The design and development of a proto-type non-contact measurement system for craniofacial surgery

Abstract

The design and calibration procedures of a new craniofacial soft tissue imaging system are developed. The designed procedures of the system are based on the integration of stereophotogrammetry and three-dimensional laser scanning techniques to capture high resolution stereo images and three-dimensional surface model of craniofacial morphology, respectively. The stereophotogrammetric technique introduced a new multi-camera system consisting of eight high resolution digital cameras, high speed shutter synchronisation controller and a specially built patients chair with photogrammetric control frame. The novel technique dubbed as natural features technique is developed and examined to improve the accuracy of craniofacial landmarks measurements via captured stereo images. The laser scanning system consists of two high resolution Minolta Vivid 910 laser scanners. The important factor of using the scanner for craniofacial mapping is completely tested. Finally, the registration accuracy of the point clouds datasets captured from the laser scanners is improved by using photogrammetric targets that setup on the photogrammetric control frame. Both imaging systems are calibrated independently using special built camera calibration test field (to calibrate the camera system) and calibrated test objects (to calibrate the laser scanning system). The results show that the developed novel technique achieves an accuracy of 0.6mm at one standard deviation. The registration of point clouds datasets by using photogrammetric targets achieves an accuracy of less than 0.3mm at one standard deviation. The results also show that the use of laser scanning system for craniofacial mapping depends on several significant factors such as scan distance, camera focal length, laser beam intensity, scanning resolution, convergence angle and the number of overlapping scans. As a conclusion, the developed imaging system is a unique system for capturing craniofacial soft tissue datasets as the system captured various datasets (stereo images and three dimensional surface model) when the patient sits at the patient’s chair with the head placed at the middle of the photogrammetric control frame.