Segmentation of Melanocytic Lesion Images Using Gamma Correction with Clustering of Keypoint Descriptors

Abstract

The early detection of skin cancer, especially through the examination of lesions with malignant characteristics, has been reported to significantly decrease the potential fatalities. Segmentation of the regions that contain the actual lesions is one of the most widely used steps for achieving an automated diagnostic process of skin lesions. However, accurate segmentation of skin lesions has proven to be a challenging task in medical imaging because of the intrinsic factors such as the existence of undesirable artifacts and the complexity surrounding the seamless acquisition of lesion images. In this paper, we have introduced a novel algorithm based on gamma correction with clustering of keypoint descriptors for accurate segmentation of lesion areas in dermoscopy images. The algorithm was tested on dermoscopy images acquired from the publicly available dataset of Pedro Hispano hospital to achieve compelling equidistant sensitivity, specificity, and accuracy scores of 87.29%, 99.54%, and 96.02%, respectively. Moreover, the validation of the algorithm on a subset of heavily noised skin lesion images collected from the public dataset of International Skin Imaging Collaboration has yielded the equidistant sensitivity, specificity, and accuracy scores of 80.59%, 100.00%, and 94.98%, respectively. The performance results are propitious when compared to those obtained with existing modern algorithms using the same standard benchmark datasets and performance evaluation indices.

Preprocessing Effects on Performance of Skin Lesion Saliency Segmentation

“…However, there is no solitary preprocessing method that can generally be applied to resolve the different types of undesirable artifacts inherent in dermoscopic images. Consequently, several studies have incorporated multiple stages of preprocessing to tackle the heterogeneous properties of dermoscopic images to improve segmentation results [ 2 , 8 , 9 , 17 , 23 , 26 , 32 , 33 , 37 , 39 , 40 , 41 , 47 ]. The dependency on manifold stages of preprocessing confines the generalizability of the existing segmentation methods and increases their computational complexities.…”

“…The CLAHE is widely recognized as the best method among the prevailing enhancement methods for preprocessing of medical images [ 40 ]. In addition, literature has shown evidence of preprocessing stages based on histogram [ 33 ], mean subtraction [ 31 ], deep learning [ 34 ], multiscale decomposition [ 21 ], adaptive gamma correction [ 23 ], Z-score transformation [ 52 ], and Frangi Vesselness filter [ 41 ]. The artifact removal and image enhancement algorithms are generally executed before the actual segmentation and postprocessing methods are applied to suppress the leftover noise.…”

“…Due to their elegance and efficiency, a wide gamut of unsupervised segmentation methods has been developed to date and applied for the analysis of skin lesions. These methods include thresholding [ 17 , 19 ], clustering [ 20 , 21 , 22 , 23 ], region merging [ 8 , 10 , 39 ], and saliency [ 19 , 24 , 25 , 27 , 28 , 29 , 38 ]. More recently, a privacy-preserving segmentation method based on the partially homomorphic permutation ordered binary (POB) number system was proposed for the segmentation of encrypted skin lesions [ 53 ].…”

“…Myriads of image segmentation methods have been proposed in the literature for skin lesions. They include thresholding [ 17 , 19 ], clustering [ 20 , 21 , 22 , 23 ], statistical region merging [ 8 , 10 ], saliency [ 19 , 24 , 25 , 26 , 27 , 28 , 29 ], and deep learning [ 2 , 30 , 31 , 32 , 33 , 34 , 35 ]. Although multifarious image segmentation methods exist in the literature, accurate segmentation of skin lesions is still a challenging open problem because of the heterogeneous properties of dermoscopic images.…”

“…A generalized solitary method to handle the various types of undesirable artifacts is still an unresolved problem in the disciplines of image processing and computer vision. Hence, myriads of studies have relied heavily on different stages of preprocessing to address the countless types of undesirable artifacts and uneven color contrast of dermoscopic images [ 2 , 6 , 9 , 17 , 20 , 23 , 26 , 37 , 39 , 40 , 41 ]. The demand for the inclusion of multiple preprocessing phases can adversely affect the performance of a segmentation method while limiting its practical applications in real clinical settings.…”