Intra cerebral haemorrhage (ICH) segmentation using CT scan image texture analysis.

In recent decades, the automatic segmentation of medical images has become necessary for many researchers. Pre-processing and computer graphic techniques are a big part of processing medical data and changing the format to make it understandable. The problem of this study is how to segment each slice of Computed Tomography (CT) image based on the pixel features of suspect stroke images, especially in intracerebral type strokes. The purpose of this study was to segment based on the feature pixel features based on the statistical data of suspect Intra Cerebral Haemorrhage (ICH) features so that normal and abnormal tissue characteristics could be seen due to bleeding. This research method develops segmentation based on statistical features. CT images require pre-processing (pre-processing) to image enhancement to meet the CT image requirements with isolation, region growing, and logical operators (OR and AND). Morphological methods (opening and closing) with logical operators produce good results. These results can help that the most uncomplicated ICH stroke segmentation process, the thresholding process, is used to extract the ICH stroke regions from the brain CT images. A median filter is applied to remove noise from the image. This study used statistical features to calculate the first-order histogram to detect the ICH stroke area. [ABSTRACT FROM AUTHOR]