Design and Development of SyMPET: System on chip Modular readout for high-resolution TOF-PET

项目总结 NALU Science LLC(NSL)建议开发和优化基于SiPM的、 低功耗、高通道密度、波形数字化读出微芯片，用于TOF-PET，将 提高图像质量，为PET脑提供更准确和精确的量化 成像，有可能显著改善神经退行性疾病的早期诊断 同时还允许在开发个性化患者成像时具有更大的灵活性 战略。NSL获得专利的波形数字化“片上系统”读出ASIC技术 从目前的状态大幅改进TOF-PET成像的潜力。在第一阶段， 我们将在使用NSL技术获得的先前经验和知识的基础上 开发最先进的高能和核物理探测器以大幅改进 PET成像读出系统通过提高信噪比、图像对比度和质量，降低 曝光时间/剂量，并降低系统成本，以推动广泛接受。这些 改进最初将集中在脑PET成像上，但可以扩展到全身 系统。我们将利用NSL现有的低功耗、低成本WFD ASIC设计组合， 已经被证明可以在大粒子物理探测器中工作，实现PET专用的WFD ASIC针对脑部PET扫描仪进行了优化，但同样适用于全身。我们最初会 开发闪烁晶体中光产生和传输的详细分析模型， 使用传感器和读出电子设备的真实蒙特卡罗模拟，以便得出 完全优化的TOF-PET WFD读出芯片的基准技术指标 作为基本上基于派生自 一个或多个现有的NSL芯片设计。NSL的“片上系统”WFD架构，完全 可随机访问的模拟存储、输入触发和片上控制功能，允许 处理诸如吞吐量等设计问题的多个高效机制， 在WFC ASIC的技术规范中，速度和缓冲区长度将是优化的关键 在满足性能目标的同时，满足 需要的物理和其他特性，如尺寸、重量、功率和成本 任何逼真的TOF-PET系统。我们将与哈米德·萨贝特博士(哈佛)合作，定义一个 真实的全信号链读出模型，并随后评估其结果以生成 大幅改进WFD读数的ASIC技术规范和体系结构设计 相对于当前技术水平的TOF-PET系统的ASIC。

PROJECT SUMMARY Nalu Scientific LLC (NSL) proposes to develop and optimize the design of a SiPM-based, low-power, high channel density, waveform-digitizing readout microchip for TOF-PET that will increase image quality and provide more accurate and precise quantization for PET brain imaging, with the potential to significantly improve early diagnosis of neurodegenerative disease while also allowing greater flexibility in the development of personalized patient imaging strategies. NSL’s patented waveform-digitizing “System on Chip” readout ASIC technology has the potential to substantially improve TOF-PET imaging from its current state. During Phase I, we will build on the prior experience and knowledge we have gained using NSL’s technology in developing state-of-the-art high energy and nuclear physics detectors to substantially improve PET imaging readout systems through increased SNR, image contrast and quality, reduced exposure times/dose, and reduced system cost to drive broad acceptance. These improvements will initially focus on brain PET imaging but can be expanded to whole-body systems. We will leverage NSL’s existing portfolio of low-power, low-cost WFD ASIC designs, already proven to work in large particle physics detectors, to implement a PET-specific WFD ASIC optimized for brain PET scanners but equally applicable to whole-body. We will initially develop detailed analytic modeling of light production and transport in scintillating crystals, along with realistic Monte Carlo simulations of sensor and readout electronics in order to derive baseline technical specifications for both a fully optimized TOF-PET WFD readout chip as well as a “bare-bones” implementation substantially based on circuit design elements derived from one or more existing NSL chip designs. NSL’s “System on Chip” WFD architecture, with fully random accessible analog storage, input triggering, and on-chip control capability, allows for a number of highly effective mechanisms to cope with design issues such as e.g., throughput, speed, and buffer length, will be crucial to optimize in the technical specification of a WFC ASIC which meets performance goals while simultaneously fulfilling the stringent constraints on physical and other characteristics such as size, weight, power, and cost that will be required in any realistic TOF-PET system. We will collaborate with Dr. Hamid Sabet (Harvard) to define a realistic full signal chain + readout model and subsequently evaluate its results to generate ASIC technical specifications and architectural design for a substantially improved WFD readout ASIC for TOF-PET systems relative to the current state of the art.