes...
متن کاملNon-melanoma skin cancer diagnosis with a convolutional neural network
Background: The most common types of non-melanoma skin cancer are basal cell carcinoma (BCC), and squamous cell carcinoma (SCC). AKIEC -Actinic keratoses (Solar keratoses) and intraepithelial carcinoma (Bowen’s disease)- are common non-invasive precursors of SCC, which may progress to invasive SCC, if left untreated. Due to the importance of early detection in cancer treatment, this study aimed...
متن کاملAnalysis of Human Skin Hyper-Spectral Images by Non-negative Matrix Factorization
This article presents the use of Non-negative Matrix Factorization, a blind source separation algorithm, for the decomposition of human skin absorption spectra in its main pigments: melanin and hemoglobin. The evaluated spectra come from a Hyper-Spectral Image, which is the result of the processing of a Multi-Spectral Image by a neural network-based algorithm. The implemented source separation ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2013