Computational Investigation of Stroke Lesion Segmentation from Flair/DW Modality MRI

Brain stroke (BS) is one of the dangerous diseases in human community, which causes the temporary or permanent disability and also death. Hence, in clinical level, the screening procedure BS is usually done using with brain images. MRI is a widely considered imaging procedure to record the internal organs of human body for clinical analysis purpose. This paper presents a soft computing procedure to examine the BS from the MRI of Diffused Weigh (DW) and Flair. This paper implements the combination of thresholding and Localized Active Contour (LAC) segmentation procedure to extract the stroke section. The evaluation of the stroke section if done using Haralick texture features. This paper also provides a performance evaluation of DW and Flair images. The outcome of this research confirms that, flair modality supports better result compared with the DW.