Automation in healthcare is necessary to improve efficiency and productivity. Researchers deal with many time-consuming tasks which directly affect their daily work. This project arises from the need of the fat-free muscle area (FFMA) evaluation in the erector spinae muscle as a prognostic marker in cirrhotic patients. In the same way, a study of this area is a marker of sarcopenia, or muscle loss. The aim of the work is to develop a medical application able to facilitate the erector spinae muscle segmentation to quantify FFMA. A set of magnetic resonance images (MRI) have been the given data to study the issue and make the tests during software development. Due to the dissimilarity between the MRIs, it was decided to provide the user with a two-alternative method: automatic and semi-automatic segmentation. To achieve the tool, it has been done a thorough investigation based on the physiological issue, deep research on the possible image processing techniques and segmentation methods to implement the most suitable ones to the final code. The programming language is MATLAB and it is the same software used to design the Graphical User Interface (GUI). The final application has been successfully performed, even though it has some inaccuracies. The automatic segmentation misses in the images which the intensity value of the erector spinae muscle pixels is closed to the ones of surrounding anatomic parts. Nevertheless, semi-automatic segmentation visibly improves the total muscle area localization. The FFMA results deviate from the current method in clinical investigation, it is needed more data to determine the fiability of the new proposed method.