太赫兹计算层析成像研究进展（特邀）

太赫兹波作为一种穿透性强、具有非电离性和惧水性的电磁波,可以穿透多种非金属、非极性介质材料。太赫兹计算层析成像技术基于傅里叶中心切片定理和直线传播模型,通过记录不同投影角度下的强度数据,采用滤波反投影等重建算法获得样品三维吸收系数分布和内外部结构信息分布。随着太赫兹成像器件的不断发展和应用场景的拓展,已发展出多种照明模式、成像光路和重建算法,并已在文物保护、骨密度测量和无损检测领域开展了应用探索。概述太赫兹计算层析技术的基本原理,并从提高重建质量、分辨率和采集效率三方面具体介绍太赫兹计算层析成像技术的最新研究。     

连续波THz-CT在医学成像中的应用研究

太赫兹已经被用于对三维物体进行计算机辅助层析成像,但具有很高的折射损耗。设计了基于耿氏二极管和返波振荡器的连续波断层成像实验装置,对聚四氟乙烯模型进行扫描获得投影数据,然后利用滤波反投影算法对投影数据进行重建得到断面图像,用于研究折射对投影及图像的影响。最后对人体牙齿样本进行扫描和重建,探讨THz-CT在医学成像特别是牙齿等骨骼成像中应用的可行性和前景。     

连续波THz-CT 的折射问题研究

设计了基于返波振荡器的连续波太赫兹CT 成像实验装置,实现了用连续太赫兹波对三维物体的CT 断层成像.虽然可以得到较好的重建图像,但由于折射率大于1.5 的材料对太赫兹射线折射较强而影响成像质量.首先通过对聚四氟乙烯模型进行扫描获得投影数据,并利用滤波反投影算法对投影数据进行重建得到断层图像,然后分析了高折射对投影探测及图像重建的影响.最后对人体牙齿样本进行投影扫描和断层重建,获得的断面图像可以反映牙齿内部结构,但由于牙齿高折射率的影响,成像精度和分辨率有待进一步提高.而且牙釉质的厚度会严重影响太赫兹波在牙齿中的透射深度.研究结果可以为提高连续波太赫兹CT 成像质量提供参考,促进其在医学成像特别是牙齿等骨骼成像中的应用.     

太赫兹计算机辅助层析图像重构算法仿真研究

由于太赫兹波具有独特的性质,使得太赫兹成像技术成为目前的研究热点。太赫兹层析成像可以获得物体横截面的分布信息并可获得物体的三维重构图像,因此也受到了广泛关注。文中对连续太赫兹层析成像进行了仿真研究。分别使用滤波反投影算法（FBP）和改进的联合代数重建算法（MSART）进行图像重构,并且分析比较了高斯低通滤波（GLPF）以及数学形态学等数字图像处理方法对改善重构图像质量的效果。仿真结果表明:在文中的仿真条件下,使用MSART算法及相应的图像处理方法所需的最少投影方向数可以达到9,与真实成像实验结果相吻合。     

基于光子混频连续太赫兹波的厚度检测

为了将光子混频的连续太赫兹波透射成像系统应用于样品厚度检测中,采用该系统获得的相位信息对样品进行了2维厚度测量。利用两个外腔半导体结构的激光器搭建了基于光子混频的连续太赫兹波透射成像系统,并利用X-Y 2维电动平移台放置样品进行点点扫描成像。该系统可同时获得样品的幅度信息和相位信息,在太赫兹波辐射频率0.47THz时,系统信噪比可达68dB。计算得出的厚度值与实际样品的厚度值最大相差0.02mm;另外还分析了平行平面样品干涉效应和样品不同透射强度对厚度测量的影响。结果表明,样品折射率越高,平行平面干涉效应对厚度测量影响越大;样品透射系数越大,测量精度也越高。当样品太赫兹波透射系数大于0.5时,厚度测量精度优于2.0%。     

BWO连续太赫兹波成像检测

返波管（BWO）连续太赫兹波成像方法是一种新的无损检测方法。实验过程中把样品放在X-Z二维电控平移台上进行扫描成像,透过样品的太赫兹波强度信息由热释电探测器接收,再经由电脑成像。该文给出了应用0.71 THz的连续太赫兹波对打孔铝板、公交卡、校园卡的内部结构和对隐藏在信封内硬币等物体和信封内纸片上的铅字迹的成像实验研究事例,并且测知该系统能够分辨出最小直径为1.5 mm的小孔。并且对成像图像进行了数字图像处理,结论表明,优化处理后的图像更加直观明显,提高了连续波成像的应用能力。揭示了BWO连续太赫兹波成像系统在无损检测和安检领域是实际有效的。     

连续太赫兹波合成孔径数字全息成像方法

太赫兹波数字全息成像技术结合了太赫兹成像技术和数字全息成像技术的优势,是一种新型相衬成像技术,具有光源相干性要求低、光路结构简单、实时定量获取物光波复振幅信息等特点,非常适用于太赫兹波段成像。影响数字全息成像技术分辨力的因素有多种,其中一个主要因素是探测器靶面尺寸的大小,因此,提出合成孔径方法扩大探测器靶面尺寸,提高太赫兹数字全息的成像分辨力。文中搭建了连续太赫兹波同轴数字全息成像装置,获取了样品高质量、高分辨力的振幅和相衬图像。实验结果有效说明了合成孔径方法可以提高太赫兹数字全息成像分辨力。     

采用CMOS太赫兹波探测器的成像系统

太赫兹波成像技术在生物医疗和安全检测等领域具有广阔的应用前景。针对新一代信息技术对便携式太赫兹波成像设备的需求，设计了基于CMOS太赫兹波探测器的成像系统。该系统包括一款CMOS太赫兹波探测器、片外模数转换器（ADC）、FPGA数字信号处理器、二位步进机、四个抛物面镜和太赫兹波辐射源等。CMOS太赫兹波探测器集成了片上贴片天线以及作为检波元件的NMOS晶体管，探测器由180 nm标准CMOS工艺制成。太赫兹波探测器的输出被片外模数转换器（ADC）采集并转换为数字信号，该数字信号被FPGA采集并传输到电脑上成像。所有上述元件均被装备在印刷线路板（PCB）上以减小系统体积。该系统实现了透射式太赫兹波扫描成像而无需斩波-锁相技术，并给出在860 GHz的太赫兹波照射下隐藏在信封内部金属的成像结果。     

一种用于人体安检的三维稀疏太赫兹快速成像算法

太赫兹全息成像技术在人体安检成像、隐匿武器检测、无损检测等领域具有广阔的应用前景。该文提出了一种能够避免产生距离混叠现象的3维稀疏太赫兹快速成像体制和成像算法。该方法通过对太赫兹雷达的3维成像几何及相应回波模型的分析,利用随机稀疏频点数据来有效消除距离混叠现象;同时利用对稀疏回波数据的频谱搬移和插值来获得包含目标完整信息的3维数据,以实现对目标的3维分辨成像。通过对0.2 THz波段的仿真和实验数据的重建,验证了算法的正确性和有效性。      

深度学习用于连续太赫兹同轴数字全息重建

由于太赫兹面阵探测器像元数少，且目标像素数较少，全息图的衍射效应明显，因此其重建较可见光全息图重建困难。研究两种将深度学习用于二维连续太赫兹同轴数字全息振幅重建的方法，并与传统的角谱法（ASM）和带切趾的振幅约束相位恢复算法（APRA）进行对比。第一种是端对端的U-net网络重建方法（H-UnetM），即网络输入图像为全息图；第二种是角谱法加U-net网络重建方法（AS-UnetM）。仿真研究表明，对于记录距离15～20 mm、分辨率0.3～0.5 mm目标的2.52 THz全息图，AS-UnetM重建优于APRA，而H-UnetM仅优于ASM但不如APRA。最后通过真实实验加以验证，结果表明H-UnetM能够重建目标，但部分背景噪声也被突出，而采用ASUnetM在目标附近的重建效果最佳。     

Micro-CT of Pseudocneorhinus bifasciatus by projection X-ray microscopy

The projection X-ray microscope utilises a very small X-ray source emitted from a thin (0.1-3 microm) target metal film excited by the focused electron beam of a scanning electron microscope. When an object is placed just below the target metal film, the diverging X-rays enlarge the shadow of the object. Because no X-ray optics such as a zone-plate is used, the focal depth is, in principle, infinitely large. We exploited this to apply projection X-ray microscopy to three-dimensional (3-D) structure analysis by means of cone-beam computed tomography. The projection images of a small arthropod (Pseudocneorhinus bifasciatus, 5 mm in length), was recorded at 3 degrees increments over the whole range (360 degrees) of a stepping-motor-controlled sample rotator. A 3-D image was reconstructed from corn-beam projections using a filtered back-projection algorithm. The reconstructed 3-D image showed in detail the internal structure of an opaque object.     

Three-dimensional surface reconstruction using optical flow formedical imaging

The recovery of a three-dimensional (3-D) model from a sequence of two-dimensional (2-D) images is very useful in medical image analysis. Image sequences obtained from the relative motion between the object and the camera or the scanner contain more 3-D information than a single image. Methods to visualize the computed tomograms can be divided into two approaches: the surface rendering approach and the volume rendering approach. In this paper, a new surface rendering method using optical flow is proposed. Optical flow is the apparent motion in the image plane produced by the projection of real 3-D motion onto the 2-D image. The 3-D motion of an object can be recovered from the optical-flow field using additional constraints. By extracting the surface information from 3-D motion, it is possible to obtain an accurate 3-D model of the object. Both synthetic and real image sequences have been used to illustrate the feasibility of the proposed method. The experimental results suggest that the proposed method is suitable for the reconstruction of 3-D models from ultrasound medical images as well as other computed tomograms     

A global energy function for the alignment of serially acquired slices

An accurate, computationally efficient, and fully automated algorithm for the alignment of two-dimensional (2-D) serially acquired sections forming a three-dimensional (3-D) volume is presented. The approach relies on the optimization of a global energy function, based on the object shape, measuring the similarity between a slice and its neighborhood in the 3-D volume. Slice similarity is computed using the distance transform measure in both directions. No particular direction is privileged in the method avoiding global offsets, biases in the estimation and error propagation. The method was evaluated on real images [medical, biological, and other computerized tomography (CT) scanned 3-D data] and the experimental results demonstrated its accuracy as reconstuction errors are less than one degree in rotation and less than one pixel in translation.     

Improved SPECT quantitation using fully three-dimensional iterative spatially variant scatter response compensation

The quality and quantitative accuracy of iteratively reconstructed SPECT images improves when better point spread function (PSF) models of the gamma camera are used during reconstruction. Here, inclusion in the PSF model of photon crosstalk between different slices caused by limited gamma camera resolution and scatter is examined. A three-dimensional (3-D) projector back-projector (proback) has been developed which models both the distance dependent detector point spread function and the object shape-dependent scatter point spread function of single photon emission computed tomography (SPECT). A table occupying only a few megabytes of memory is sufficient to represent this scatter model. The contents of this table are obtained by evaluating an analytical expression for object shape-dependent scatter. The proposed approach avoids the huge memory requirements of storing the full transition matrix needed for 3-D reconstruction including object shape-dependent scatter. In addition, the method avoids the need for lengthy Monte Carlo simulations to generate such a matrix. In order to assess the quantitative accuracy of the method, reconstructions of a water filled cylinder containing regions of different activity levels and of simulated 3-D brain projection data have been evaluated for technetium-99m. It is shown that fully 3-D reconstruction including complete detector response and object shape-dependent scatter modeling clearly outperforms simpler methods that lack a complete detector response and/or a complete scatter response model. Fully 3-D scatter correction yields the best quantitation of volumes of interest and the best contrast-to-noise curves.     

Exact (spiral + circles) scan region-of-interest cone beam reconstruction via backprojection

We present a (spiral + circles) scan cone beam reconstruction algorithm in which image reconstruction proceeds via backprojection in the object space. In principle, the algorithm can reconstruct sectional region-of-interest (ROI) in a long object. The approach is a generalization of the cone beam backprojection technique developed by Kudo and Saito in two aspects: the resource-demanding normalization step in the Kudo and Saito's algorithm is eliminated through the technique of data combination that we published earlier, and the elimination of the restriction that the detector be big enough to capture the entire cone beam projection of the ROI. Restricting the projection data to the appropriate angular range required by data combination can be accomplished by a masking process. Because of the simplification resulting from the elimination of the normalization step, the most time-consuming operations of the algorithm can be approximated by the efficient step of line-by-line ramp filtering the cone beam image in the direction of the scan path, plus a correction image. The correction image, which can be computed exactly, is needed because data combination is not properly matched at the mask boundary when ramp filtering is involved. Empirical two-dimensional (2-D) point spread function (PSF) is developed to improve matching with the correction image which is computed with finite samplings. The use of transition region to further improve matching is introduced. The results of testing the algorithm on simulated phantoms are presented.     

Three-dimensional electrical impedance tomography based on thecomplete electrode model

In electrical impedance tomography an approximation for the internal resistivity distribution is computed based on the knowledge of the injected currents and measured voltages on the surface of the body. It is often assumed that the injected currents are confined to the two-dimensional (2-D) electrode plane and the reconstruction is based on 2-D assumptions. However, the currents spread out in three dimensions and, therefore, off-plane structures have significant effect on the reconstructed images. In this paper we propose a finite element-based method for the reconstruction of three-dimensional resistivity distributions. The proposed method is based on the so-called complete electrode model that takes into account the presence of the electrodes and the contact impedances. Both the forward and the inverse problems are discussed and results from static and dynamic (difference) reconstructions with real measurement data are given. It is shown that in phantom experiments with accurate finite element computations it is possible to obtain static images that are comparable with difference images that are reconstructed from the same object with the empty (saline filled) tank as a reference.     

A Hybrid Compression Method for Integral Images Using Discrete Wavelet Transform and Discrete Cosine Transform

Integral imaging (II) is a promising three-dimensional (3-D) imaging technique that uses an array of diffractive or refractive optical elements to record the 3-D information on a conventional digital sensor. With II, the object information is recorded in the form of an array of subimages, each representing a slightly different perspective of the object In order to obtain high-quality 3-D images, digital sensors with a large number of pixels are required. Consequently, high-quality II involves recording and processing large amounts of data. In this paper, we present a compression method developed for the particular characteristics of the digitally recorded integral image. The compression algorithm is based on a hybrid technique implementing a four-dimensional transform combining the discrete wavelet transform and the discrete cosine transform. The proposed algorithm outperforms the baseline JPEG compression scheme applied to II and a previous compression method developed for II based on MPEG II.     

3-D computer simulations of resolution effects of multidetector point focusing SPECT imaging

The authors present an efficient algorithm and the results of its application in simulating the three-dimensional (3-D) projection data resulting from a 3-D distribution of radioactivity. The algorithm was applied to a series of geometrical mathematical phantoms and to a realistic mathematical brain phantom. The authors simulated the projection data from a multidetector single-photon emission computed tomography (SPECT) system with point focusing collimators. The simulated projection data were then reconstructed using the manufacturer's software. The objects simulated included simple geometrical solids such as spheres and sheets, as well as the distribution of muscarinic cholinergic receptors in a realistic brain slice. Spheres were chosen as a model for brain structures such as caudate nucleus, thalamus, and cerebellum; sheets were selected as representing lateral cortical gray matter regions. The results of these simulations indicate the existence of significant qualitative and quantitative artifacts in reconstructed human brain images.     

A cone-beam reconstruction algorithm for circle-plus-arc data-acquisition geometry

In cone-beam computerized tomography (CT), projections acquired with the focal spot constrained on a planar orbit cannot provide a complete set of data to reconstruct the object function exactly. There are severe distortions in the reconstructed noncentral transverse planes when the cone angle is large. In this work, a new method is proposed which can obtain a complete set of data by acquiring cone-beam projections along a circle-plus-arc orbit. A reconstruction algorithm using this circle-plus-arc orbit is developed, based on the Radon transform and Grangeat's formula. This algorithm first transforms the cone-beam projection data of an object to the first derivative of the three-dimensional (3-D) Radon transform, using Grangeat's formula, and then reconstructs the object using the inverse Radon transform. In order to reduce interpolation errors, new rebinning equations have been derived accurately, which allows one-dimensional (1-D) interpolation to be used in the rebinning process instead of 3-D interpolation. A noise-free Defrise phantom and a Poisson noise-added Shepp-Logan phantom were simulated and reconstructed for algorithm validation. The results from the computer simulation indicate that the new cone-beam data-acquisition scheme can provide a complete set of projection data and the image reconstruction algorithm can achieve exact reconstruction. Potentially, the algorithm can be applied in practice for both a standard CT gantry-based volume tomographic imaging system and a C-arm-based cone-beam tomographic imaging system, with little mechanical modification required.