Automatic Brain Stroke Detection using Histogram Based Classification Methods (original) (raw)

A computer aided stroke detection techniques are useful for diagnosing brain tumors or strokes. Human brain stroke is the rapid loss o f brain functions due to hemorrhage. Image classification and segmentation are used to explore different types of strokes. Since stroke detection using classification or segmentation is automated approaches, which will minimize the detection time. Histogram or Centroid based segmentation methods like K-Means, Mean-shift segmentation fail to detect optimal regions from high resolution images. In the high resolution images, segmentation main aim is to divide the image into a set of non-overlapping regions based on stroke features. Traditional approaches have been investigated to get an optimal solution for the stroke detection. Automated brain stroke detection approaches are difficult due to variations in size, type, shape and location of strokes. Histogram based stroke detection can be used to find the stroke in the left or right symmetrical regions. ...