Spot Brain Tumors Using MRI Scans

• Published in: 2023 9th International Conference on Advanced Computing and Communication Systems (ICACCS) Vol. 1; pp. 2403 - 2407
• Main Authors: K, Elavarasi, R, Balaji Ragavendra Raj
• Format: Conference Proceeding
• Language: English
• Published: IEEE 17.03.2023
• Subjects: Classification algorithms; Image segmentation; Key Terms Include Segmentation; Learning Vector Quantization; Magnetic resonance imaging; Magnetic Resonance Imaging of The Brain; Robustness; Self-Organizing Feature Map; Stationary Wavelet Transform; Training; Vector quantization; Wavelet transforms
• ISSN: 2575-7288
• DOI: 10.1109/ICACCS57279.2023.10112730

The vast majority of cells in the human body develop and divide in a well-regulated manner to generate new cells, which are necessary for the body to maintain its healthy and functional state. However, cells may undergo uncontrolled and rapid proliferation when they lose the ability to self-regulate their growth, leading to the formation of abnormal growths or tumors in the body's tissues. Brain tumors, for example, arise due to the excessive proliferation of cells in the brain. This paper intends to evaluate the present status of segmentation techniques utilized in biomedical image processing and suggest means to enhance them. The primary objective is to analyze the current status of semiautomatic and automatic methods for segmenting anatomical medical images, including their pros and cons. Through this evaluation, we hope to provide insight into techniques that can improve the accuracy and efficiency of segmentation methods. To achieve this goal, we employ database testing and training to detect brain cancers and classify them into different stages. Segmentation in the context of testing often involves the use of spatial fuzzy c-means (FCM), a clustering algorithm that partitions an image into different regions based on both intensity and spatial information. Our study will explore various ways to enhance the performance of segmentation methods, including the use of prior knowledge, combining multiple methods, and incorporating expert feedback. By improving the accuracy and efficiency of segmentation techniques, we can potentially provide better diagnosis, treatment, and monitoring of brain tumors and other medical conditions.