Soft tissue scanning for patient registration in image-guided surgery

Prior to an image-guided surgical intervention, a correlation between the patient's data set and the surgical site is required. This study introduces a markerless registration method for craniomaxillofacial surgery that is based on a high-resolution laser scan of the patient's skin surface. The Surgical Segment Navigator SSN++ rejects contaminated surface measurements in a way similar to the bluescreen technique. Acquisition of the spatial position and the corresponding surface color of each laser-scanned point facilitates this bluescreen method, removing points with a defined surface color, e.g., blue or green points. The accuracy of the laser-scan-based registration was measured via additional intraoral titanium-markers. These markers served only to check the accuracy of the markerless registration process. In twelve patients, the stability and accuracy of the data set alignment was evaluated for high-(300,000 surface points), medium-, and low-resolution (down to 3,750 surface points) laser scanning. The accuracy of the registration technique was best for high-resolution laser scanning (mean deviation 1.1 mm; maximum deviation 1.8 mm). Low-resolution laser scans revealed inaccuracies up to 6 mm.