Abstract
1154
Introduction: Due to the important physiological functions and relatively complex anatomical structures of the tumor-involved regions of head-and-neck cancer (HNC), there is a need for high sensitivity and high spatial resolution PET system to evaluate small lymph nodes (<5 mm), establish how far the tumor has invaded locally, and guide the decision to resect a tumor rather than irradiate and deliver chemotherapy. In order to design with empathy and have easy integration with the clinical workflow, we are developing a two-panel geometry dedicated HNC PET scanner with adjustable spacing [1]. Cadmium zinc telluride (CZT) is used as the detector material, that can achieve better detector Compton scattering recovery due to better spatial and energy resolution [2]. In this article, we demonstrate the low electronic noise level of the proposed HNC dedicated PET system.
Methods: The detector crystal is based on 4.0 × 4.0 × 0.5 cm3 CZT. Each CZT crystal has 39 anode strips, 38 steering electrode strips, and 8 cathode strips. Two identical CZT crystals are assembled to a common flexible circuit based on anode-cathode-cathode-anode stacking to construct a 4.0 × 4.0 × 1.0 cm3 CZT detector module [3]. In order to maintain the high energy resolution of CZT crystals, in a scaled-up PET system with a large number of CZT detectors (300 CZT crystals), it is crucial to keep the noise of the back-end electronic circuits as low as possible. To quantify the contribution of the electronic noise level of the external circuit on the energy resolution and to estimate a lower bound for energy resolution, first, we used a square wave pulse as the excitation source for providing charge injection in each channel. The square wave pulse frequency was 1 kHz with 500 mV peak-to-peak amplitude with no offset, which simulates approximately equivalent charge induced by a 511 keV photon in the CZT crystal. In our system studies, data was acquired utilizing 40 RENA boards and two fan-in boards (each fan-in board with 20 RENA boards). This experiment was repeated 5 times to calculate the standard deviation. In each stage, the spectral peak with a full width at half maximum (FWHM) was calculated in both ADC and keV units for all the channels.
Results: We summarize in Table 3 the contribution of electronic noise of the back-end readout circuit on the energy resolution using a test pulse. The RENA boards only produce a spectral peak with FWHM of 0.96±0.19% at 511 keV (7.85±1.53 ADC units or 4.92±0.96 keV after conversion). Comparing this result with previously reported results, testing 18 RENA boards, no significant cross-talk and degradation were observed [4]. Compared to the previous CZT-based system [3], our proposed system has a relatively better performance.
Conclusions: We are developing a dedicated head-and-neck cancer PET scanner at UC Santa Cruz. The electronic noise level of part of the head-and-neck dedicated CZT based PET system has been characterized, which showed the low-noise and no significant cross-talk of readout electronics as we scale-up the system with more boards. We will scale the detector module up to a two-panel system and characterize its energy resolution and coincidence timing performance in the presentation. Funding support: We acknowledge the funding support from the NIBIB of the NIH under Award Number R01EB028091. References: [1] Zhang, H. et al, 2020. Penalized maximum-likelihood reconstruction for improving limited-angle artifacts in a dedicated head and neck PET system. Physics in Medicine & Biology. [2] Yang, S. et al, 2019. Effect of CZT System Characteristics on Compton Scatter Event Recovery. IEEE Transactions on Radiation and Plasma Medical Sciences, 4(1), pp.91-97. [3] Abbaszadeh, S. et al, 2017. New-generation small animal positron emission tomography system for molecular imaging. Journal of Medical Imaging, 4(1), p.011008. [4] Y. Wang et al, “Back-end readout electronic design and initial results: a head and neck dedicated PET system based on CZT,” Proc. SPIE Medical Imaging, 2021.