Segmentation of Skin Cancer Images Based on Gradient Vector Flow (GVF) Snake (original) (raw)
Related papers
Journal of Visual Communication and Image Representation, 2013
Segmentation of ultrasound (US) images of breast cancer is one of the most challenging problems of the modern medical image processing. A number of popular codes for US segmentation are based on a generalized gradient vector flow (GGVF) method proposed by Xu and Prince. The GGVF equations include a smoothing term (diffusion) applied to regions of small gradients of the edge map and a stopping term to fix and extend large gradients appearing at the boundary of the object. The paper proposes two new directions. The first component is diffusion as a polynomial function of the intensity of the edge map. The second component is the orientation score of the vector field. The new features are integrated into the GGVF equations in the smoothing and the stopping term. The algorithms, having been tested by a set of ground truth images, show that the proposed techniques lead to a better convergence and better segmentation accuracy with the reference to conventional GGVF snakes. The adaptive multi-feature snake does not require any hand-tuning. However, it is as efficient as the standard GGVF with the parameters selected by the ''brutal force approach''. Finally, proposed approach has been tested against recent modifications of GGVF, i.e. the Poisson gradient vector flow, the mixed noise vector flow and the convolution vector flow. The numerical tests employing 195 synthetic and 48 real ultrasound images show a tangible improvement in the accuracy of segmentation.
An Approach to Edge Detection in Images of Skin Lesions by Chan-Vese Model
2013
Nowadays there is a great interest in the application of computational systems for the analysis of skin lesions. These systems allow the dermatologist to prevent the development of malignant lesions. The development of the systems has occurred due to the increase of skin cancer cases. In the characterization of skin lesions it is necessary to segment the images accurately. Thus the features and edges information of the lesion can be extracted and used by a classifier or by a dermatologist for a better classification. When images are acquired in a non-systematic and non-controlled way there may be a segmentation problem. In this case the skin lesion of images can have different sizes and various type of noises, such as the hair. These factors can affect the detection of the lesion edges and complicate its characterization. One solution would be to apply a smoothing filter to reduce noise before the segmentation step. Segmentation techniques adapted to each type of image can be used to solve the problem of diversified images, such as images with different sizes lesions, reflexions and light intensities. In this paper is proposed a computational method to assist the dermatologists in the diagnosis of skin lesions by digital images. It was used the anisotropic diffusion technique for the preprocessing of the images in order to remove the noises. The Chan-Vese model was used to segment the lesions. The next step consists of the application of morphological filters to eliminate outside and inside noises from the object, that remained in the segmented images, and also to smooth their edges. This approach allowed to minimize noise problems and edge detection to different cases of skin lesions images, such as melanoma, melanocytic nevi and seborrheic keratosis. The segmentation achieved 94.36% of accuracy for the three types of skin lesions.
Segmentation of Melanoma Skin Lesions Using Anisotropic Diffusion and Adaptive Thresholding
Proceedings of the 2018 8th International Conference on Biomedical Engineering and Technology, 2018
Segmentation is the first and most important task in the diagnosis of skin cancer using computer-aided systems and due to complex structure of skin lesions, the automated process may lead to a completely different diagnosis. In this paper, a novel segmentation method of skin lesions is proposed which is both effective and simple to implement. Smoothing of skin lesions in original image plays a pivotal role to secure an accurate segmented image. Anisotropic Diffusion Filter (ADF) is used in the initial stage to smooth images with preserved edges. Adaptive thresholding is then applied to segment the skin lesion of the image by binarizing it. The morphological operations are applied for further enhancement and final segmented image is obtained by applying proposed boundary conditions in which objects are selected on basis of distance. The proposed technique is tested on over 300 images and averaged results are compared with existing methods like L-SRM, Otsu-R, Otsu-RGB and TDLS. The proposed method achieved an average accuracy of 96.6%. Visual results for selected images also depicted better performance of proposed method even in the presence of bad illumination and rough skin lesions in the image.
A wide spread of algorithms for automatic segmentation of dermoscopic images
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2013
Currently, there is a great interest in the development of computer-aided diagnosis (CAD) systems for dermoscopic images. The segmentation step is one of the most important ones, since its accuracy determines the eventual success or failure of a CAD system. In this paper, different kinds of algorithms for the automatic segmentation of skin lesions in dermoscopic images were implemented and evaluated, namely automatic thresholding, k-means, mean-shift, region growing, gradient vector flow (GVF), and watershed. The segmentation methods were evaluated with three distinct metrics, using as ground truth a database of 50 images manually segmented by an expert dermatologist. Among the implemented segmentation approaches, the GVF snake method achieved the best segmentation performance.
Segmentation of skin cancer images
Image and Vision Computing, 1999
An automatic method for segmentation of images of skin cancer and other pigmented lesions is presented. This method first reduces a color image into an intensity image and approximately segments the image by intensity thresholding. Then, it refines the segmentation using image edges. Double thresholding is used to focus on an image area where a lesion boundary potentially exists. Image edges are then used to localize the boundary in that area. A closed elastic curve is fitted to the initial boundary, and is locally shrunk or expanded to approximate edges in its neighborhood in the area of focus. Segmentation results from 20 randomly selected images show an average error that is about the same as that obtained by four experts manually segmenting the images. ᭧
Segmentation of brain tumors in MRI images using multi-scale gradient vector flow
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2011
The gradient vector flow (GVF) algorithm has been used extensively as an efficient method for medical image segmentation. This algorithm suffers from poor robustness against noise as well as lack of convergence in small scale details and concavities. As a cure to this problem, in this paper the idea of multi scale is applied to the traditional GVF algorithm for segmentation of brain tumors in MRI images. Using this idea, the active contour is evolved with respect to scaled edge maps in a multi scale manner. The edge detection performance of the modified GVF algorithm is further enhanced by applying a threshold-based edge detector to improve the edge map. The Bspline snake is selected for representation of the active contour, due to its ability to capture corners and its local control. The results showed an improvement of 30% in the accuracy of tumor segmentation against traditional GVF and 10 % as compared to Bspline GVF in the presence of noise, besides the repeatability of the alg...
Journal on Information Technology in Healthcare
Objective: To evaluate the application of active contours (snakes) in computer-assisted medical image analysis and to propose a parametric active snake model for segmenting pigmented skin lesions in dermatological images. The aim is to create a model that can identify the edges of a dermatological lesion and thus help differentiate benign lesions. i.e. with regular edges from malignant lesions (irregular edges).
A Novel Method for Skin Lesion Segmentation
International Journal of Information, Security and System Management, 2015
Skin cancer has been the most usual and illustrates 50% of all new cancers detected each year. If they detected at an early stage, treatment can become simple and economically. Accurate skin lesion segmentation is important in automated early skin cancer detection and diagnosis systems. The aim of this study is to provide an effective approach to detect the skin lesion border on a purposed image. A novel method based on image processing is proposed that combines the edge detection and the thresholding technique for skin lesions detection from skin region in an image. The distributions of edge and the proposed thresholding method provide a good discrimination of skin lesions. The evaluation of the proposed method is based on the comparison with the Otsu and Rosin segmentation as the most application methods. The performance of the designed system is evaluated with 30 test images, and the experimental results demonstrate the effectiveness of the proposed mole localization scheme.
Nonlinear smoothing of skin lesions images driven by derivative filters
2010
Image segmentation is an important step to suitable extraction of features of objects from images. However, the presence of noise interferes in segmentation quality; for example, by generating the detection of false edges (or false borders). To diminish the problems caused by the presence of noise in images, various smoothing techniques have been proposed to pre-process the original images. Those methods reduce the noise presented in input images, but they can also strongly affect the borders of the objects, leading to the loss of important details, such as the original roughness of the contours or the elimination of the borders of small objects. Among the existing smoothing techniques, one of the most promising is based on the use of anisotropic diffusion, which allows a selective smoothing that decreases the undesirable effects caused by noise presented in the input image and preserves the edges of the objects. However, the success of this smoothing method is strongly reliant on the number of iterations performed that depends on the input image. In this work, we propose the use of derivative filters for the definition of the appropriate number of iterations adopted by the smoothing method based on anisotropic diffusion, when it is applied for the removal of noise usually present in images of skin lesions. The experimental results demonstrate that the developed solution is promising, being able to determine the adequate number of iterations for smoothing the input images avoiding the excessive loss of details of the borders of the lesions presented in images.