To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Computer-aided Decision Support System for Classification of Pigmented Skin Lesions |
|
Author |
Qaisar Abbas |
| Citation |
Vol. 16 No. 4 pp. 9-15 |
|
Abstract |
Computer-aided decision support systems (CAD-SS) are widely used for providing clinical decision to medical experts. A few CAD-SS systems are developed to diagnosis all categories of pigmented skin lesions (PSLs) using dermoscopy images. These systems are limited to specific categories of PSLs, computationally slow and defined non-effective visual features. In this paper, an improved CAD-SS system is proposed by integrating of effective color and texture features for recognizing of 12 categories of skin lesions. Compared to state-of-the-art systems, the optimal features are extracted from each of PSLs images in a perceptual-oriented color space (CIEL*a*b*). These visual features are then classified by using Learn++ architecture. The CAD-SS system is tested on a total of 360 lesions and achieved an average value of area under the receiver operating curve (AUC) of 0.97. Experimental results indicate that the CAD-SS system yields higher accuracy for recognition of pigmented skin lesions and the CAD-SS system can helpful for less experience dermatologists to take their clinical decisions. |
|
Keywords |
Skin Cancer, Computer-aided diagnosis system, Incremental learning, Perceptual color space, Hill-climbing, Color and texture features. |
|
URL |
