A deep-learning-based scatter correction with water equivalent path length map for digital radiography.

We proposed a new deep learning (DL) model for accurate scatter correction in digital radiography. The proposed network featured a pixel-wise water equivalent path length (WEPL) map of subjects with diverse sizes and 3D inner structures. The proposed U-Net model comprises two concatenated modules: one for generating a WEPL map and the other for predicting scatter using the WEPL map as auxiliary information. First, 3D CT images were used as numerical phantoms for training and validation, generating observed and scattered images by Monte Carlo simulation, and WEPL maps using Siddon's algorithm. Then, we optimised the model without overfitting. Next, we validated the proposed model's performance by comparing it with other DL models. The proposed model obtained scatter-corrected images with a peak signal-to-noise ratio of 44.24 ± 2.89 dB and a structural similarity index measure of 0.9987 ± 0.0004, which were higher than other DL models. Finally, scatter fractions (SFs) were compared with other DL models using an actual phantom to confirm practicality. Among DL models, the proposed model showed the smallest deviation from measured SF values. Furthermore, using an actual radiograph containing an acrylic object, the contrast-to-noise ratio (CNR) of the proposed model and the anti-scatter grid were compared. The CNR of the images corrected using the proposed model are 16% and 82% higher than those of the raw and grid-applied images, respectively. The advantage of the proposed method is that no actual radiography system is required for collecting training dataset, as the dataset is created from CT images using Monte Carlo simulation.

Crystal optics simulations for delineation of the three-dimensional cellular nuclear distribution using analyzer-based refraction-contrast computed tomography.

Refraction-contrast computed tomography (RCT) using a refractive angle analyzer of Si perfect crystal can reconstruct the three-dimensional structure of biological soft tissue with contrast comparable to that of stained two-dimensional pathological images. However, the blurring of X-ray beam by the analyzer has prevented improvement of the spatial resolution of RCT, and the currently possible observation of tissue structure at a scale of approximately 20 µm provides only limited medical information. As in pathology, to differentiate between benign and malignant forms of cancer, it is necessary to observe the distribution of the cell nucleus, which is approximately 5-10 µm in diameter. In this study, based on the X-ray dynamical diffraction theory using the Takagi-Taupin equation, which calculates the propagation of X-ray energy in crystals, an analyzer crystal optical system depicting the distribution of cell nuclei was investigated by RCT imaging simulation experiments in terms of the thickness of the Laue-case analyzer, the camera pixel size and the difference in spatial resolution between the Bragg-case and Laue-case analyzers.

Ring artifact removal for differential phase-contrast X-ray computed tomography using a conditional generative adversarial network.

PURPOSE: The integration process used as a pre-processing step in the reconstruction of differential phase-contrast X-ray CT (d-PCCT) causes the measurement noise to propagate throughout the projection image, which is leading to increased ring artifacts (RA) in the reconstructed image. It is difficult to eliminate the RA using conventional RA removal methods that were developed for the absorption-based CT field. We propose an effective method that can remove RA of d-PCCT images. METHODS: The proposed method uses Laplacian images reconstructed from second-derivative projections of d-PCCT. This method is based on a conditional generative adversarial network (cGAN), whose loss function is designed by adding the L1- and L2-norm to the original cGAN. The training data were taken from a numerical phantom generated by a d-PCCT imaging simulator. To validate the applicability of the trained network, we tested its RA removal effect on test data from numerical phantoms generated randomly and actual experimental data. RESULTS: The results of numerical validation using numerical phantoms showed that the proposed method improved the RA removal effect compared to conventional methods. In addition, image comparison by visual evaluation showed that only the proposed method was able to remove RA while preserving original structures in the actual biological d-PCCT images. CONCLUSION: We proposed a cGAN-based method for RA removal that exploits the physical properties of d-PCCT. The proposed method was able to completely remove RA from d-PCCT images on both simulated data and biological data. We believe that this method is useful for the observation of various types of biological soft tissue.

Usefulness of X-ray dark-field imaging in the evaluation of local recurrence after nipple-sparing mastectomy.

PURPOSE: Our previous study suggests that the cross-sectional morphology of ducts and branching of ducts in the nipple are associated with the presence of breast cancer. In this study, we evaluated whether cross-sectional morphology and duct branching of human nipple obtained by X-ray dark-field imaging tomographic technique (XDFI-CT) could predict the likelihood of the presence of intraductal cancer into the nipple. METHODS: A total of 51 nipple specimens were obtained from consecutive total mastectomies performed for breast cancer in Nagoya Medical Center. After reconstructing 3D images of the nipple using XDFI-CT, the cross-sectional images and the 3D arrangement of ducts were extracted. These cross-sectional images of ducts were classified into four patterns based on the status of the lumen without being informed of pathology results. RESULTS: Of the four patterns, the distended ducts with heterogenous content were highly correlated with the presence of ductal carcinoma in situ confirmed by histopathology. The total number of orifices identified in the 51 specimens was 1298, and 182 (14%) at the tip and 19 (1.5%) at least 5 mm depth from the tip were composed of two or more ducts. CONCLUSIONS: Anatomy of nipple ducts is essential to evaluate risk of local recurrence after nipple-sparing mastectomy because cancerous spread occurs within the duct of the same segment of the mammary duct-lobular system in the in situ stage. The 3D microscale anatomy of nipple ducts revealed by XDFI-CT provides useful information to assess the risk of breast cancer involvement at the preserved portion in nipple-sparing mastectomy.

X-ray Dark-Field Imaging (XDFI)-a Promising Tool for 3D Virtual Histopathology.

X-ray dark-field imaging (XDFI) utilizing a thin silicon crystal under Laue case enables visualizing three-dimensional (3D) morphological alterations of human tissue. XDFI uses refraction-contrast derived from phase shift rather than absorption as the main X-ray image contrast source to render 2D and 3D images of tissue specimens in unprecedented detail. The unique features of XDFI are its extremely high sensitivity (approximately 1000:1 compared to absorption for soft tissues under X-ray energy of around 20 keV, theoretically) and excellent resolution (8.5 µm) without requiring contrast medium or staining. Thus, XDFI-computed tomography can generate 3D virtual histological images equivalent to those of stained histological sections pathologists observe under low-power light microscopy as far as organs and tissues selected as samples in preliminary studies. This paper reviews the fundamental principles and the potential of XDFI, describes two optical setups for XDFI with examples, illustrates features of XDFI that are salient for histopathology, and presents XDFI examples of refraction-contrast images of atherosclerotic plaques, musculoskeletal tissue, neuronal tissue, and breast cancer specimens. Availability of this X-ray imaging in routine histopathological evaluations of tissue specimens would help guide clinical decision making by highlighting suspicious areas in unstained, thick sections for further sampling and analysis using conventional histopathological techniques. XDFI is a promising tool for 3D virtual histopathology.

X-ray dark-field phase-contrast imaging: Origins of the concept to practical implementation and applications.

The basic idea of X-ray dark-field imaging (XDFI), first presented in 2000, was based on the concepts used in an X-ray interferometer. In this article, we review 20 years of developments in our theoretical understanding, scientific instrumentation, and experimental demonstration of XDFI and its applications to medical imaging. We first describe the concepts underlying XDFI that are responsible for imparting phase contrast information in projection X-ray images. We then review the algorithms that can convert these projection phase images into three-dimensional tomographic slices. Various implementations of computed tomography reconstructions algorithms for XDFI data are discussed. The next four sections describe and illustrate potential applications of XDFI in pathology, musculoskeletal imaging, oncologic imaging, and neuroimaging. The sample applications that are presented illustrate potential use scenarios for XDFI in histopathology and other clinical applications. Finally, the last section presents future perspectives and potential technical developments that can make XDFI an even more powerful tool.

Improving contrast and spatial resolution in crystal analyzer-based x-ray dark-field imaging: Theoretical considerations and experimental demonstration.

PURPOSE: This paper describes and experimentally validates a methodology for improving contrast and spatial resolution of the x-ray dark-field imaging (XDFI) by cutting the monochromator-collimator asymmetrically and thinning the Laue angle analyzer. METHODS: We measure the spatial resolution of our XDFI setup using a test object consisting of wolfram tungsten meshes and compare it with the theoretical prediction. Using x-ray dark-field computed tomography of breast cancer specimens (lobular carcinoma in situ), we demonstrate that the resolution of XDFI is sufficient for histopathologic analysis. RESULTS: Our experimental results show that the overall spatial resolution of XDFI can be improved by approximately a factor of 2 when these modifications are implemented. The reconstructed images of breast cancer specimens provide sufficient details for radiologic histopathology. CONCLUSIONS: By cutting the monochromator-collimator and thinning the Laue angle analyzer, XDFI can achieve the resolution sufficient for radiologic histopathology.

Improvement of the depth resolution of swept-source THz-OCT for non-destructive inspection.

We construct a terahertz swept source optical coherence tomography system using a continuous-wave diode multiplier source in the 600-GHz band for defect inspection in multilayer objects and evaluate its performance. Using this system, we image a multilayer plastic sample to demonstrate the effectiveness of nondestructive three-dimensional imaging. To enhance the depth resolution, we apply an annihilating filter to the analysis and confirm that two surfaces of a 1-mm-thick plastic plate can be resolved. In addition, the repeatability of measured thicknesses is 0.22 mm. These values are approximately one-half and one-tenth of the resolution achievable by conventional Fourier analysis, respectively.

Cardiac fiber tracking on super high-resolution CT images: a comparative study.

Purpose: High-resolution cardiac imaging and fiber analysis methods are required to understand cardiac anatomy. Although refraction-contrast x-ray CT (RCT) has high soft tissue contrast, it cannot be commonly used because it requires a synchrotron system. Microfocus x-ray CT ( µ CT ) is another commercially available imaging modality. Approach: We evaluate the usefulness of µ CT for analyzing fibers by quantitatively and objectively comparing the results with RCT. To do so, we scanned a rabbit heart by both modalities with our original protocol of prepared materials and compared their image-based analysis results, including fiber orientation estimation and fiber tracking. Results: Fiber orientations estimated by two modalities were closely resembled under the correlation coefficient of 0.63. Tracked fibers from both modalities matched well the anatomical knowledge that fiber orientations are different inside and outside of the left ventricle. However, the µ CT volume caused incorrect tracking around the boundaries caused by stitching scanning. Conclusions: Our experimental results demonstrated that µ CT scanning can be used for cardiac fiber analysis, although further investigation is required in the differences of fiber analysis results on RCT and µ CT .

Three-dimensional microanatomy of human nipple visualized by X-ray dark-field computed tomography.

PURPOSE: The three-dimensional (3D) structure of the human nipple has not been fully clarified. However, its importance has increased in recent years because it has become common practice to preoperatively explore the spread of breast cancer to the nipple with needle biopsy, ductoscopy, and/or ductal lavage for nipple-sparing mastectomy. Here, we demonstrated that X-ray dark-field computed tomography (XDFI-CT) is a powerful tool for reconstructing the 3D distribution pattern of human lactiferous ducts non-destructively, without contrast agent, and with high tissue contrast. METHODS: Nipples amputated from mastectomy specimens of 51 patients with breast cancer were visualized three-dimensionally by XDFI-CT. First, CT images and conventionally stained tissue sections were compared to demonstrate that XDFI-CT provides 3D anatomical information. Next, the number of ducts in the nipple and the number of ducts sharing an ostium near the tip of the nipple were measured from the volume set of XDFI-CT. Finally, the 3D distribution pattern of the ducts was determined. RESULTS: XDFI-CT can provide images almost equivalent to those of low-magnification light microscopy of conventional hematoxylin-eosin-stained histological sections. The mean number of ducts in all cases was 28.0. The total number of ducts sharing an ostium near the tip of the nipple was 525 of 1428. The 3D distribution patterns of the ducts were classified into three types that we defined as convergent (22%), straight (39%), or divergent (39%). CONCLUSIONS: XDFI-CT is useful for exploring the microanatomy of the human nipple and might be used for non-invasive nipple diagnosis in the future.

Imaging with ultra-small-angle X-ray scattering using a Laue-case analyzer and its application to human breast tumors.

PURPOSE: In this study, we demonstrate a novel imaging technique, based on ultra-small-angle X-ray scattering (USAXS) that uses a Laue-case Si wafer as the angle analyzer. METHODS: We utilized the (1 1 1) diffraction plane of a 356 µm thick, symmetrically cut Si wafer as the angle analyzer, denoted by A[L]. With this device, we performed USAXS imaging experiments using 19.8 keV synchrotron X-rays. The objects we imaged were formalin-fixed, paraffin-embedded breast tumors (an invasive carcinoma and an intraductal papilloma). During image acquisition by a charge-coupled device (CCD) camera, we varied the rotation angle of the analyzer in 0.02″ steps from -2.40″ to +2.40″ around the Bragg angle. The exposure time for each image was 2 s. We determined the amount of ultra-small-angle X-ray scattering from the width of the intensity curve obtained for each local pixel during the rotation of the analyzer. RESULTS: We acquired USAXS images of malignant and benign breast tumor specimens using the A[L] analyzer; regions with larger USAXS form brighter areas in the image. We varied the sensitivity of the USAXS image by changing the threshold level of the object rocking curve. CONCLUSIONS: The USAXS images can provide information about the internal distribution of closely packed scattering bodies in a sample with reasonable sensitivity. This information differs from that obtainable through refraction-contrast imaging. Although further validation studies will be necessary, we conclude that USAXS imaging using a Laue-case analyzer may have significant potential as a new diagnosis technique.

Evaluation of retinal nerve fiber layer defect using wide-field en-face swept-source OCT images by applying the inner limiting membrane flattening.

PURPOSE: The assessment of retinal nerve fiber layer defects (RNFLDs) is a useful part of glaucoma care. Here, we obtained en-face images of retinal layers below the inner limiting membrane (ILM) with swept source-optical coherence tomography (SS-OCT), and measured RNFLD angle with new software. METHODS: This study included 105 eyes of 105 normal tension glaucoma (NTG) patients (age, 59.8 ± 13.2). Exclusion criteria were best-corrected visual acuity < 0.5, axial length > 28 mm, non-glaucoma ocular disease, and systemic disease affecting the visual field. We obtained 12 x 9 mm 3D volume scans centered on the macula with SS-OCT (DRI OCT-1, Topcon), and from these scans, created 3 averaged en-face images, each comprising 7 horizontal en-face images (total thickness: 18.2 µm). We labeled these averaged images, according to their depth below the ILM, as en-face images 1 (shallowest), 2 (middle) and 3 (deepest). In each image, a circle was drawn centered on the disc, with a radius of half the distance between the centers of the disc and macula. The investigator marked points where the edge of the RNFLD intersected this circle, and RNFLD angle (RNFLDA) was calculated with new software. Finally, we analyzed the association between RNFLDA, cpRNFLT, weighted RGC count (wrgc) and Humphrey field analyzer (HFA)-measured mean deviation (MD) and hemifield total deviation (TD), both overall and in each hemifield. RESULTS: En-face image 2 had the highest interclass reproducibility for measuring RNFLDA (intra-rater intraclass correlation coefficient (ICC): 0.988, inter-rater ICC: 0.962). The correlation coefficients with RNFLDA were: HFA MD, -0.60; superior TD, -0.73; inferior TD, -0.69; overall cpRNFLT, -0.27; superior hemifield cpRNFLT, -0.39; and inferior hemifield cpRNFLT, -0.53 (all p<0.001). CONCLUSIONS: RNFLDA measured in SS-OCT images had high reproducibility and was correlated to glaucoma severity. Our new method may be a valuable future part of glaucoma care.

Multi-pinhole fluorescent x-ray computed tomography for molecular imaging.

We propose a multi-pinhole fluorescent x-ray computed tomography (mp-FXCT) technique for preclinical molecular imaging that can provide the complete data necessary to produce 3-D tomographic images during anaesthesia. In this method, multiple projections are simultaneously acquired through a multi-pinhole collimator with a 2-D detector and full-field volumetric beam to accelerate the data acquisition process and enhance the signal-to-noise ratios of the projections. We constructed a 15-pinhole mp-FXCT imaging system at beamline ARNE-7A at KEK and performed preliminary experiments to investigate its imaging properties using physical phantoms and a non-radioactive I imaging agent. The mp-FXCT system could detect an I concentration of 0.038 mg/ml, the minimum required for in-vivo imaging, at a spatial resolution of about 0.3 mm during a data acquisition time of 90 min, which is less than the time for which anaesthesia is effective and suggests that preclinical molecular imaging is feasible with mp-FXCT.

Dual-energy fluorescent x-ray computed tomography system with a pinhole design: Use of K-edge discontinuity for scatter correction.

We propose a pinhole-based fluorescent x-ray computed tomography (p-FXCT) system with a 2-D detector and volumetric beam that can suppress the quality deterioration caused by scatter components. In the corresponding p-FXCT technique, projections are acquired at individual incident energies just above and below the K-edge of the imaged trace element; then, reconstruction is performed based on the two sets of projections using a maximum likelihood expectation maximization algorithm that incorporates the scatter components. We constructed a p-FXCT imaging system and performed a preliminary experiment using a physical phantom and an I imaging agent. The proposed dual-energy p-FXCT improved the contrast-to-noise ratio by a factor of more than 2.5 compared to that attainable using mono-energetic p-FXCT for a 0.3 mg/ml I solution. We also imaged an excised rat's liver infused with a Ba contrast agent to demonstrate the feasibility of imaging a biological sample.

Dark-Field Imaging: Recent developments and potential clinical applications.

This paper describes an X-ray phase contrast imaging technique using analyzer-based optics called X-ray Dark-Field Imaging that has been under development for the past 10years. We describe the theory behind XDFI, the X-ray optics required for implementing it in practice, and algorithms used for 2D, 2.5D, and 3D image reconstruction. The XDFI optical chain consists of an asymmetrically cut, Bragg-type monochromator-collimator that provides a planar monochromatic X-ray beam, a positioning stage for the specimens, a Laue-case angle analyzer, and one or two cameras to capture the dark and bright field images. We demonstrate the soft-tissue discrimination capabilities of XDFI by reconstructing images with absorption and phase contrast. By using a variety of specimens such as breast tissue with cancer, joints with articular cartilage, ex-vivo human eye specimen, and others, we show that refraction-based contrast derived from XDFI is more effective in characterizing anatomical features, articular pathology, and neoplastic disease than conventional absorption-based images. For example, XDFI of breast tissue can discriminate between the normal and diseased terminal duct lobular unit, and between invasive and in-situ cancer. The final section of this paper is devoted to potential future developments to enable clinical and histo-pathological applications of this technique.

The feasibility of 11C-PIB-PET/CT for amyloid plaque burden: validation of the effectiveness of CT-based partial volume correction.

INTRODUCTION: Amyloid positron-emission tomography (PET) imaging with 11C-Pittsburgh compound B (PiB) is an effective tool for assessing brain amyloid deposits. PET imaging, however, can suffer from the partial volume effect (PVE). PVE has been corrected using MRI (magnetic resonance imaging) image data. However, correction of the PVE of PET using MRI usually requires two separate procedures, imposing a burden on patients and leading to low throughput and inefficient diagnoses. The advent of PET/computed tomography (PET/CT) may potentially overcome these problems and offer higher throughput and reliable quantification of amyloid plaques and assessment of Alzheimer disease (AD). METHODS: We investigated the feasibility of correcting PVE in amyloid PET using CT, obtained by PET/CT, instead of MRI. We demonstrated the efficacy of partial volume correction (PVC) based on CT by comparing the results of CT-based PVC and those of MRI-based PVC using images acquired from AD patients and controls. RESULTS: Both methods were able to perform PVC. Slight but significant differences between standard uptake volume ratio (SUVR) values were noted between the two modalities; these were attenuated by constant multiplication. CONCLUSION: CT will potentially replace MRI for PVC, allowing the use of a single PET/CT scanner for amyloid plaque quantitation.

Non-invasive quality evaluation of confluent cells by image-based orientation heterogeneity analysis.

In recent years, cell and tissue therapy in regenerative medicine have advanced rapidly towards commercialization. However, conventional invasive cell quality assessment is incompatible with direct evaluation of the cells produced for such therapies, especially in the case of regenerative medicine products. Our group has demonstrated the potential of quantitative assessment of cell quality, using information obtained from cell images, for non-invasive real-time evaluation of regenerative medicine products. However, image of cells in the confluent state are often difficult to evaluate, because accurate recognition of cells is technically difficult and the morphological features of confluent cells are non-characteristic. To overcome these challenges, we developed a new image-processing algorithm, heterogeneity of orientation (H-Orient) processing, to describe the heterogeneous density of cells in the confluent state. In this algorithm, we introduced a Hessian calculation that converts pixel intensity data to orientation data and a statistical profiling calculation that evaluates the heterogeneity of orientations within an image, generating novel parameters that yield a quantitative profile of an image. Using such parameters, we tested the algorithm's performance in discriminating different qualities of cellular images with three types of clinically important cell quality check (QC) models: remaining lifespan check (QC1), manipulation error check (QC2), and differentiation potential check (QC3). Our results show that our orientation analysis algorithm could predict with high accuracy the outcomes of all types of cellular quality checks (>84% average accuracy with cross-validation).

In Vitro Validation of an Artefact Suppression Algorithm in X-Ray Phase-Contrast Computed Tomography.

X-ray phase-contrast tomography can significantly increase the contrast-resolution of conventional attenuation-contrast imaging, especially for soft-tissue structures that have very similar attenuation. Just as in attenuation-based tomography, phase contrast tomography requires a linear dependence of aggregate beam direction on the incremental direction alteration caused by individual voxels along the path of the X-ray beam. Dense objects such as calcifications in biological specimens violate this condition. There are extensive beam deflection artefacts in the vicinity of such structures because they result in large distortion of wave front due to the large difference of refractive index; for such large changes in beam direction, the transmittance of the silicon analyzer crystal saturates and is no longer linearly dependent on the angle of refraction. This paper describes a method by which these effects can be overcome and excellent soft-tissue contrast of phase tomography can be preserved in the vicinity of such artefact-producing structures.

Limited view reconstruction for differential phase-contrast computed tomography.

This paper describes an algebraic reconstruction algorithm that uses total variation (TV) regularization for differential phase contrast computed tomography (DPC-CT) using a limited number of views. In order to overcome over-flattening inherent in TV regularization, a two-step reconstruction process is used: we first reconstruct tomographic images of gradient refractive index from differential projections with TV regularization; these images are then used to compute tomographic images of refractive index by solving the Poisson equation. We incorporate TV regularization in the reconstruction process because the distribution of gradient refractive index is much more flattened than the refractive index. Simulations of the proposed method demonstrate that it can achieve satisfactory image quality from a much smaller number of projections than is required by the Nyquist sampling theorem. We experimentally prove the feasibility of the proposed method using dark field imaging optics at PF-14C beamline at the Photon Factory, KEK. The differential phase contrast projection data was experimentally acquired from a biological sample and DPC-CT images were reconstructed. We show that far fewer projections are needed when the proposed algorithm is used.

Image-based focused counting of dividing cells for non-invasive monitoring of regenerative medicine products.

Despite the growing numbers of successful applications in regenerative medicine, biotechnologies for evaluating the quality of cells remain limited. To evaluate the cultured cells non-invasively, image-based cellular assessment method holds great promise. However, although there are various image-processing algorithms, very few studies have focused to prove the effectiveness of phase contrast images with risk assessment example that reflects actual difficulties in regenerative medicine products. In this study, we developed a simple image-processing method to recognize the number of dividing cells in time-course phase-contrast microscopic images, and applied this method to assess the irregular proliferation behavior in normal cells. Practically, as a model, rapid proliferating human fibrosarcoma cells were mixed in normal human fibroblasts in the same culture dish, and their sarcoma existence was evaluated. As a result, the existence of sarcoma population in normal cell sample could be feasibly detected within earliest period of cell culture by their irregular rise of accumulated counts of dividing cells. Our image-processing technique also illustrates the technical effectiveness of combining intra-frame and inter-frame image processing to accurately count only the dividing cells. Our concept of focused counting of dividing cells shows a successful example of image-based analysis to quickly and non-invasively monitor the regular state of regenerative medicine products.