Academic
Publications
Border detection in dermoscopy images using hybrid thresholding on optimized color channels

Border detection in dermoscopy images using hybrid thresholding on optimized color channels,10.1016/j.compmedimag.2010.08.001,Computerized Medical Ima

Border detection in dermoscopy images using hybrid thresholding on optimized color channels   (Citations: 2)
BibTex | RIS | RefWorks Download
Automated border detection is one of the most important steps in dermoscopy image analysis. Although numerous border detection methods have been developed, few studies have focused on determining the optimal color channels for border detection in dermoscopy images. This paper proposes an automatic border detection method which determines the optimal color channels and performs hybrid thresholding to detect the lesion borders. The color optimization process is tested on a set of 30 dermoscopy images with four sets of dermatologist-drawn borders used as the ground truth. The hybrid border detection method is tested on a set of 85 dermoscopy images with two sets of ground truth using various metrics including accuracy, precision, sensitivity, specificity, and border error. The proposed method, which is comprised of two stages, is designed to increase specificity in the first stage and sensitivity in the second stage. It is shown to be highly competitive with three state-of-the-art border detection methods and potentially faster, since it mainly involves scalar processing as opposed to vector processing performed in the other methods. Furthermore, it is shown that our method is as good as, and in some cases more effective than a dermatology registrar.
Journal: Computerized Medical Imaging and Graphics - COMPUT MED IMAGING GRAPH , vol. 35, no. 2, pp. 105-115, 2011
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...Recent methods include histogram thresholding followed by region growing [4], JSEG algorithm based on color quantization and spatial segmentation [5], two-stage k-means++ clustering followed by region merging [6], global thresholding on optimized color channels followed by morphological operations [7], Hybrid thresholding [8], etc...
    • ...The performance of existing border-detection methods have been commonly evaluated by using either a single metric (e.g., border error [5], [6]) or a pair of metrics (e.g., sensitivity and specificity [8])...
    • ...In order to determine the weights, the three different approaches explained in Section IV are implemented using the results of the five recent border-detection methods: namely, Hybrid (W30B30) [8], DTEA [4], JSEG [5], KPP [6], and Global thresholding on XoYoR color channel [7], on the image set of 55 dermoscopy images...
    • ...2) Solving the System of Nonlinear Equations: To reconfirm the optimal weights α, β, σ, and δ ,w e have also solved the system of nonlinear equations defined by (23), subject to the constraints given in (16), using the borders drawn by the Hybrid border-detection method [8]...
    • ...3) Grid Search: A grid search, as explained in Section IV-C, is also applied on borders resulting from the Hybrid method [8] to find the weights yielding maximum OWPI...
    • ...Fig. 7 shows the five different borders resulting from applying the aforementioned five border-detection methods (DTEA [4], JSEG [5], KPP [6], global thresholding on XoYoR color channel [7], and Hybrid [8]) on a sample dermoscopy image...

    Rahil Garnaviet al. Optimized Weighted Performance Index for Objective Evaluation of Borde...

    • ...In our preliminary study [8] we proposed a hybrid thresholding method, which combines global and adaptive local thresholding and uses color optimization process to detect the lesion borders, and demonstrated its superiority over other state-of-the-art skin lesion segmentation algorithms...
    • ...Manual borders drawn by dermatologists tend to surround the borders produced by automated method [3,8]...
    • ...diagnosis. We tackle the problem by proposing a two-stage histogram thresholding method [8] which successively identifies the core-lesion and edge-lesion areas and here we highlight the success of the method in detecting the edge-lesion area...
    • ...The local threshold is calculated for every pixel over its surrounding window and the process is stopped when the initial pixel is revisited. Further details may be found in [8]...

    Rahil Garnaviet al. Global versus Hybrid Thresholding for Border Detection in Dermoscopy I...

Sort by: