A CT Image Analysis Method for Evaluating Hematoma Enlargement in Acute Hypertensive Cerebral Hemorrhage

Hemorrhage(1) enlargement occurs in about one-third of patients with acute hypertensive cerebral hemorrhage (AHCH) and leads to high rates of morbidity and mortality. An effective CT imaging analyses method for AHCH enlargement is very important for medical diagnosis. The proposed system in this paper has been implemented by: CT scans input, registration, filter, segmentation and calculation. Registration based on local mutual information and segmentation based on Markov random field model are the key algorithms for effectively detecting the enlargement of AHCH. The system was developed in QT 5.0 using C++. Using a non-standard CT's database, the sensitivity, specificity and accuracy of the system are 95%, 90% and 92%. Compared with manual segmentation, the statistical results of volume change showed that there was no significant difference (p=0.48, p>0.05). Our proposed method also can improve the efficiency and reduce the consuming-labor of the physician.