Multiparametric transrectal ultrasound and photoacoustic imaging of human prostate

摘要 前列腺癌（PCa），随着发病率的逐年上升，已成为最常见的诊断 美国男性的癌症侵袭性PCa的准确诊断对患者的生存至关重要。 经直肠超声（TRUS）引导活检，目前的标准程序，用于评估存在和 PCa的侵略性，遭受低核心产量，导致采样不足和分级不足的临床 重大肿瘤。为了填补PCa诊断中长期存在的技术空白，我们提出开发一种新的 多模态成像平台，该平台将新兴的光声（PA）成像技术与 建立了经直肠超声（TRUS）多参数PCa诊断和分级。与唯一 能够映射反映病理学的结构/解剖、机械、功能和分子信息 在整个人体前列腺的体内条件下，非侵入性地，提出的TRUS和PA（即TRUSPA） 多参数成像可以灵敏地检测空间分布的前列腺肿瘤，更重要的是， 区分侵袭性肿瘤和非侵袭性肿瘤。因此，多参数TRUSPA（mpTRUSPA）， 当以与当前TRUS相同的方式进行时，可以通过以下方式大大改善活检采样的指导 确保在前列腺的最具侵袭性的恶性区域内高效地进行核心提取。 中心假设是实时mpTRUSPA成像-显示基于结构的B型超声 信息，基于能量多普勒超声的血管血流，基于剪切波弹性成像（SWE）的组织硬度， 以及用于绘制缺氧和血管生成的光谱PA图像-可以灵敏地表征癌症 通过绘制PCa的一组结构、机械和功能生物标志物， 可以随后引导活检取样。这一假设将被客观和严格的审查，通过 针对临床相关PCa动物模型和人类受试者的拟议研究，具体目的如下： 目的1：开发和验证mpTRUSPA成像系统，用于研究持久的前列腺模拟体模 以及侵袭性和非侵袭性前列腺癌的小鼠模型。目标2：开发和验证人工智能 (AI)能够进行信号去噪和定位、前列腺分割、伪影校正的框架，以及 实时地对深部组织派信号进行光谱解混。目的3：筛选一组PCa成像生物标志物， 在临床环境中可通过AI增强型mpTRUSPA成像系统在体内访问。 一旦PCa的mpTRUSPA成像性能通过拟议的动物模型研究得到验证， 和人类受试者，这种负担得起的和用户友好的技术将准备适当的I/II期临床 在下一个供资周期进行试验。TRUSPA，具有独特的临床检测和鉴别能力 重要的PCa非侵入性，在真实的时间，可以无缝地适应到目前的临床程序的图像引导 活检，并具有很大的潜力，以改变模式的PCa诊断。

ABSTRACT Prostate cancer (PCa), with an annually increasing incident rate, has become the most commonly diagnosed cancer in American men. The accurate diagnosis of aggressive PCa is critical for the survival of patients. Transrectal ultrasound (TRUS) guided biopsy, the current standard procedure for evaluating the presence and aggressiveness of PCa, suffers from low core yield, leading to under-sampling and under-grading of clinically significant tumors. To fill the long-standing technical gap in PCa diagnosis, we propose to develop a novel multimodal imaging platform which integrates the emerging photoacoustic (PA) imaging technique with the established transrectal ultrasound (TRUS) for multiparametric PCa diagnosis and grading. With the unique capability to map structural/anatomical, mechanical, functional, and molecular information reflecting pathological conditions over the entire human prostate in vivo, non-invasively, the proposed TRUS and PA (i.e. TRUSPA) multiparametric imaging can sensitively detect spatially distributed prostate tumors and, more importantly, differentiate aggressive tumors from non-aggressive ones. Hence, multiparametric TRUSPA (mpTRUSPA), when performed in the same way as in current TRUS, could largely improve the guidance of biopsy sampling by ensuring highly effective core extraction within the most aggressive malignant regions of the prostate. The central hypothesis is that real-time mpTRUSPA imaging – displaying B-mode US based structural information, Power Doppler US based vascular flow, shear wave elastography (SWE) based tissue stiffness, and spectral PA images for mapping hypoxia and angiogenesis - can sensitively characterize cancer aggressiveness in vivo by mapping a group of structural, mechanical, and functional biomarkers of PCa, and can subsequently guide the biopsy sampling. This hypothesis will be objectively and rigorously examined via the proposed studies on clinically relevant PCa animal models and human subjects using the following specific aims: Aim 1: Develop and validate mpTRUSPA imaging system using studies on durable prostate mimicking phantoms and mouse models of aggressive and non-aggressive PCa. Aim 2: Develop and validate an artificial intelligence (AI) framework capable of signal denoising and localization, prostate segmentation, artifact correction, and spectral unmixing of deep tissue PAI signals in real-time. Aim 3: Validate a set of PCa imaging biomarkers that are accessible in vivo by the AI enhanced mpTRUSPA imaging system in a clinical setting. Once the performance of mpTRUSPA imaging of PCa is validated via the proposed studies on animal models and human subjects, this affordable and user-friendly technology will be ready for appropriate phase I/II clinical trials in the next funding cycle. TRUSPA, with the unique capability of detecting and differentiating clinically significant PCa non-invasively, in real time, can seamlessly fit into the current clinical procedure of image-guided biopsy, and holds great potential to shift the paradigm of PCa diagnosis.