3-D calibration method for freehand ultrasound image with high precision based on string-beads phantom

As extremely important techniques of MRCAS, Z-string-beads phantom and optimization algorithm for manual imaging coordinates are investigated to improve the precisions of 3-D ultrasound (US) calibration which defines the spatial relationship between the US image plane and an external tracking device attached to the US probe. The phantom consists of accurate position beads and two layers of Z-string planar arrays. The beads can rapidly guide the user to align the probe accurately for imaging, so their coordinates are directly used in the algorithm for homologous points matching without any transformations; the optimized algorithm incorporates the imaging points coordinates (mass centre coordinates of each point) in multiple US images that corrects the manual marking coordinates of the imaging points. The calibration parameters are calculated by matching of homologous points in the image and the tracking device and the 2-D US image can be converted into 3-D US coordinates. The results show that this method can simplify the calibration algorithm, reduce the influences of the US beam thickness on the precision and improve the calibration precision and reconstruction accuracy of the 3-D US system. The calibration experiments are demonstrated with a magnetic locator, but the method could be suitable for other tracking devices.