Selection and Evaluation of Color/Depth Camera for Imaging Surgical Stoma.

We aim to develop a system to evaluate pouch attachment to and removal from stoma (artificial anus). For preliminary evaluation, we devoted this study to verify if a color/depth camera can accurately capture the stoma shape. Specifically, a 3D scanner was used to take precise images of three stoma models made of human tissue-mimicking gel and with diameters of 3, 6, and 9 cm to obtain the corresponding point clouds. Then, three stoma models were imaged using a color/depth camera to obtain the point clouds. We used the “CloudCompare” software and its implementation of the iterative closest point algorithm to compare the point clouds. Furthermore, the point clouds in the 3D Euclidean space were represented in a polar coordinate system with origin at the center of gravity of the point cloud to obtain the corresponding histograms. Finally, each point cloud and histogram captured by depth camera and 3D scanner were compared. The evaluation results showed that a larger stoma model provides a smaller difference between the point clouds. Nevertheless, the histograms from the three models were similar. Therefore, histogram comparison may be effective for shape recognition of real stoma from point clouds toward the extraction of information for a system to guide pouch management.