Evaluation of a new multipurpose whole-body CzT-based camera: comparison with a dual-head Anger camera and first clinical images

Abstract

Background: Evaluate the physical performance of the VERITON CzT camera (Spectrum Dynamics, Caesarea, Israel) that benefits from new detection architecture enabling whole-body imaging compared to that of a conventional dual-head Anger camera.

Methods: Different line sources and phantom measurements were performed on each system to evaluate spatial resolution, sensitivity, energy resolution and image quality with acquisition and reconstruction parameters similar to those used in clinical routine. Extrinsic resolution was assessed using ^99mTc capillary sources placed successively in air, in a head and in a body phantom filled with background activity. Spectral acquisitions for various radioelements used in nuclear medicine (^99mTc, ¹²³I, ²⁰¹Tl, ¹¹¹In) were performed to evaluate energy resolution by computing the FWHM of the measured photoelectric peak. Tomographic sensitivity was calculated by recording the total number of counts detected during tomographic acquisition for a set of source geometries representative of different clinical situations. Sensitivity was also evaluated in focus mode for the CzT camera, which consisted of forcing detectors to collect data in a reduced field-of-view. Image quality was assessed with a Jaszczak phantom filled with 350 MBq of ^99mTc and scanned on each system with 30-,20-,10- and 5-min acquisition times.

Results: Extrinsic and tomographic resolution in the brain and body phantoms at the centre of the FOV was estimated at 3.55, 7.72 and 6.66 mm for the CzT system and 2.47, 7.75 and 7.72 mm for the conventional system, respectively. The energy resolution measured at 140 keV was 5.46% versus 9.21% for the Anger camera and was higher in a same manner for all energy peaks tested. Tomographic sensitivity for a point source in air was estimated at 236 counts·s^-1·MBq^-1 and increased to 1159 counts·s^-1·MBq^-1 using focus mode, which was 1.6 times and 8 times greater than the sensitivity measured on the scintillation camera (144 counts·s^-1·MBq^-1). Head and body measurements also showed higher sensitivity for the CzT camera in particular with focus mode. The Jaszczak phantom showed high image contrast uniformity and a high signal-to-noise ratio on the CzT system, even when decreasing acquisition time by 6-fold. Representative clinical cases are shown to illustrate these results.

Conclusion: The CzT camera has a superior sensitivity, higher energy resolution and better image contrast than the conventional SPECT camera, whereas spatial resolution remains similar. Introduction of this new technology may change current practices in nuclear medicine such as decreasing acquisition time and activity injected to patient.

Keywords: Molecular imaging; Multipurpose CzT-camera; SPECT; Whole-body imaging.

Fig. 1

General camera architecture showing the different movements of the 12 detectors (a), schematic principle of the focus mode showing the reduced swipe motion of detectors to a predefined region of interest shown in dark grey (b), design of the detection column consisting of an array of 16 by 128 pixel units (c) and picture of the system (d)

Fig. 2

Estimated radial spatial resolution of Anger and CzT cameras as a function of the distance to the centre of the field of view (ranging from 0 to 9 cm) measured in air (a) in a head phantom (from IEC Standard 61675-2) (b) and in a body phantom (from IEC Standard 61675-1) (c)

Fig. 3

Axial reconstructed slices (similar scale with min and max threshold respectively set to 0 and 125%) of the Jaszczak phantom obtained for CzT (a) and Anger cameras (b) centred on cold spheres (top rows) or capillaries (bottom rows) for 30, 20, 10 or 5 min of acquisition time (left to right)

Fig. 4

Image contrast measured as a function of sphere diameter at different acquisition times for CzT and Anger cameras (a). Total recorded counts (b), integral uniformity (c), and RMS noise (d) calculated for different acquisition times for CzT and Anger cameras

Fig. 5

Example of a manually drawn VOIs for the thyroid study (a): spherical VOIs of 5 mm of diameter were drawn on lesion corresponding to a cold nodule (blue) and normal tissue (yellow) VOIs were drawn on CzT camera images (left) and copied to Anger camera images (right). The resulting CNRs were computed for each clinical case and both cameras (b)

Fig. 6

SPECT imaging of 30-min duration performed successively on CzT (with (a) and without (b) CT image fusion) and Anger cameras (c) 48-h after intravenous injection of 5.27 MBq of ²²³Ra. An uptake on a right shoulder bone metastasis was visible on axial, coronal and sagittal slices (left to right)