Segmentation principles are used to construct personalised three-dimensional computer models of anatomical structures from imaging modalities such as CT, MRI and ultrasound. The segmentation is usually done by placing voxels of a certain organ in a separate mask. Depending on the image quality and the organ this proces is performed automatically or semi-automatically. The segmented model forms the basis for further functional analysis using either flow or structural simulations. Image segmentation can be done on virtually all regions that can be distinguished in a scans including bone, blood vessel and airways.
The segmentation and three-dimensional reconstruction generally provides a very informative visual aid to assess the organ’s condition. In addition to the visualisation also length and volume measurements could be performed on the models providing the opportunity to quantify certain regions and document the properties (eg tumour size) for follow up.
Case report: stent placement in kyphoscoliosis patient
To illustrate the clinical application of the segmentation principles a case study is considered in a 16-year old kyphoscoliosis patient. Please click here for more details.