Brain tumor diagnosis based on Zernike moments and support vector machine optimized by chaotic arithmetic optimization algorithm. (April 2023)

• Record Type: Journal Article
• Authors: Zheng, Nan; Zhang, Guoying; Zhang, Yang; Sheykhahmad, Fatima Rashid
• Is Part Of: Biomedical signal processing and control. Volume 82(2023)
• Publication Date: 2023-04
• Subjects: Computer-aided diagnosis -- Brain tumor -- Zernike moments -- Support vector machine -- Chaotic arithmetic optimization algorithm; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104543
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: An automated method for brain tumor diagnosis based on MR image. For preprocessing the MRI, noise reduction and skull removing are established. Zernike moments are used for features extraction. An optimized version of support vector machine is used for classification. A new version of Arithmetic Optimization Algorithm is used to optimization. Abstract: The abnormal growth of cells causes brain tumor in the human brain. These tumors can cause to cancer, which is one of the death reasons worldly. Early diagnosis of the tumor and estimation of its progression based on the MRI image will assist physicians save lives. Here, an automated method is presented in MRI images to find and diagnosis of the tumors. The proposed method includes four major steps including preprocessing, image segmentation, feature extraction, and classification. The first step includes two phases: one phase is for reduction of noise and the second is to remove the skull parts that can be decrease the diagnosis accuracy. The second step is to segment the region of interest based on an optimized Kapur thresholding followed by mathematical morphology. Then, Zernike moments are employed to drive the main image characteristics, and finally, an optimized classification methodology based on support vector machine is used for final diagnosis. The optimization of image segmentation and classification is established based on an improved version of Arithmetic Optimization Algorithm to provide a system with high efficiency. For validation of the suggested method, it is performed to Figshare dataset and the results are compared with six new techniques from the literature to show the system effectiveness.