A neural network based approach on melanoma skin cancer detection | 31228
Journal of Proteomics & Bioinformatics

Journal of Proteomics & Bioinformatics
Open Access

ISSN: 0974-276X

+44 1223 790975

A neural network based approach on melanoma skin cancer detection

6th International Conference & Expo on Proteomics

March 29-31, 2016 Atlanta, USA

Aso Bakr Mohammed

Sulaimanyah University, Iraq

Scientific Tracks Abstracts: J Proteomics Bioinform

Abstract :

High rate of melanoma skin cancer brought up the necessity of finding more innovative methods for diagnosis and cure. Approximately each year 9000 dies due to melanoma skin cancer. Unfortunately the rate of people diagnosed with melanoma is increasing at a rate of 3% and if it is not diagnosed in the first stage of cancer, there will be only 5% chance of survival. Also the cost of treatment will increase dramatically based on the stage of cancer. Due to importance of melanoma skin cancer, it is crucial to detect any sign of disease in early stages. Old dermotoscope devices can help physicians to have a better judgment about any abnormal moles that appears on skin, but not sufficient enough. New methods of melanoma skin cancer detection aim to investigate more features from the mole until the final decision is made. Here we present various image processing techniques to detect the A, B, C & D features of skin cancer which are used in an advanced neural network to come up with the task of cancer detection, the accuracy of results were promising, proving our method has advanced over other algorithm in both part of detecting features from images along with neural network decision performance.

Biography :

Aso Bakr Mohammed has completed his Bachelor’s degree from Sulaimanyah University, College of Science in Biology Department. He currently works as Medical Laboratory Technologist at Shar Hospital and he has published two papers.