A Preclinical SPECT System with Optimized Data Acquisition for Imaging Larger Animals

Objectives: Our aim is to create an optimized, adaptive preclinical SPECT system for imaging larger animals including rabbits. Conventional SPECT systems acquire data with only one imaging configuration, whereas this adaptive system employs an adaptive aperture design with the goal of obtaining rich data sets with varying levels of resolution, sensitivity, and field of view. Our objective is also to optimize the acquisition by digitally capturing the voltage waveforms off each PMT for every gamma-ray interaction, increasing the information content of the resulting signals and thereby improving the maximum-likelihood (ML) estimation of event parameters.

Methods: The preclinical adaptive SPECT imager named SmartCam is based on modifying a single-camera GE MaxiCamera 400T SPECT system, which consists of a 10-mm-thick, 470-mm-NaI(Tl) scintillation crystal, a fused silica light guide, and 61 PMTs arranged in a hexagonally packed pattern, each with their own acquisition circuitry. Pinhole collimation is used, with a modified aperture size (2 mm rather than 8 mm in the original system). In order to tap into the raw low-level signals before they reach the original analog Anger-logic network, active buffer amplifier circuits were introduced that intercept these signals. After amplification, they are sent off-camera via 50-Ω-terminated coaxial cables to be digitally sampled at a sampling rate of 66 MHz. Triggering is programmed on sums of the individual 12 bit signals, and complete waveforms are sent to a dedicated computer for processing. Once there, we implement an ML estimation algorithm using a generalized probability distribution function. A mean detector response function (MDRF) was acquired by translating a ^99mTc calibration source across the detector and acquiring data at small steps. This MDRF was then integrated into the ML estimation. The adaptive aperture design is comprised of 9-2 mm diameter pinholes of equal size arranged in a 3x3 grid. Each pinhole has a shutter that can be individually opened and closed. Determination of which pinholes will be open will be based on in-line image-quality assessments. Results: We have accomplished the retrofit of a commercial clinical SPECT camera with state of the art waveform capture electronics to create a new system specifically tailored to preclinical studies. We have developed ML-estimation methods for the waveform data sets that maximize information content. We are currently optimizing the system to image rabbit models with vascular diseases, including vulnerable lesions of atherosclerosis in the hyperlipidemic rabbit aorta and elastase-induced aneurysm. Comparative studies can be performed using non-adaptive acquisitions on the same SmartCam imaging system.

Conclusions: Updating of classic Anger-camera systems with state-of-the-art electronics and adaptive pinhole apertures are an effective approach to developing disease-specific preclinical systems. The utility of adaptation and the performance of the SmartCam will be demonstrated by vascular images of rabbit models. Funding:This work was partially supported by This work was partially supported by the NIH/NIBIB Grant P41-EB002035 “The Center for Gamma-Ray Imaging”. Neil Momsen was partially supported by grants: Biomedical Imaging and Spectroscopy Fellowship, NIH T32-EB000809, and Cardiovascular Training Grant, NIH T32HL007955.