Methods Of Extracting Sub-Cortical Structures From Human Brain Scans – A Survey

The DGM - Deep Grey Matter in the human brain is responsible for various activities such as memory, cognition, learning, etc. The problem in any of these functions indicates neurological diseases like Alzheimer's disease (AD), Cerebral palsy, etc. The image of DGM helps to study the functions of various organs. To extract the structure of DGM, many algorithms are applied. This paper presents automatic and semi-automatic segmentation techniques. The commonly used techniques are categorized into Atlas-based, Learning-based and intensity-based. The performance of such algorithms is compared with the existing tools like FreeSurfer, LesionTOADS and FIRST. The Quantitative metrics like Jaccard and Dice are also used to validate the performance of a technique