Computerized grading of brain tumors supplemented by artificial intelligence

• Published in: Soft computing (Berlin, Germany) Vol. 24; no. 10; pp. 7827 - 7833
• Main Authors: Aruna, S. K., Sindhanaiselvan, K., Madhusudhanan, B.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.05.2020; Springer Nature B.V
• Subjects: Accuracy; Algorithms; Artificial Intelligence; Brain; Brain cancer; Computational Intelligence; Control; Datasets; Diagnosis; Engineering; Gabor filters; Image segmentation; Literature reviews; Magnetic resonance imaging; Mathematical Logic and Foundations; Mechatronics; Medical imaging; Methodologies and Application; Robotics; Search algorithms; Support vector machines; Swarm intelligence; Tumors; Wavelet transforms; Brain tumor; Artificial bee colony; Support vector machine; Magnetic resonance imaging; Medical image segmentation
• ISSN: 1432-7643; 1433-7479
• DOI: 10.1007/s00500-019-04403-7

For effective diagnosis of health conditions, there is a need to process medical images to obtain meaningful information. The diagnosis of brain tumors begins with magnetic resonance imaging (or MRI) scan. This is followed by segmentation of the medical images so obtained which can prove cumbersome if it were to be performed manually. Determining the best approach to do segmentation remains challenge among multiple computerized approaches. This paper combines both the identification and classification of tumors from the MRI results and is backed by a cloud-based framework to provision the same. The phase of extraction of features includes the utilization of a Hadoop framework and Gabor filter along with variations in terms of orientation and scale. Artificial bee colony algorithm and support vector machine classifier have been used to designate the degree of optimal features and categorize the same. The grading of brain tumors from MRI images can be fulfilled by the aforementioned approach. The said approach is believed to deliver promising results in terms of accuracy, which has also been verified experimentally.