RATIONALE AND OBJECTIVES. Accurate registration of computed tomography (CI) images to the patient is crucial for computer-assisted surgery, Markers used in roentgen stereo-grammetric analysis (RSA) can be located in a CT scan using a novel approach and also can be located physically, Roentgen stereo-grammetric analysis data act as ''ground truth'' for the three-dimensional marker locations. METHODS. Two foam-bone phantoms were marked, The markers were scanned seven times with RSA, three times with axial CT, and contacted four times with a coordinate measuring arm, Root-mean-square (RMS) errors were derived for the registrations. RESULTS. Computed tomography and RSA data register to 0.15 mm RMS error, Computed tomography and arm data register to 0.25 mm, The markers are biocompatible, and the coordinate measuring arm is usable in an operating room. CONCLUSIONS. Typical in vitro registration errors are approximately 2 mm, The authors have developed a marker that provides superior registration, and a procedure that can be used for in vivo studies.
Use of a biocompatible fiducial marker in evaluating the accuracy of computed tomography image registration
CARAMELLA, DAVIDE;
1996-01-01
Abstract
RATIONALE AND OBJECTIVES. Accurate registration of computed tomography (CI) images to the patient is crucial for computer-assisted surgery, Markers used in roentgen stereo-grammetric analysis (RSA) can be located in a CT scan using a novel approach and also can be located physically, Roentgen stereo-grammetric analysis data act as ''ground truth'' for the three-dimensional marker locations. METHODS. Two foam-bone phantoms were marked, The markers were scanned seven times with RSA, three times with axial CT, and contacted four times with a coordinate measuring arm, Root-mean-square (RMS) errors were derived for the registrations. RESULTS. Computed tomography and RSA data register to 0.15 mm RMS error, Computed tomography and arm data register to 0.25 mm, The markers are biocompatible, and the coordinate measuring arm is usable in an operating room. CONCLUSIONS. Typical in vitro registration errors are approximately 2 mm, The authors have developed a marker that provides superior registration, and a procedure that can be used for in vivo studies.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
