Research Information System
Tıp Teknolojileri Ulusal Kongresi,, Nevşehir, Turkey, 25 - 27 September 2014, pp.85-88, (Full Text)
Malignant Melanoma (MM) is the third most frequently
occurring human skin cancer. It is the deadliest form of all
skin cancers and arises from cancerous growth in pigmented
skin lesion. Research in the field melanoma is very important
because it has a good chance to be cure if treated in the early
stages. Nowadays melanoma computer-aided diagnosis
research is very popular with based on computer algorithms.
A software solution designed to analyze photographs of the
lesions where the dermatologists will capture the image of a
melanoma using a digital dermathoscopy will process the
image and provide an output diagnosis in an automated
manner.
ABCD (A: asymmetry, B: border, C: colour, D: differential
structures) score calculation, lesion image need some preprocessing. Because of the lesion image as like thick hair,
illumination, thin blood vessels, skin lines were affecting
computer based algorithm. In this study, we have proposed a
new method for pre-processing to remove hairs and
illumination effects from dermathoscopic image to calculate
the ABCD score successfully by software. We compare the
performance of a classical method to find the border of the
lesion by using before and after filtering. Our experimental
results indicate that proposed method can eliminate the effect
of hairs and illumination