Three-Dimensional Visualisation Of Ct-Data For Surgical Planning In Trachea Resection: Proof Of Concept

In this study we designed a new software to segment and visualise CT-data for surgical planning of trachea resection Resection planning requires exact anatomic knowledge in all three dimensions for each individual patient. Semiautomatic image analysis algorithms like live-wire, region-growing and watershed transform were integrated and set in a fixed segmentation order to simplify segmentation. The planning software contained multiple interaction possibilities to demonstrate the results in a 3-dimensional kind. Measurement tools to compute extents, values and distances were integrated. Segmentation was done by surgeons themselves. The segmentation results were compared with the intraoperative and histopathologic findings. The benefit was proofed by using the ICCAS assessment system "Level of Quality" (LOQ). This system allows an evaluation of the gain of information for surgical planning compared to standard two-dimensional representation methods. Eight patients with stenosis of the trachea were evaluated in this study, where three patients were undergoing resection of stenosis. The average time required for segmentation and visualisation was about 35 minutes, indeed professional image processors needed less time for post processing than clinical surgeons. Segmentation results were proved to correspond with intraoperative and histopathologic findings. The results of preoperative surgical evaluation of trachea stenosis based on two-dimensional CT-data or diagnostic tracheoscopy differed from histopathologic findings. In all cases stenosis was estimated too long. Measurements based on computer-guided three-dimensional reconstruction corresponded exactly to the histopathologic findings. The LOQ value of 70 showed a significant gain of information and thereby a benefit in comparison to standard two-dimensional representation methods of CT-data. In some cases surgical intervention was only made possible by preoperative planning on reconstructed CT-data. Interdisciplinary discussions were shortened and simplified by showing three-dimensional visualisation results to colleagues. Besides the received results were used for patients' information and students' teaching.