An improved Internet-based melanoma screening system with dermatologist-like tumor area extraction algorithm

Iyatomi, Hitoshi; Oka, Hiroshi; Celebi, M. Emre; Hashimoto, Masahiro; Hagiwara, Masafumi; Tanaka, Masaru; Ogawa, Kōichi

doi:10.1016/j.compmedimag.2008.06.005

Cited by 218 publications

(135 citation statements)

References 34 publications

Supporting

Mentioning

135

Contrasting

Order By: Relevance

“…The proposed scheme is Wavelet Transformation for image improvement, denoisingand Histogram analysis and finally fuzzy interference system for final decision of skin type based. Combination of ABCD rule and wavelet coefficient shown classification accuracy by 60%.Hitoshi Iyatomi et al [7] proposed an internet based melanoma screeing system which involves area extraction algorithm which calculates characterization of the tumor. The sensitivity achieves is 85.9% and Specificity is 86%.Nagaraj et al [8] have proposed a technique called automatic segmentation of skin lesion in traditional macroscopic images.…”

Section: Complete Life Cycle Of An Automated System For the Detecmentioning

confidence: 99%

Melanoma Detection from Dermoscopic Images Using Gaussian Mixtures and Artificial Neural Network

Sharma¹

2017

IJARCSSE

View full text Add to dashboard Cite

The malignant melanoma is the most deadly form of skin cancer. Early diagnosis is an important that increases chances of successful cureas there is rapid course of the disease. It is costly for dermatologist to screen every patient for melanoma. There is a need of system to assess the lesion area based on dermatological photographs. Image processing andComputer analysis are the efficient tools which helps medical diagnosis. So to get the effective results and information of various stages of the infected portion, there is need of features of that particular area .So the feature extraction phase is hugely dependent on the region detected which has the disease. So suitablesegmentation algorithm is required which can affectivity detect the skin melanoma pixels in the information image. In this work, we proposed an algorithm which can detect the pixels which have melanoma region and ordinary skin. In proposedwork Gaussian mixtures posterior algorithm is presented. Which chooses some candidates from different regions of the images which have different intensity values and further Gaussian models have been built from the chosen places. In the end Artificalneural network has been implemented to get final segmentation results. Experimental results show that the algorithm gives 97% accuracy results on the tested database images.Keywords-Melanoma, Segmentation, GMM, ANN. I. INTRODUCTIONComputer-aided diagnosis (CAD) systems have been quickly being developed over the previous decade for skin cancer classification. In fact, an impressive role of CAD systems is to give a "second opinion" to the dermatologists in their decision making for successful diagnosis of patients. In dermatology, the significant types of lesions for skin cancer are divided into malignant melanoma and non-melanoma. Melanoma is a most frequent type of the melanocytes it is a type of cell that is found in the skin"s epidermis and its incidence has been increasing over last few decade. Melanoma mostly occurs on the trunk or lower extremities and is the most malignant form of skin cancer [1] .A cancerous lesion in the pigment bearing basal layer of theskin is the most treatable and with the cure rate of 100% in the early stage. Theproblem consists in identifying the small percentage of skin lesion that develops melanoma [2]. The advanced cutaneous melanoma is still downhearted, but when examined at early stages it can be cured without complications. The separation of early melanoma from other non-malignant pigmented skin lesions is important even for experienced dermatologists. In several cases primary care physicians miscalculate melanoma in its early stage [3].The main idea of image segmentation is to group pixels in homogeneous regions and the approach used is "common feature". The features can be represented by the space of color, texture and gray level [4]. In this paper, the problems of skin image segmentation using a hybrid path for the separation of pigmented skin lesions from normal skin and the feature extraction from the separated regi...

show abstract

Section: Complete Life Cycle Of An Automated System For the Detecmentioning

confidence: 99%

Melanoma Detection from Dermoscopic Images Using Gaussian Mixtures and Artificial Neural Network

Sharma¹

2017

IJARCSSE

View full text Add to dashboard Cite

show abstract

“…Iyatomi et al [25] built the site in 2004, which is still in operation. They acquired and analyzed 1,258 dermoscopic images via the site, and achieved high sensitivity and speci city of 85.9% and 86.0%, respectively [26]. First, their system extracts the melanoma area from the image using image analysis methods such as thresholding and clustering.…”

Section: Diagnostic Image Analysismentioning

confidence: 99%

Recent Advances in Diagnostic Technologies for Melanoma

Nagaoka

2016

ABE

View full text Add to dashboard Cite

Melanoma is a malignant tumor occurring typically in the skin. Advanced melanoma has very poor prognosis, and is also called the king of cancer . On the other hand, early-stage melanomas are known to be cured completely by relatively simple treatments. Therefore, early detection and treatment of melanomas is very important. However, many melanomas are still diagnosed by subjective evaluation of a physician, in other words, visual inspection. Recently, the demands for development of quantitative melanoma diagnostic technology are increasing. Much research on melanoma diagnosis has progressed in response to these demands. This review introduces some recent advances in melanoma diagnosis using spectral imaging techniques, impedance measurement and diagnostic image analysis, and also brie y describes a novel biological agent against melanoma developed in Japan. A high-quality system using multispectral imaging technique has attracted attention, but the low speci city has been regarded as a problem. It should be noted that descriptions of medical contents are limited because this review is written for engineers and researchers.

show abstract

“…In [40] a service on the Internet was introduced to upload dermoscopy images for on-line extraction of the tumor area and calculation of 428 global features (color, symmetry, border, and texture ABCD) for the characterization of the lesion. The extracted features classified the lesion as melanoma or nevus using a neural network classifier achieving a sensitivity of 85.9% and a specificity of 86.0% on a set of 1258 dermoscopy images using cross-validation.…”

Section: Related Workmentioning

confidence: 99%

Search for Resolution Invariant Wavelet Features of Melanoma Learned by a Limited ANN Classifier

Surówka¹

2017

View full text Add to dashboard Cite

Abstract. This article addresses the Computer Aided Diagnosis (CAD) of melanoma pigmented skin cancer. We present back-propagated Artificial Neural Network (ANN) classifiers discriminating dermoscopic skin lesion images into two classes: malignant melanoma and dysplastic nevus. Features used for our classification experiments are derived from wavelet decomposition coefficients of the image. Our research objective is i) to select the most efficient topology of the hidden layers and the network learning algorithm for full-size and downgraded image resolutions and, ii) to search for resolution-invariant topologies and learning methods. The analyzed classifiers should be fit to work on ARM-based hand-held devices, hence we take into account only limited learning setups.

show abstract

An improved Internet-based melanoma screening system with dermatologist-like tumor area extraction algorithm

Cited by 218 publications

References 34 publications

Melanoma Detection from Dermoscopic Images Using Gaussian Mixtures and Artificial Neural Network

Melanoma Detection from Dermoscopic Images Using Gaussian Mixtures and Artificial Neural Network

Recent Advances in Diagnostic Technologies for Melanoma

Search for Resolution Invariant Wavelet Features of Melanoma Learned by a Limited ANN Classifier

Contact Info

Product

Resources

About