Brain Ischemic Stroke Segmentation: A Survey

In this survey paper a comprehensive survey of segmentation techniques used for localizing brain stroke area (differentiating it with normal region of brain) has been presented. In recent times, automated systems are preferred over manual and semi automated systems. Segmentation or localization is not a challenge, challenge lies behind producing an automated, reproducible and clinically accepted method which works well in case of abnormalities too and can be used for further processing like computing volume, growth prediction and treatment etc. As in today's time there is no place for erroneous interpretations but large number of stroke cases overburden the radiologists and may result in wrong analysis. So, an efficient automated segmentation method is need of an hour and is preferred over conventional systems. The most relevant methods whether manual, semi automated or automated for segmentation are discussed. So, it sets up a picture where there is scope of improvement to reach a goal.