HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Scheiber, C. Articles by Grucker, D. Search for Related Content PubMed PubMed Citation Articles by Scheiber, C. Articles by Grucker, D.

Nuclear Medicine Image Registration by Spatially Noncoherent Interferometry

Institut de Physique Biologique, Faculté de Midecine, Université Louis Pasteur, Strasbourg
CDO-Optimet, Paris, France

Correspondence: For correspondence or reprints contact: Christian Scheiber, MD, Institut de Physique Biologique, Faculté de Médecine, UPRES-A 7004 CNRS, 4 rue Kirschleger, Strasbourg Cedex, F-67085 France.

ABSTRACT

This article introduces a technique for obtaining high-resolution body contour data in the same coordinate frame as that of a rotating camera, using a miniature range finder, the conoscope, mounted on the camera gantry. One potential application of the technique is accurate coregistration in longitudinal brain SPECT studies, using the face of the patient (or "mask"), instead of SPECT slices, to coregister subsequent acquisitions involving the brain. Methods: Conoscopic holography is an interferometry technique that relies on spatially incoherent light interference in birefringent crystals. In this study, the conoscope was used to measure the absolute distance (Z) between a light source reflected from the skin and its observation plane. This light was emitted by a 0.2-mW laser diode. A scanning system was used to image the face during SPECT acquisition. The system consisted of a motor-driven mirror (Y axis) and the -camera gantry (1 profile was obtained for each rotation step, X axis). The system was calibrated to place the conoscopic measurements and SPECT slices in the same coordinate frame. Results: Through a simple and robust calibration of the system, the SE for measurements performed on geometric shapes was less than 2 mm, i.e., less than the actual pixel size of the SPECT data. Biometric measurements of an anthropomorphic brain phantom were within 3%–5% of actual values. The mask data were used to register images of a brain phantom and of a volunteer's brain, respectively. The rigid transformation that allowed the merging of masks by visual inspection was applied to the 2 sets of SPECT slices to perform the fusion of the data. Conclusion: At the cost of an additional low-cost setup integrated into the -camera gantry, real-time data about the surface of the head were obtained. As in all other surface-based techniques (as opposed to volume-based techniques), this method allows the match of data independently from the dataset of interest and facilitates further registration of data from any other source. The main advantage of this technique compared with other optically based methods is the robustness of the calibration procedure and the compactness of the sensor as a result of the colinearity of the projected beam and the reflected (diffused) beams of the conoscope. Taking into account the experimental nature of this preliminary work, significant improvements in the accuracy and speed of measurements (up to 1000 points/s) are expected.

Key Words: brain • registration • surface matching • SPECT