Magnetic Resonance Imaging (MRI) is today one of the most important medical imaging technologies in contemporary health care. It produces sharp images of the body interiors, especially the soft tissue, i.e. brain, spinal cord and organs. However, detection of some important aspects, including tumours, lesions and organ boundaries in MRI images, remains a problem. The reasons which lead to such difficulties are the issues of image noise, low contrast, brightness, and the complicated body structures. As stated by the current research paper, the spatial domain algorithms could be capable of locating characteristics of the MRI images. The primary methods which are going to be introduced are edge detection, region-based segmentation, and morphological operations. The working mechanism of each method, the advantages/disadvantages of each method and how that method contributes towards feature detection will be taken into consideration in the paper. The practical sense of these algorithms is proven in the section on clinical uses. Last but not least, the paper renders the future potential and contemporary limitations of enhancing the methods of the spatial domain along the coalescence of artificial intelligence and mainstream image processing.