PURPOSE: In this paper a new, easy-to-manufacture and easy-to-use ultrasound (US) probe calibration phantom for 3D freehand scanning is presented and evaluated, together with a new method for achieving an accurate and user-robust calibration using virtual plane. METHODS: The phantom allows the optically tracked US probe to perform two rotations and two translations while keeping the image of a tensioned wire in the image plane. This approach allows obtaining a sharp image of the wire independently from the probe pose. The virtual plane allows the calibration algorithm to converge minimizing the required number of US probe tracked poses. The US image and position data are synchronized via a CORBA interface, created within the Image Guided Surgery Toolkit (IGSTK) framework. The calibration algorithm and the calibration protocol were evaluated in a set of experiments carried out by different test-users. RESULTS: The calibration method proved to be accurate and precise: 3D point reconstruction accuracy resulted 0.2 mm as mean value, while the precision was 0.4 mm as standard deviation. CONCLUSIONS: The technique showed to be suitable for medical applications from morphological diagnosis to intraoperative surgical planning adaption.
PURPOSE: In this paper a new, easy-to-manufacture and easy-to-use ultrasound (US) probe calibration phantom for 3D freehand scanning is presented and evaluated, together with a new method for achieving an accurate and user-robust calibration using virtual plane. METHODS: The phantom allows the optically tracked US probe to perform two rotations and two translations while keeping the image of a tensioned wire in the image plane. This approach allows obtaining a sharp image of the wire independently from the probe pose. The virtual plane allows the calibration algorithm to converge minimizing the required number of US probe tracked poses. The US image and position data are synchronized via a CORBA interface, created within the Image Guided Surgery Toolkit (IGSTK) framework. The calibration algorithm and the calibration protocol were evaluated in a set of experiments carried out by different test-users. RESULTS: The calibration method proved to be accurate and precise: 3D point reconstruction accuracy resulted 0.2 mm as mean value, while the precision was 0.4 mm as standard deviation. CONCLUSIONS: The technique showed to be suitable for medical applications from morphological diagnosis to intraoperative surgical planning adaption.
Authors: E De Momi; G Ferrigno; G Bosoni; P Bassanini; P Blasi; G Casaceli; D Fuschillo; L Castana; M Cossu; G Lo Russo; F Cardinale Journal: Int J Comput Assist Radiol Surg Date: 2015-07-17 Impact factor: 2.924