DEVELOPMENT OF SLOW SCAN CAMERA & INTERFACE

这个核心研发项目的目标是开发一个系统，用于 NCMIR IVEM在分辨率、速度等方面进行了必要的增强 和灵敏度为用户提供一种电子读出装置 与使用新一代2k x 2k的胶片相媲美或更好 CCD成像系统。这个子项目对此非常重要 研究计划。这将使我们能够改进图像采集 计算机化三维重建的速度和精度 厚重生物标本的可视化。我们最近做了 完成了新的镜头耦合摄像系统。它已安装完毕 在IVEM上，正在接受测试。这其中有五个部分 项目：1)闪光屏开发：两项美国专利 (#5,401,964和#5,594,253)已发给我们，其中包括一份 1997年1月发布。一种基于薄箔的P20荧光屏具有 是格兰特科学公司为这个系统制作的，根据我们的 专利设计。该设计首先被描述和表征在 1994(Fan和Ellisman，超微显微镜55：7-14,1994)，并优化 1996年(Fan等人，超镜66：11-19,1994)；2)透镜耦合 系统，由光学研究协会根据我们的 规格，由廷斯利实验室制造，是 已交付并已安装在IVEM上。设计目标是 对光学设计师和制造商都具有挑战性，以及 在第一次交付镜片后，廷斯利不得不进行了修改， 由于透镜未能达到我们光学系统中的一些设计目标 台架测试。在此之后，性能有了显著的提高 修改。MOD仿真传递函数(MTF)为55% 奈奎斯特频率，并且在整个视场内几乎是平坦的， 这是一个直径超过10厘米的区域。整体灯光 透过率为83%，超过了80%的设计目标。这个 光学系统的分辨率和继电效率很好地匹配 闪烁屏幕的分辨率和系统提供的分辨率 超过了光纤耦合系统的可能；3) 所采用的ccd芯片在技术上比现在先进。 商业上可用，并作为协作式 与麻省理工学院林肯实验室和美国空军的研究工作。 该设备采用了非常先进的技术，具有8个高带宽 可以并行读出的端口。虽然只有四个端口 在我们的实施中使用，我们仍然将实现 与我们当前的1k x 1k设备相比，速度更快，但能够 想象一个超过我们目前的CCD成像器两倍大小的阵列；4) 设计了与CCD摄像控制器的计算机接口，并进行了实验验证 实施。该接口采用基于Unix的工作站 到DataCube MV200图像处理器控制摄像头和 对图像进行多路分解和组装。一个新的图形界面具有 是为相机的使用而设计的。5)机械集成 相机组件是在室内使用一套3D实体设计的 建模/二维CADCAM软件工具。整个系统最初是 在3D中建模，以允许可视化和验证总 先整合后建设。从最终的优化模型， 为施工生成了工程示意图。这个 该系统的组件包括：真空兼容的落地法兰 它支撑和定位闪烁体屏幕和含铅玻璃 窗户，可调镜头支架五金，机械隔离， 陀螺仪相机支架外壳，允许亚微米 CCD芯片的居中和调整，自动旋转，以及 精确对焦，以及允许快速对焦的相机外壳适配器 交换2kx2k和1kx1k摄像头。初步测试 表明该成像系统的整体性能为 比胶片敏感得多，也比 400keV光纤耦合CCD系统(数据来自亚利桑那州立大学 大学)。进一步的量化评估正在开始进行。 与亚利桑那州的约翰·斯宾塞博士和左建明博士合作 州立大学。我们还计划探索应用程序的使用 可替代CCD的专用集成电路(ASIC)探测器 用来做透射电子显微镜成像。ASIC探测器是由Xong博士开发的 UCSD的Nguyen-Huu和他的X射线结晶学同事 申请。我们最近对这个装置进行了电子测试。 在80-400keV能量范围内的探测(Fan等人，由 超微显微镜，1997)，结果非常令人鼓舞。一个 基于ASIC的成像系统将具有许多优于 基于CCD的成像系统(参见第4A2.2节)。徐博士将 在这个项目上与我们合作。

The objective of this core TR&D project is to develop a system for the NCMIR IVEM with the necessary enhancements in resolution, speed and sensitivity to provide users with an electronic readout device comparable to or better than film using a next generation of 2k x 2k CCD imaging system . This subproject is very important to this research program. It will allow us to improve the image acquisition rate and precision for computerized 3-D reconstruction and visualization of thick biological specimens. We have recently completed the new lens coupled camera system. It has been installed on the IVEM and is undergoing testing. There are five parts to this project: 1) scintillating screen development: two US patents (#5,401,964 and #5,594,253) have been issued to us including one just issued in January, 1997. A thin foil-based P20 phosphor screen has been made for this system by Grant Scientific according to our patented design. The design was first described and characterized in 1994 (Fan and Ellisman, Ultramicroscopy 55:7-14, 1994) and optimized in 1996 (Fan et al, Ultramicroscopy 66:11-19, 1994) ; 2) Lens coupling system, designed by Optical Research Associates according to our specifications, and manufactured by Tinsley Laboratories, was delivered and has been installed on the IVEM. The design goal was challenging for both the optical designer and the manufacturer, and Tinsley had to make a modification after the lens was first delivered, as the lens failed to meet some of the designed goals in our optical bench test. The performance was significantly improved after the modification. The mod EMulation transfer function (MTF) is 55% at the Nyquist frequency, and is nearly flat across the entire field of view, which is an area over 10 cm in diameter. The overall light transmittance is 83%, exceeding the design goal of 80%. The resolution and relay efficiency of the optical system match well with that of the scintillating screen and the system delivers resolution exceeding that possible with a fiber-optically coupled system; 3) The CCD chip being employed is technologically more advanced than what is commercially available and was provided as part of a collaborative research effort with MIT's Lincoln Laboratory and the US Air Force. This device employs very advanced technology and has 8 high bandwidth ports which may be read out in parallel. Although only four ports are being used in our implementation, we will still achieve a substantial speedup as compared to our current 1k x 1k device, yet are able to image an array more than 2x the size of our current CCD imager; 4) The computer interface to the CCD camera controller has been designed and implemented. The interface employs a Unix-based workstation coupled to DataCube MV200 image processor to control the camera and to demultiplex and assemble the image. A new graphical interface has been designed for use of the camera. 5) Mechanical integration of the camera components was designed in house using a suite of 3D Solid Modeling/2D CADCAM software tools. The complete system was initially modeled in 3D to allow for visualization and validation of total integration prior to construction. From the final, optimized model, engineering schematics were generated for construction. The components of the system include: a vacuum compatible drop flange which supports and positions the scintillator screen and leaded glass window, adjustable lens support hardware, a mechanically isolated, gyroscopic camera support housing which allows for sub-micron centering and adjustment of the CCD chip, automated rotation, and precision focusing, and a camera housing adapter which allows quick swapping of the 2kx2k and 1k x1k camera heads. Preliminary tests indicate that the overall performance of this imaging system is considerably more sensitive than film and better than a fiber-optically coupled CCD system at 400 keV (Data from Arizona State University). A further quantitative evaluation is beginning conducted in collaboration with Drs. John Spence and Jian Ming Zuo at Arizona State University. We also plan to explore the use of an Application Specific Integrated Circuit (ASIC) detector as an alternative to a CCD for TEM imaging. The ASIC detector was developed by Dr. Xuong Nguyen-Huu of UCSD and co-workers for X-ray crystallography applications. We have recently tested the device for electron detection in the energy range of 80-400 keV (Fan et al, in review by Ultramicroscopy, 1997), and the results are very encouraging. An ASIC-based imaging system will possess many advantages over the CCD-based imaging systems (see Section 4A2.2). Dr. Xuong will collaborate with us on this project.