Multimodal visualization of craniofacial tumors of large dimensions

Abstract

A special visualization approach for tumors in the cranio-maxillofacial region, e.g. skull base infiltrating tumors of large dimensions, was developed for diagnosis support as well as surgery and therapy planning.

Methods: For complicated craniofacial tumors, mostly, high resolution CT and specially adapted MRI protocols are available. For later rendering of the skeletal anatomy, the CT is registered on the MRI. Next, the tumor is roughly segmented from the MRI data with highly exaggerated safety margin. Thereafter, these voxels are isolated from the original MRI data. After detailed image processing, the CT data (skull, dark color) and the voxels of the tumor (highly lucent color) are simultaneously visualized with special transfer functions. The tumor is rendered with dynamically changing transparency. By the highly exaggerated safety margin, explicit segmentation of the tumor boundary is avoided which makes the overall approach less error prone. The dark color visualization of the skull based on CT data provides a good anatomical orientation whereby the tumor becomes clearly visible. The dynamically changing transparency allows varying impressions of the tumor. By the highly exaggerated margin, the vascular structures around the tumor are included to the rendering.

Results & Discussion: The approach has already been successfully applied for a series of patients with benign or malign craniofacial tumors or lesions of large dimensions where pre- and post-op radiological data were analyzed. By experienced cranio-maxillofacial surgeons and neurosurgeons, the provided images and movies were rated as very helpful for efficient diagnosis support and surgery planning. The approach was validated by direct comparison versus anatomical findings during surgery.

Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.