[Computer graphics methods for 3-dimensional imaging of intrapulmonary space-occupying lesions of CT and MRI images]

To improve the pre-operative planning of interventional procedures in thorax surgery different methods of computer-graphical visualization of intrapulmonary lesions and adjacent anatomical structures based on CT and MR data were realized and compared to each other.

In 21 patients with intrapulmonary lesions the image data was segmented by interactive and automatic algorithms and different reconstruction techniques were applied (maximum intensity projection; color-encoded and transparent surfaces; volume rendering). Based on these three-dimensional reconstructions, different views from arbitrary perspectives (including simulated endoscopic images) were generated and animated film sequences of the 3D scene were displayed with 30 frames/second.

For all patients under review, a high-quality presentation of the relevant structures was obtained by use of the applied computer-graphical techniques. Even combinations of different visualization methods in one image can be generated with the software we developed.

The various methods for image segmentation allow a fast and comfortable processing even of large data sets. The calculated values of tumor surface and volume can be used for quantitative studies and therapy control. The planning of surgical and interventional procedures can be supported by the simultaneous visualization of the intrathoracic lesion and the surrounding structures.