Single-pixel diffuse optical tomography scanner for dynamic functional imaging

合理和客观的扩散光学断层扫描(DOT)是一种低成本的成像方式，可以提供组织吸收的三维(3D)图像，在近红外范围内，主要是由于血红蛋白和水的吸收。通过测量组织在多个波长的吸收，可以生成断层功能图像并用于检测局部的生理异常。正是这种非侵入性成像组织生理参数(并使用安全、便携的仪器)的能力使DOT成为一项如此有吸引力的技术。虽然DOT有许多优点，但它作为一种功能成像方式的广泛采用受到了当前空间和时间分辨率之间的权衡的限制。也就是说，当前的方法可以在空间分辨率较低的情况下高速获取数据，或者在毫米分辨率但时间分辨率较差的情况下获取数据。*本研究计划的目标是解决这一关键限制。*研究计划*1)我们将基于：*i)早期光子成像：这些光子沿着被限制在源和探测器之间的视线附近的路径传播(准模拟X射线)，这减少了图像模糊并显著提高了空间分辨率。*ii)广场图案化照明：数字微镜装置(DMD)由可单独打开或关闭的2D微镜阵列组成，将用于在组织的表面上产生广场图案化照明。我们将利用DMDS技术的最新进展，生产能够以非常高的速度(高达23 KHz)和高光输出(高达48%)产生光图案的单元，以建立快速高效的照明系统。*iii)压缩检测：通过组织传输的光将被投影到第二个DMD(对光强度进行空间编码)，然后由单个光电探测器集成。请注意，我们将使用单像素相机来生成3D图像，而无需扫描或多路传输。*2)我们将开发一个基于早期光子在组织中传播的准确蒙特卡罗模型的图像重建软件。*3)扫描仪的空间分辨率将使用具有空间分辨特征的组织模拟体模进行量化。*4)系统的动态成像能力将在局部缺血的动物模型中进行测试。*Impact*本提案中详细说明的研究计划将导致开发一种点扫描仪能够以毫米分辨率快速对厚组织(厚度高达60毫米)进行3D成像。这种扫描仪将使分子示踪剂的动态成像能够开发出方法，用于i)小动物分子成像，ii)监测炎性关节炎的治疗反应，以及iii)预测乳腺癌的化疗反应。

Rational and Objective****Diffuse optical tomography (DOT) is a low-cost imaging modality that can provide three-dimensional (3D) images of tissue absorption, which, in the near-infrared range, is mainly due to hemoglobin and water. By measuring tissue absorption at multiple wavelengths, tomographic functional images can be generated and used to detect localized physiological abnormalities. It is this capacity to noninvasively image tissue physiological parameters (and to do so using safe, portable instruments) that makes DOT such an attractive technology. While DOT has numerous advantages, its widespread adoption as a functional imaging modality has been restricted by the current tradeoff between spatial and temporal resolution. That is, current methods can either acquire data at high speed with poor spatial resolution, or at millimeter resolution but with poor temporal resolution.****The objective of this research program is to address this critical limitation.****Research Plan****1) We will build a DOT scanner based on:**** i) Early-photons imaging: These photons travel along paths that are confined close to the line of sight between the source and detector (quasi-mimicking X-rays), which reduces image blurring and significantly improves spatial resolution.**** ii) Wide-field patterned illumination: A digital micro-mirror device (DMD), made of a 2D array of micro-mirrors that can be individually turned “ON” or “OFF”, will be used to generate wide-field patterned illumination on the surface of the tissue. We will leverage recent advances in DMDs technology that produced units that can generate light patterns at very high speed (up to 23 kHz) with high light output (up to 48%), to build a fast and efficient illumination system.*** iii) Compressive detection: Light transmitted through the tissue will be projected onto a 2nd DMD (to spatially encode the light intensity), then integrated by a single photodetector. Notice that we will be using a single-pixel camera to generate 3D images without scanning or multiplexing.****2) We will develop an image reconstruction software based on accurate Monte Carlo modeling of early-photons propagation in tissue.****3) The spatial resolution of the scanner will be quantified using tissue-mimicking phantoms with spatially-resolved features.****4) The dynamic imaging capability of the system will be tested in an animal model of localized ischemia.****Impact****The research program detailed in this proposal will lead to the development of a DOT scanner capable of fast 3D imaging of thick tissue (up to 60-mm thickness) with millimeter resolution. Such a scanner will enable dynamic imaging of molecular tracers to develop methods for i) small animal molecular imaging, ii) monitoring treatment response in inflammatory arthritis, and iii) predicting response to chemotherapy in breast cancer.***