Segmentation and volume calculation in acute stroke

Stroke is a major health problem and the most common imaging method in stroke therapy is magnetic resonance (MR). The diffusion weighted imaging (DWI) and perfusion-weighted imaging (PWI) are quickly becoming the main parts of acute stroke therapy. In this study, the MR images of acute stroke patients, which gives the damaged volume of the brain and the potential volume to be damaged, were calculated by a mathematical method and estimated by texture methods. In this study, the texture methods, which are Gray Level Co-Occurrence Matrix (GLCM), Gabor, and Laws Texture Energy Measure (LTEM) have been used for feature extraction. Extracted features were employed to estimate DWI and PWI volumes that give how much brain was damaged, and how much brain could be damaged, respectively. Obtained results were compared with the mathematical model and the literature.