Abstract: The intracranial hematoma, especially acute intracranial hematoma, is one of intracranial injuries which do harm to human life and health, so accurate segmentation of intracranial hematoma area has significant clinical value. Segmentation of hematoma regional medical images is a key technology to realize intracranial hematoma 3D reconstruction and volume calculation, and the problem with segmentation is how to improve the accuracy. It established a matterelement model for medical images of intracranial hematoma, and proposed a research method which combined extension detecting technology of focusing on matter with the fuzzy C-means (FCM) clustering algorithm, which prevents the FCM clustering algorithm from falling into local optimization in segmentation of intracranial hematoma medical images.  Therefore, this method effectively improves the accuracy of segmentation.

Key words: extension detecting technology; matter focusing; fuzzy Cmeans clustering; intracranial hematoma; segmentation method