Title: Bio-inspired approaches for classification of benign and malignant tumour of the skin

Authors: Aman Gautam; Usha Chouhan

Addresses: Department of Bioinformatics, Maulana Azad National Institute of Technology, Bhopal, Bhopal, 462003, India ' Department of Mathematics, Bioinformatics and Computer Applications, Maulana Azad National Institute of Technology, Bhopal, Bhopal, 462003, India

Abstract: The malignant tumour is the deadliest skin disease. The reasons for malignant growth are carcinogenic cells that can spread to the different parts of the skin. There are three primary types of skin diseases: Basal Carcinoma Cancer, Squamous Carcinoma Cancer, and Melanoma. The initial two are known as non-melanoma skin malignant growth or NMSC. 90% of melanoma has occurred due to sunlight. This exposure has increased due to thin ozone layer and tanning bed is another cause of skin cancer. The motivation behind this paper is to classify the Benign and Malignant tumour dependent on dermoscopic images from ISIC datasets and utilising five machine learning methods (MLP, SVM, RF, KNN, and LR). Correctly classified instances were found in 87%, 83%, 86%, 73% and 70% for MLP, SVM, RF, KNN, and LR respectively. The accuracy achieved higher than the current methodologies.

Keywords: tumour images; medical decision support system; machine learning technique; tumour types; classification methods; performance measures; supervised learning; features extraction methods.

DOI: 10.1504/IJBRA.2021.120197

International Journal of Bioinformatics Research and Applications, 2021 Vol.17 No.5, pp.424 - 433

Received: 22 Jan 2019
Accepted: 09 Aug 2019

Published online: 11 Jan 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article