Segmentation of MRI Images using Clustering Algorithms

Even though we live in a digital age, many hospitals and medical facilities still process MRI images manually. Such analysis must always be performed by trained medical staff who are experts in the field, but they may still miss something or misjudge a finding. This can lead to unnecessary examinations, costs and misdiagnosis. This is especially true with brain tumors. They often contain many abnormalities in size, shape and location. This makes it difficult to fully understand the nature of the tumor. A professional neurosurgeon is also needed to analyses the MRI results, which may not be available everywhere. Mainly in developing countries, there is often a lack of qualified specialists and the lack of knowledge about tumors makes diagnosis difficult and lengthy. This paper includes a description of what cluster analysis is, a description of selected clustering algorithms, tests on MRI images affected by a brain tumor, solution design using image segmentation, and evaluation of the tests using the selected evaluation metrics. Test results can be used in further research or applied to similar medical images to assist medical personnel in detecting objects in the human body.