Image Processing-based Analysis of Neural Network Segmented Brain Tumor MRI Images for Treatment Planning

Various fields are absolutely supported by technological advancements. Every week, millions of imaging procedures are performed all over the world. Correct segmentation and analysis based on height, width and area are very crucial in brain tumor patients. This paper presents the automatic calculation of maximum height and width, tumor area in pixels and pixel coordinates which give the exact location of tumor tissues. This work is an expansion of our previous work on automatic tumor segmentation using deep learning. Due to tumors’ varying nature and shape, it is difficult but very important to get exact tumor dimensions. Tumor dimensions are important components in the assessment of prognosis. Doctors can determine the severity of a patient’s brain tumor and provide appropriate treatment before it becomes harmful by taking regular measurements of the tumor size in the patient’s brain. So, automatic segmentation and analysis are very helpful for doctors to avoid prolonged processes and for patients in the time reduction.