Brain cancer detection in the MRI using the characteristics of the area

• Main Authors: Faraj, Abbas Mones, Saleh, Hayder Adnan, Abdul-Zahra, Dalael Saad, Alkhafaji, Mohammed Ayad
• Format: Conference Proceeding
• Language: English
• Subjects: Cell division; Feature extraction; Image processing; Magnetic resonance imaging; Medical imaging; Tumors
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0157402

When normal cell division and proliferation are disrupted by imbalance and disorder, abnormal and uncontrollable growth of some of these cells occurs, resulting in a mass or tumor. The current study intends to segment MRI scans to acquire a tumor-specific area of interest, which will aid clinicians in diagnosing images and offer an image ready for the CAD system's next steps of feature extraction and classification. The suggested methodology worked by conducting preset image processing, such as screening medical images and then transforming them to binary form. The image was segmented in the second stage by first determining a description of the components that make up the image and then determining their properties, with the area and toughness attributes being used to determine the tumor area, which represents the most solid area among all the image elements. Finally, morphological techniques wer e used to smooth out the edges and fill in the gaps in the study region. The procedure has been proved to reach a percentage of more than 98 percent on all photos in the database used.