Automatic Detection of Brain Tumors Using Genetic Algorithms with Multiple Stages in Magnetic Resonance Images

Authors

Keywords: MRI brain tumour, GLCM, SURF, Genetic optimization, advanced machine learning.

Abstract

Biomedicine is still working to solve the problem of detecting brain tumours, one of the biggest problems in the profession today. With improved technology or instrument, early diagnosis of brain cancers is feasible. Classifying brain tumour kinds using patent brain pictures enables automation in automated procedures. Furthermore, the suggested new method is utilised to tell the difference between brain tumours and other brain diseases. To split the tumour and other brain areas, the input picture is first pre-processed. After this, the pictures are divided into different colours and levels, and then they are run through the Gray Level Co-Occurrence and SURF extraction methods to uncover the important details in the photographs. Using genetic optimization, the retrieved characteristics are made smaller. For training and testing tumour classification, the cut-down characteristics are used using an advanced learning technique. The technique's accuracy, error, sensitivity, and specificity are all evaluated alongside the current method. The method has a 90%+ accuracy rate, with less than 2% inaccuracy for all kinds of cancers. Finally, the specificity and sensitivity of every kind are above 90% and 50% correspondingly. Using a genetic algorithm to support the approach is more efficient, since the method it uses has both higher accuracy and specificity than the other techniques.

Downloads

Published
02.10.2023
Issue
Section
Articles

How to Cite

Annam, K., G, S. K., P, A. B., & Domala, N. (2023). Automatic Detection of Brain Tumors Using Genetic Algorithms with Multiple Stages in Magnetic Resonance Images. Journal of Automation, Mobile Robotics and Intelligent Systems, 16(4), 36-43. https://doi.org/10.14313/JAMRIS/4-2022/31