Title :
P5C-4 Image Guidance Using Camera and Ultrasound Images
Author :
Khosravi, Sara ; Rohling, Robert ; Lawrence, Peter
Author_Institution :
Univ. of British Columbia, Vancouver
Abstract :
Previous work of estimating needle trajectories using un-calibrated camera-based tracking ("touchless" needle guide) is extended here to real ultrasound imaging. Using 4D ultrasound, this technique can be used as a needle guide during percutaneous procedures. A pair of miniature cameras captures images and the ultrasound system captures a volume as the needle moves through the workspace. The 2D position and orientation of the needle is calculated from the camera images and the 3D position of the needle is estimated from the ultrasound volume. A database is created from these values and a non-parametric learning algorithm determines the best model to the data. Then, in practice, the model can be used to estimate the projected needle trajectory within the ultrasound volume with only camera images of the needle above the skin surface. This study achieved an average error of 0.94 mm in position and 3.93deg in orientation.
Keywords :
bioMEMS; biomedical ultrasonics; image sensors; learning (artificial intelligence); medical computing; microsensors; needles; nonparametric statistics; position measurement; skin; stereo image processing; 2D needle position calculation; 4D ultrasound technique; image guidance; miniature cameras; needle guidance; needle orientation; nonparametric learning algorithm; percutaneous procedures; projected needle trajectory estimation; skin surface; ultrasound images; ultrasound volume; Anatomy; Biomedical optical imaging; Calibration; Cameras; Image databases; Image reconstruction; Needles; Optical imaging; Trajectory; Ultrasonic imaging;
Conference_Titel :
Ultrasonics Symposium, 2007. IEEE
Conference_Location :
New York, NY
Print_ISBN :
978-1-4244-1384-3
Electronic_ISBN :
1051-0117
DOI :
10.1109/ULTSYM.2007.566