A fast spin echo two-point Dixon technique and its combination with sensitivity encoding for efficient T2-weighted imaging

A fast spin echo two-point Dixon (fast 2PD) technique was developed for efficient T2-weighted imaging with uniform water and fat separation. The technique acquires two interleaved fast spin echo images with water and fat in-phase and 180° out-of-phase, respectively, and generates automatically separate water and fat images for each slice. The image reconstruction algorithm uses an improved and robust region-growing scheme for phase correction and achieves consistency in water and fat identification between different slices by exploiting the intrinsic correlation between the complex images from two neighboring slices. To further lower the acquisition time to that of a regular fast spin echo acquisition with a single signal average, we combined the fast 2PD technique with sensitivity encoding (SENSE). Phantom experiments show that the fast 2PD and SENSE are complementary in scan efficiency and signal-to-noise ratio (SNR). In vivo data from scanning of clinical patients demonstrate that T2-weighted imaging with uniform and consistent fat separation, including breath-hold abdominal examinations, can be readily performed with the fast 2PD technique or its combination with SENSE.     

Long-term evolution in the tropospheric concentration of chlorofluorocarbon 12 (CCl2F2) derived from high-spectral resolution infrared solar absorption spectra: retrieval and comparison with in situ surface measurements

The average tropospheric volume mixing ratios of chlorofluorocarbon 12 (CCl₂F₂) have been retrieved from high-spectral resolution ground-based infrared solar-absorption spectra recorded from March 1982 to October 2003 with the McMath Fourier transform spectrometer at the US National Solar Observatory facility on Kitt Peak in southern Arizona (31.9°N, 111.6°W, 2.09 km altitude). The retrievals are based on fits to the unresolved ν₈ band Q-branches near using the SFIT2 retrieval algorithm. The annual increase rate was equal to (16.88±1.37) parts per trillion (10^-12) by volume at the beginning of the time series, March 1982, or (4.77±0.04)%, 1 sigma, declining progressively to (2.49±1.24) parts per trillion, by volume at the end, October 2003, or (0.46±0.24)%, 1 sigma. Average tropospheric mixing ratios from the solar spectra have been compared with average surface flask and in situ sampling measurements from the Climate Monitoring and Diagnostics Laboratory (CMDL) station at Niwot Ridge, CO, (USA) (40.0°N, 105.5°W, 3013 m altitude). The average ratio and standard deviation of the monthly means of the retrieved tropospheric mixing ratios relative to the CMDL surface mixing ratios is (1.01±0.03) for the overlapping time period. Both datasets demonstrate the progressive impact of the Montreal protocol and its strengthening amendments on the trend of CCl₂F₂, though a tropospheric decrease has yet to be observed.     

基于水样类型识别的光谱COD测量方法

基于紫外吸收光谱的COD测量方法,尽管具有快速、实时、免试剂、无污染等优势。但该方法对于组分多变的水样适应性不强,构建的单一计算模型不能适用于所有待测水样类型,导致其在复杂环境下测量准确度较低,从而限制了其应用领域。本研究提出一种基于水样类型识别的测量方法。其过程包括:动态识别水样类型→自动选择相应的"吸光度(Auv)-COD"算法模型→计算COD。该方法有效提高了紫外光谱法COD测量的准确度和适用性。该研究在传统的光谱识别技术的基础上,针对COD实际测量的特点加以改进。选取水样吸光度曲线的形貌特征作为水样类型的表征参数,利用LM-BP神经网络作为识别算法。并引入了"历史数据队列"、"历史识别因子"的概念,在此基础上形成了级联的神经网络结构。该算法实现了COD测量应用中的高准确度的光谱识别,进而提高了复杂环境下COD测量的精度。大量实验测试和结果表明,与传统的光谱识别技术相比,该方法在COD测量应用中具有更好的鲁棒性和准确性。水样类型识别准确率达98%以上。同时算法结构简单,计算量小,适用于资源受限的小型化COD测量仪。当仪器在复杂多变的水环境中进行测量时,采用该算法测量得到的COD精度有显著的提高。该方法的提出为光谱COD测量法在水体组分多变场合的应用及提高其测量精度提供了技术保证,可望解决传统紫外光谱COD测量法难以适应变化和复杂水环境应用的问题。     

Using ultrasound CBE imaging without echo shift compensation for temperature estimation

Clinical trials have demonstrated that hyperthermia improves cancer treatments. Previous studies developed ultrasound temperature imaging methods, based on the changes in backscattered energy (CBE), to monitor temperature variations during hyperthermia. Echo shift, induced by increasing temperature, contaminates the CBE image, and its tracking and compensation should normally ensure that estimations of CBE at each pixel are correct. To obtain a simplified algorithm that would allow real-time computation of CBE images, this study evaluated the usefulness of CBE imaging without echo shift compensation in detecting distributions in temperature. Experiments on phantoms, using different scatterer concentrations, and porcine livers were conducted to acquire raw backscattered data at temperatures ranging from 37 °C to 45 °C. Tissue samples of pork tenderloin were ablated in vitro by microwave irradiation to evaluate the feasibility of using the CBE image without compensation to monitor tissue ablation. CBE image construction was based on a ratio map obtained from the envelope image divided by the reference envelope image at 37 °C. The experimental results demonstrated that the CBE image obtained without echo shift compensation has the ability to estimate temperature variations induced during uniform heating or tissue ablation. The magnitude of the CBE as a function of temperature obtained without compensation is stronger than that with compensation, implying that the CBE image without compensation has a better sensitivity to detect temperature. These findings suggest that echo shift tracking and compensation may be unnecessary in practice, thus simplifying the algorithm required to implement real-time CBE imaging.     

A Study on Usefulness Comparison of Multiple-Shot FSE PROPELLER and IDEAL-FSE Techniques for Magnetic Susceptibility Artifact

The aim of this study was to visualize multiple-shot fast spin echo (FSE) images using a periodically rotated overlapping parallel line with enhanced reconstruction (PROPELLER) technique. An iterative decomposition of water and fat with echo asymmetry and a least squares estimation (IDEAL) technique were also performed to reduce the image distortion or susceptibility artifact depending on the difference in magnetic susceptibility to the surrounding tissues caused by metal insertion. The utility of this technique was examined quantitatively. A ferromagnetic image was generated from all the metals, but the IDEAL technique caused less image distortion than the PROPELLER technique. The 3-point IDEAL technique, which used the difference in the signal phase of fat and water, required more time for the examination and image reconstruction than the PROPELLER technique, which was based on a rotating blade in k-space. On the other hand, the IDEAL technique was more useful for reducing the susceptibility artifacts. The use of a proper technique in clinical trials based on these results is expected to provide better clinical information for imaging diagnoses.     

Phase-sensitive fat suppression steady-state free procession sequence with phase correction

Robust and fast fat suppression is a challenge in balanced steady-state free precession （SSFP） magnetic resonance imaging. Although single-acquisition phase-sensitive SSFP can provide fat-suppressed images in short scan time, phase errors, especially spatially-dependent phase shift, caused by a variety of factors may result in misplacement of fat and water voxels. In this paper, a novel phase correction algorithm was used to calibrate those phase errors during image reconstruction. This algorithm corrects phase by region growing, employing both the magnitude and the phase information of image pixels. Phantom and in vivo imagings were performed to validate the technique. As a result, excellent fat-suppressed images were acquired by using single-acquisition phase-sensitive SSFP with phase correction.     

Phase-sensitive fat suppression steady-state free procession sequence with phase correction

Robust and fast fat suppression is a challenge in balanced steady-state free precession (SSFP) magnetic resonance imaging. Although single-acquisition phase-sensitive SSFP can provide fat-suppressed images in short scan time, phase errors, especially spatially-dependent phase shift, caused by a variety of factors may result in misplacement of fat and water voxels. In this paper, a novel phase correction algorithm was used to calibrate those phase errors during image reconstruction. This algorithm corrects phase by region growing, employing both the magnitude and the phase information of image pixels. Phantom and \textit{in vivo} imagings were performed to validate the technique. As a result, excellent fat-suppressed images were acquired by using single-acquisition phase-sensitive SSFP with phase correction.     

小波包熵的复杂体系近红外光谱信息提取

有用信息提取是复杂体系近红外检测的重点和难点之一。由于复杂体系光谱中存在各种噪声、基线漂移、谱带重叠及复杂背景的干扰,常规方法不能准确地从光谱中获得有用信息。为此,将小波包变换(DWPT)和信息熵理论相结合--小波包熵(EWPIE)提取复杂体系光谱中的有用信息。思路是采用小波包变换对光谱信号进行多频带分解,根据有用信号与噪声的频带分布特点,基于信息熵理论滤除干扰的频率分量,采用正交校正法(OSC)剔除与被测组分无关的信息,然后对处理后的频率分量进行重构,从而实现复杂体系有用信息的准确提取。通过对复杂体系光谱数据建立多元校正模型来验证该方法的效果。采用牛奶的近红外光谱数据,以牛奶中脂肪和蛋白质浓度为研究对象,建立了偏最小二乘法(PLS)模型。结果显示,牛奶中脂肪和蛋白质的预测均方根误差(RMSEP)分别为0.132%和0.121%,与单纯的DWPT和OSC相比,EWPIE能够有效地提取有用信息,避免了无用信息的干扰,明显提高了模型的预测精度,对复杂体系的准确检测具有一定的理论意义和实际应用价值。     

Dynamic view-dependent visualization of unstructured tetrahedral volumetric meshes

Abstract  

Visualization of large volumetric datasets has always been an important problem. Due to the high computational requirements of volume-rendering techniques, achieving interactive rates is a real challenge. We present a selective refinement scheme that dynamically refines the mesh according to the camera parameters. This scheme automatically determines the impact of different parts of the mesh on the output image and refines the mesh accordingly, without needing any user input. The view-dependent refinement scheme uses a progressive mesh representation that is based on an edge collapse-based tetrahedral mesh simplification algorithm. We tested our view-dependent refinement framework on an existing state-of-the-art volume renderer. Thanks to low overhead dynamic view-dependent refinement, we achieve interactive frame rates for rendering common datasets at decent image resolutions.     

面向地质应用的航空高光谱CASI‐SASI数据大气校正方法对比研究

蚀变信息提取是高光谱遥感地质应用的重要内容。基于特殊吸收峰的蚀变矿物提取是蚀变信息提取的重要手段。由于大气的吸收和散射作用,为了获得更为真实的地物反射光谱,必须进行大气校正。目前,国内外针对大气校正的对比研究主要集中在大气校正前后图像的质量改善、地物分类效果的提升以及校正图像像元光谱与实际地物光谱的相关关系等方面,而对不同校正方法获得的像元光谱与实际光谱吸收峰位的对应情况则很少讨论,这对于依赖吸收峰特征进行蚀变矿物提取的地质遥感极为不利。利用CASI-SASI航空高光谱成像系统,采集了甘肃龙首山地区的航空高光谱遥感数据,并运用ASD光谱仪,对该地区实际地物光谱进行了测量。以此为基础,开展了FLAASH、快速大气校正（QUAC）、经验线（EMPL）等方法大气校正结果的对比研究。通过对比分析,发现FLAASH,QUAC和EMPL均能在一定程度上消除大气的影响,改善航空高光谱遥感的图像质量,但EMPL方法得到的反射率与实际反射率相关性最好。此外,运用人工目视方法开展了实际地物反射光谱的吸收峰位与不同校正方法得到的对应像元反射光谱的吸收峰位的对比研究,发现不同校正方法得到的像元光谱的吸收峰位与实际峰位均存在不同程度的差异,虽然EMPL对吸收峰位的保留效果最好,但依然有“漏峰”的现象。据此,提出运用多种大气校正方法开展综合研究,以提高不同类型的蚀变带定位准确度。     

Image stack alignment in full‐field X‐ray absorption spectroscopy using SIFT_PyOCL

Full‐field X‐ray absorption spectroscopy experiments allow the acquisition of millions of spectra within minutes. However, the construction of the hyperspectral image requires an image alignment procedure with sub‐pixel precision. While the image correlation algorithm has originally been used for image re‐alignment using translations, the Scale Invariant Feature Transform (SIFT) algorithm (which is by design robust versus rotation, illumination change, translation and scaling) presents an additional advantage: the alignment can be limited to a region of interest of any arbitrary shape. In this context, a Python module, named SIFT_PyOCL, has been developed. It implements a parallel version of the SIFT algorithm in OpenCL, providing high‐speed image registration and alignment both on processors and graphics cards. The performance of the algorithm allows online processing of large datasets.     

Dual-acquisition phase-sensitive fat-water separation using balanced steady-state free precession

Balanced steady-state free precession (SSFP) sequences use fully re-focussed gradient waveforms to achieve a high signal and useful image contrast in short scan times. Despite these strengths, the clinical feasibility of balanced SSFP is still limited both by bright fat signal and by the signal voids that result from off-resonance effects such as field or susceptibility variations. A new method, dual-acquisition phase-sensitive SSFP, combines the signals from two standard balanced SSFP acquisitions to separate fat and water while simultaneously reducing the signal voids. The acquisitions are added in quadrature and then phase corrected using a simple algorithm before fat and water can be identified simply by the sign of the signal. This method is especially useful for applications at high field, where the RF power deposition, spatial resolution requirements and gradient strength limit the minimum repetition times. Finally, dual-acquisition phase-sensitive SSFP can be combined with other magnetization preparation schemes to produce specific image contrast in addition to separating fat and water signals.     

基于空间特征的光谱分类算法研究

随着成像光谱仪器的广泛应用,利用数据立方体进行物质分类与识别成为一项重要的研究内容,分类算法对最终的目标识别准确度与精度具有很大的决定作用。目前常见的分类算法主要利用了光谱维信息,从光谱匹配的角度进行物质分类。由于仪器探测的物质反射光谱不仅反映了物质种类,还与物质表面的几何结构,表面粗糙度等有关,因此仅仅利用物质的反射光谱进行物质分类识别具有一定的误差。该文在利用可见光反射光谱进行分类的基础上,结合图像空间特征,对分类过程进行控制,达到提高分类准确度的目的。利用该分类算法进行真假叶片识别,结果表明其具有较好的空间连续性,很大程度上克服了"麻点"效应,验证了算法的有效性。     

Radiography registration for mosaic tomography

A hybrid method of stitching X‐ray computed tomography (CT) datasets is proposed and the feasibility to apply the scheme in a synchrotron tomography beamline with micrometre resolution is shown. The proposed method enables the field of view of the system to be extended while spatial resolution and experimental setup remain unchanged. The approach relies on taking full tomographic datasets at different positions in a mosaic array and registering the frames using Fourier phase correlation and a residue‐based correlation. To ensure correlation correctness, the limits for the shifts are determined from the experimental motor position readouts. The masked correlation image is then minimized to obtain the correct shift. The partial datasets are blended in the sinogram space to be compatible with common CT reconstructors. The feasibility to use the algorithm to blend the partial datasets in projection space is also shown, creating a new single dataset, and standard reconstruction algorithms are used to restore high‐resolution slices even with a small number of projections.     

基于光谱分类的端元提取算法研究

目前成熟的端元提取算法是基于单形体几何学的像元纯度指数(PPO)算法,N-FINDR,VCA等算法.这些算法从图像所有像元中提取纯光谱,具有提取速度慢、精度不高的缺点;部分算法需要进行光谱降维,不利于小目标信息的提取.该文提出先利用基于空间特征的光谱分类算法进行分类,将格个图像划分成空间相邻、光谱相似的若干类,每一类的...     

基于多光谱图像融合的掌纹识别方法

生物特征识别在信息安全领域发挥着重要作用,掌纹识别作为一种新型生物特征识别方式,具有低失真、非侵入性和高唯一性等优势。传统掌纹研究大多使用自然光成像系统以灰度格式获取,识别精度很难进一步提升。为了获得更多的身份鉴别信息,提出利用多光谱掌纹图像代替自然光掌纹图像。针对现有掌纹识别算法由于没有考虑到不同光谱的特性而导致纹理细节丢失,识别精准率低的问题,提出了一种基于多光谱图像融合的掌纹识别算法。该方法通过对不同光谱下的掌纹图像进行快速自适应二维经验模式分解（FABEMD）,将多光谱掌纹图像分解成一系列频率由高到低的二维固有模态函数(BIMF)和一个残余分量,残余分量可被视为该光谱图像低频信息的初步估计。图像采集过程中光照条件很难保持稳定,而近红外光谱图像在进行FABEMD分解时对光照变换敏感,容易导致分解后的BIMF背景信息过于冗余;因此对分解后的近红外掌纹图像进行背景重建及特征细化,在对背景冗余信息进行平滑处理的同时可以有效增强高频信息的特征表达。为避免直接融合处理后引发的图像过度曝光问题,提出对近红外特征压缩后再融合。此外,提出了一种结合了注意力机制的改进残差网络（IRCANet）,用于融合后的掌纹图像分类,在网络中引入分阶段残差结构,缓解了网络的退化问题,在学习过程中有效地减少信息丢失,对于融合后的多光谱掌纹图像,分阶段残差结构能够稳定地将图像信息在网络间传输,但对图像中的高低频信息区分效果不够显著,为了使网络关注更多区分性特征,利用特征通道间的相互依赖性,在分阶段残差结构中结合了通道注意力(Channel Attention)机制。最终,在香港理工大学（PolyU）多光谱掌纹数据集上进行的综合实验表明,该方法可以取得良好的效果,算法识别准确率能达到99.67%且具有良好的实时性。     

Error-compensating algorithms in phase-shifting interferometry: a comparison by error analysis

In this paper, two families of phase-shifting algorithms with π/2 phase steps are studied. In family I, three new algorithms are derived by using the averaging technique based on the Surrel six-sample algorithm with phase shifts of π/2. Family II includes four well-known algorithms derived by the averaging technique based on the conventional four-sample algorithm with π/2 phase steps. A polynomial model of phase-shift errors used to describe general expressions for calculation of the correct object phase via the Fourier spectra analysing method as a function of the harmonic order in the fringe signal is presented. The error-compensating properties of the algorithms in families I and II are investigated by the Fourier spectra analysing method. It is found that the averaging technique, when used in any of the algorithm with π/2 phase steps, can improve the phase-shifting algorithm property: it is insensitive to phase-shift error when the fringe signal contains the first harmonic, but it can't be used to enhance the phase-shifting algorithm properties when the fringe signal contains higher order harmonics (n2). P–V (peak–valley) phase errors are calculated by the computer simulation and tables and plots are presented, from which the algorithms in families I and II are compared. It is shown that the algorithms in family I are more insensitive to phase-shift errors when the fringe signal contains the second harmonic and the algorithms in family II are more insensitive to phase-shift errors when the fringe signal is a sinusoidal waveform.     

基于赝热光照明的单发光学散斑成像

散射介质对光的散射是当前限制光学成像深度或距离的一个严重的问题.本文首先数值模拟比较了光透过随机散射介质成像研究中常用的基于光学记忆效应(memory effect, ME)和自相关(autocorrelation, AC)方法的HIOER算法和乒乓(Ping-Pang, PP)算法的优缺点.通过对HIOER算法和PP算法的恢复效果和迭代次数进行比较,发现PP算法在保持较高恢复效果的前提下拥有更快的运行速度.实验中,利用连续HeNe激光器和旋转毛玻璃产生赝热光源,通过物镜对随机散射介质后数毫米距离内的不同形状物体进行了单帧成像,并采用PP算法成功地恢复出微米量级物体的实际图像.这一研究结果将进一步促进ME和AC方法在深层生物组织医学成像研究上的应用.最后,实验研究了不同的物镜和散射介质的间距对成像恢复的放大率、分辨率和图像强度的影响特性,并进行了详细研究.     

Synthesis,characterization and magnetic properties of MFe2O4 (M=Co,Mg, Mn,Ni) nanoparticles using ricin oil as capping agent

We focused on obtaining MFe₂O₄ nanoparticles using ricin oil solution as surfactant and on their structural characterization and magnetic properties. The annealed samples at 500 °C in air for 6 h were analyzed for the crystal phase identification by powder X-ray diffraction using CuKα radiation. The particle size, the chemical composition and the morphology of the calcinated powders were characterized by scanning electron microscopy. All sintered samples contain only one phase, which has a cubic structure with crystallite sizes of 12–21 nm. From the infrared spectra of all samples were observed two strong bands around 600 and 400 cm⁻¹, which correspond to the intrinsic lattice vibrations of octahedral and tetrahedral sites of the spinel structure, respectively, and characteristic vibration for capping agent. The magnetic properties of fine powders were investigated at room temperature by using a vibrating sample magnetometer. The room temperature M–H hysteresis loops show ferromagnetic behavior of the calcined samples, with specific saturation magnetization (M_s) values ranging between 11 and 53 emu/g.     

Multi-exponential relaxation analysis with MR imaging and NMR spectroscopy using fat-water systems

This study was undertaken to evaluate the feasibility of multiexponential relaxation data analysis to MR imaging techniques. The first part of this study contains accurate relaxation time measurements performed on a conventional spectrometer. In the second part, essentially the same measuring techniques were applied but now on standard whole body MR imaging equipment. T2 relaxation was measured using multi-echo techniques, T1 relaxation using multiple inversion recovery measurements. Manganese chloride solutions were used for verification of the single exponential model. Water and fat mixtures were considered for multi-exponentiality. Pure fat showed an intrinsic two-exponentiality in T1 and T2 relaxation. Mixtures of fat and water were analyzed and could at best be characterized by two exponentials, although at least three exponentials were known to be present. From the two-exponential fit the relative amounts of fat and water were calculated and compared with the mixture composition. Statistical criteria are discussed to discriminate between single and double exponential behavior in relaxation curves. It is concluded that the time consuming IR measurements for the determination of multiple T1 relaxation are not applicable in a clinical environment. Multiple T2 relaxation can be determined in a reasonable amount of time using multiple echo measurements in one image acquisition. It is shown that the observed values of T1 and T2 from tissues with intrinsic multiexponential relaxation behavior, measured with MR imaging or MR relaxation techniques on a whole-body imager or a conventional spectrometer, depend strongly on the way the experiments are set up and on the model accepted for data analysis.