Fig. 1.

Region growth algorithm based on sagittal T2WI images for normalized axial CE-MRI segmentation. The lesion was found in the (A) sagittal position and was manually divided. Axial CE scan was performed on the (B) focal vertebra. The focal region outlined in the sagittal position (C) was transformed into the focal region in the axial position, and the red line region was used as the initial region of the region growth algorithm for segmentation. As shown in (D), in order to cover all lesions, the region growth algorithm expanded the initial region by five times and divided the image into several different partitions represented by different colors. (E) is reserved for and shows the partition that overlaps the red line area. (F) is a 3D area where the remaining partitions in all layers are fused. The pixel with the highest degree of enhancement in the (G) 3D region is the result shown by the seed region growth algorithm of the region, and the red region is the segmented tumor region. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)