Automatic segmentation of nucleus accumbens

Segmentation of subcortical structures in the brain through MRI scans has become an increasingly important topic in contemporary medicine. An area of the brain whose possible segmentation has received particular attention is the Nucleus Accumbens (NA), which is believed to play a central role in the reward circuit. The use of a reliable automated segmentation method would represent an extremely helpful and efficient tool for medical teams, since many disorders (such as ADHD) cause variations in the volume of the NA. Consequently, the main objective of the thesis is the implementation of a robust algorithm for segmenting the Nucleus Accumbens structure. We apply state of the art segmentation methods to the NA, using three different approaches; firstly, the standard Atlas Segmentation method is used, showing generally poor results paired with long computational times and high complexity. In addition, Multi Atlas Segmentation and Adaptive Multi Atlas Segmentation methods are implemented. The results show improved accuracy and better performance than the single Atlas approach. Finally, we propose a prototype of GUI for the implementation of a semi-supervised segmentation approach, aimed at exploiting the knowledge of human experts to refine the segmentation results, and targeted to medical use.