We propose to develop a completely new single photon emission computed tomography (SPECT) system using novel technologies employing cadmium zinc telluride (CZT), customized application-specific integrated circuit (ASIC), innovative collimator design, and ultrafast graphics processing unit (GPU) specialized for heaving computation in medical image reconstruction. With our proposed SPECT system, a single photon molecular imaging technology that will not require changes of collimators for different photon energies can be realized, at least for potential applications (breast, brain, and prostate) of small volume SPECT applications targeted by the proposed imaging system. By coupling the detectors to custom-built ASIC readout electronics, we can obtain high energy-resolution signals and bin the signals in multiple specific energy windows without requiring time-consuming list-mode acquisition and postprocessing. With these specific goals in mind, we emphasize that the primary goal is to remove the energy dependency of SPECT collimators so that the SPECT technology continues to be viable in the future. Our hypotheses are: Hypothesis 1: The CZT's excellent energy resolution and stopping power can distinguish background photons (photons scattered from human body and collimator septa) from quality photons regardless of emission energy. Hypothesis 2: A single collimator that provides high sensitivity can be used for the wide variety of SPECT radiotracers. Hypothesis 3: A fast computer algorithm that can correct the collimator-dependent blurring can provide excellent spatial resolution comparable to PET spatial resolution in the wide range of SPECT radiotracers using a single collimator.
Our specific aims to test the hypotheses are:
Aim 1 : We will develop a pair of small-pitch (1.5 mm) and large-area (20 cm x 20 cm) pixelated CZT detectors and associated application-specific integrated circuits (ASICs). The ASIC-driven electronics will be designed to acquire SPECT data using multiple narrow energy windows from CZT, not requiring list-mode data acquisition.
Aim 2 : We will develop a parallel-hole collimator with holes matched with CZT pixels to maximize the detection efficiency. We will use a Monte Carlo simulation tool to design the collimator to assess energy profiles from different photon energies for SPECT imaging.
Aim 3 : We will develop novel reconstruction algorithms that will compensate energy uncertainties and collimator-dependent blurring using high-speed computing techniques such as specialized novel graphics processing unit (GPU) that is heavily parallelized for computation in medical imaging.
The goal of this project is to develop a completely new single photon emission computed tomography (SPECT) system using novel technologies employing cadmium zinc telluride (CZT), customized application-specific integrated circuit (ASIC), innovative collimator design, and graphics processing unit (GPU) specialized for parallelized computation. We will focus applications of the new SPECT system to high-resolution, high-sensitivity small volume imaging of breast, brain, and prostatic bed.
|Lee, Tzu-Cheng; Ellin, Justin R; Huang, Qiu et al. (2014) Multipinhole collimator with 20 apertures for a brain SPECT application. Med Phys 41:112501|
|Lee, Tzu-Cheng; Burghardt, Andrew J; Yao, Wei et al. (2014) Improved trabecular bone structure of 20-month-old male spontaneously hypertensive rats. Calcif Tissue Int 95:282-91|
|Bowen, Jason D; Huang, Qiu; Ellin, Justin R et al. (2013) Design and performance evaluation of a 20-aperture multipinhole collimator for myocardial perfusion imaging applications. Phys Med Biol 58:7209-26|