Detection Of Melanoma Tumor in Dermoscopic Images Using Image Segmentation and Machine Learning

DOI

10.2991/978-94-6463-602-4_21How to use a DOI?

Keywords

Melanoma Detection; Hair removal; supervised learning; Image segmentation; Image enhancement

Abstract

Skin cancer is considered one of the most dangerous and common types of cancer. Melanoma is a type of skin cancer caused by the abnormal growth of melanocytes. If the spread of melanocytes is limited to the subcutaneous layer of the skin, then it would be benign cancer. It can become deadly if the melanocytes introvert into the blood supply. The early detection of melanoma is crucial for effective treatment. Recently, implementing a machine learning algorithm on dermoscopic images has become a great tool for early detection. In this paper, we introduce an automated method for classifying melanoma from dermoscopy images into benign and malignant cancer using supervised learning methods. International Skin Imaging Collaboration (ISIC) and the Society for Imaging Informatics in Medicine (SIIM) image dataset were used for study. The images were normalized and then a hair removal algorithm was applied to eliminate occlusion caused by the patient’s hair. Image segmentation using kmean clustering to extract the skin lesion /tumor from the background, followed by feature extraction based on the morphological structure, intensity, and texture features from the segmented data. After feature extraction, the data were used to train the classifiers and tested. Comparative analysis was also performed between different classifiers such as Support Vector Machine (SVM), Random Forest (RF), Decision Tree (DT), Linear Regression (LR), and K-nearest neighbor (KNN). To test the performance of the classifiers, we calculated the precision, accuracy, specificity, F1-score, confusion matrix, and ROC curves implemented. From the study, it was found that LDA and logistic regression were able to achieve an accuracy of approx. 93% and 92%. The proposed method can detect the presence of melanoma from dermoscopic images with good accuracy, which is comparable to deep learning-based and neural network methods.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Muhammad Aamir
AU  - Muhammad Usman
AU  - Muhammad Farhan Yousuf
AU  - Muhammad Abdullah
PY  - 2024
DA  - 2024/12/24
TI  - Detection Of Melanoma Tumor in Dermoscopic Images Using Image Segmentation and Machine Learning
BT  - Proceedings of the 4th International Conference on Key Enabling Technologies (KEYTECH 2024)
PB  - Atlantis Press
SP  - 145
EP  - 149
SN  - 2589-4943
UR  - https://doi.org/10.2991/978-94-6463-602-4_21
DO  - 10.2991/978-94-6463-602-4_21
ID  - Aamir2024
ER  -