| This thesis presents a novel computer-assisted orthopaedic surgery (CAOS) method that employs 2D ultrasound as the imaging modality. Medical ultrasound is non-ionizing and real-time, and our proposed method does not require any invasive mounting procedures. Preoperatively, a set of 2D ultrasound images is acquired with the corresponding positional information of the ultrasound probe provided by a tracking system. Using calibration parameters, the position of every pixel in the images is transformed into the world coordinate frame to construct a 3D volume of the targeted anatomy for surgical planning. Intra-operatively, the surgeon takes live ultrasound images from the patient with the position of the ultrasound probe tracked in real time. A mutual-information-based registration algorithm is then used to find the closest match to the live image in the preoperative images. Because the position of the preoperative image inside the ultrasound volume is known, we are able to register the preoperative volume to the patient's anatomy.; Experiments have shown that the registration technique has sub-millimeter accuracy in localizing the best match between the intra-operative and pre-operative images, demonstrating great potential for orthopaedic applications. This method has some significant advantages over previously reported ultrasound-guided LAOS techniques: it requires no segmentation, and employs only a few ultrasound images to accurately and robustly localize the patient. Preliminary laboratory results on both a physical model of a radius bone and human subjects are presented. (Abstract shortened by UMI.)... |
