Guang-Hong Chen

Prior image constrained compressed sensing (PICCS): a method to accurately reconstruct dynamic CT images from highly undersampled projection data sets.

When the number of projections does not satisfy the Shannon/Nyquist sampling requirement, streaking artifacts are inevitable in x-ray computed tomography (CT) images reconstructed using filtered backprojection algorithms. In this letter, the spatial-temporal correlations in dynamic CT imaging have been exploited to sparsify dynamic CT image sequences and the newly proposed compressed sensing (CS) reconstruction method is applied to reconstruct the target image sequences. A prior image reconstructed from the union of interleaved dynamical data sets is utilized to constrain the CS image reconstruction for the individual time frames. This method is referred to as prior image constrained compressed sensing (PICCS). In vivo experimental animal studies were conducted to validate the PICCS algorithm, and the results indicate that PICCS enables accurate reconstruction of dynamic CT images using about 20 view angles, which corresponds to an under-sampling factor of 32. This undersampling factor implies a potential radiation dose reduction by a factor of 32 in myocardial CT perfusion imaging.

Abdominal CT With Model-Based Iterative Reconstruction (MBIR): Initial Results of a Prospective Trial Comparing Ultralow-Dose With Standard-Dose Imaging

OBJECTIVE The purpose of this study was to report preliminary results of an ongoing prospective trial of ultralow-dose abdominal MDCT. SUBJECTS AND METHODS Imaging with standard-dose contrast-enhanced (n = 21) and unenhanced (n = 24) clinical abdominal MDCT protocols was immediately followed by ultralow-dose imaging of a matched series of 45 consecutively registered adults (mean age, 57.9 years; mean body mass index, 28.5). The ultralow-dose images were reconstructed with filtered back projection (FBP), adaptive statistical iterative reconstruction (ASIR), and model-based iterative reconstruction (MBIR). Standard-dose series were reconstructed with FBP (reference standard). Image noise was measured at multiple predefined sites. Two blinded abdominal radiologists interpreted randomly presented ultralow-dose images for multilevel subjective image quality (5-point scale) and depiction of organ-based focal lesions. RESULTS Mean dose reduction relative to the standard series was 74% (median, 78%; range, 57-88%; mean effective dose, 1.90 mSv). Mean multiorgan image noise for low-dose MBIR was 14.7 ± 2.6 HU, significantly lower than standard-dose FBP (28.9 ± 9.9 HU), low-dose FBP (59.2 ± 23.3 HU), and ASIR (45.6 ± 14.1 HU) (p < 0.001). The mean subjective image quality score for low-dose MBIR (3.0 ± 0.5) was significantly higher than for low-dose FBP (1.6 ± 0.7) and ASIR (1.8 ± 0.7) (p < 0.001). Readers identified 213 focal noncalcific lesions with standard-dose FBP. Pooled lesion detection was higher for low-dose MBIR (79.3% [169/213]) compared with low-dose FBP (66.2% [141/213]) and ASIR (62.0% [132/213]) (p < 0.05). CONCLUSION MBIR shows great potential for substantially reducing radiation doses at routine abdominal CT. Both FBP and ASIR are limited in this regard owing to reduced image quality and diagnostic capability. Further investigation is needed to determine the optimal dose level for MBIR that maintains adequate diagnostic performance. In general, objective and subjective image quality measurements do not necessarily correlate with diagnostic performance at ultralow-dose CT.

Performance comparison between total variation (TV)-based compressed sensing and statistical iterative reconstruction algorithms

Of all available reconstruction methods, statistical iterative reconstruction algorithms appear particularly promising since they enable accurate physical noise modeling. The newly developed compressive sampling/compressed sensing (CS) algorithm has shown the potential to accurately reconstruct images from highly undersampled data. The CS algorithm can be implemented in the statistical reconstruction framework as well. In this study, we compared the performance of two standard statistical reconstruction algorithms (penalized weighted least squares and q-GGMRF) to the CS algorithm. In assessing the image quality using these iterative reconstructions, it is critical to utilize realistic background anatomy as the reconstruction results are object dependent. A cadaver head was scanned on a Varian Trilogy system at different dose levels. Several figures of merit including the relative root mean square error and a quality factor which accounts for the noise performance and the spatial resolution were introduced to objectively evaluate reconstruction performance. A comparison is presented between the three algorithms for a constant undersampling factor comparing different algorithms at several dose levels. To facilitate this comparison, the original CS method was formulated in the framework of the statistical image reconstruction algorithms. Important conclusions of the measurements from our studies are that (1) for realistic neuro-anatomy, over 100 projections are required to avoid streak artifacts in the reconstructed images even with CS reconstruction, (2) regardless of the algorithm employed, it is beneficial to distribute the total dose to more views as long as each view remains quantum noise limited and (3) the total variation-based CS method is not appropriate for very low dose levels because while it can mitigate streaking artifacts, the images exhibit patchy behavior, which is potentially harmful for medical diagnosis.

Fan-beam and cone-beam image reconstruction via filtering the backprojection image of differentiated projection data

In this paper, a new image reconstruction scheme is presented based on Tuys cone-beam inversion scheme and its fan-beam counterpart. It is demonstrated that Tuys inversion scheme may be used to derive a new framework for fanbeam and cone-beam image reconstruction. In this new framework, images are reconstructed via filtering the backprojection image of differentiated projection data. The new framework is mathematically exact and is applicable to a general source trajectory provided the Tuy data sufficiency condition is satisfied. By choosing a piece-wise constant function for one of the components in the factorized weighting function, the filtering kernel is one dimensional, viz. the filtering process is along a straight line. Thus, the derived image reconstruction algorithm is mathematically exact and efficient. In the cone-beam case, the derived reconstruction algorithm is applicable to a large class of source trajectories where the pi-lines or the generalized pi-lines exist. In addition, the new reconstruction scheme survives the super-short scan mode in both the fan-beam and cone-beam cases provided the data are not transversely truncated. Numerical simulations were conducted to validate the new reconstruction scheme for the fan-beam case.

Temporal resolution improvement using PICCS in MDCT cardiac imaging

The current paradigm for temporal resolution improvement is to add more source-detector units and/or increase the gantry rotation speed. The purpose of this article is to present an innovative alternative method to potentially improve temporal resolution by approximately a factor of 2 for all MDCT scanners without requiring hardware modification. The central enabling technology is a most recently developed image reconstruction method: Prior image constrained compressed sensing (PICCS). Using the method, cardiac CTimages can be accurately reconstructed using the projection data acquired in an angular range of about 120°, which is roughly 50% of the standard short-scan angular range ( ∼ 240 ° for an MDCT scanner). As a result, the temporal resolution of MDCT cardiac imaging can be universally improved by approximately a factor of 2. In order to validate the proposed method, two in vivo animal experiments were conducted using a state-of-the-art 64-slice CT scanner (GE Healthcare, Waukesha, WI) at different gantry rotation times and different heart rates. One animal was scanned at heart rate of 83 beats per minute (bpm) using 400 ms gantry rotation time and the second animal was scanned at 94 bpm using 350 ms gantry rotation time, respectively. Cardiac coronary CTimaging can be successfully performed at high heart rates using a single-source MDCT scanner and projection data from a single heart beat with gantry rotation times of 400 and 350 ms . Using the proposed PICCS method, the temporal resolution of cardiac CTimaging can be effectively improved by approximately a factor of 2 without modifying any scanner hardware. This potentially provides a new method for single-source MDCT scanners to achieve reliable coronary CTimaging for patients at higher heart rates than the current heart rate limit of 70 bpm without using the well-known multisegment FBP reconstruction algorithm. This method also enables dual-source MDCT scanner to achieve higher temporal resolution without further hardware modifications.

Prior image constrained compressed sensing: Implementation and performance evaluation

PURPOSE Prior image constrained compressed sensing (PICCS) is an image reconstruction framework which incorporates an often available prior image into the compressed sensing objective function. The images are reconstructed using an optimization procedure. In this paper, several alternative unconstrained minimization methods are used to implement PICCS. The purpose is to study and compare the performance of each implementation, as well as to evaluate the performance of the PICCS objective function with respect to image quality. METHODS Six different minimization methods are investigated with respect to convergence speed and reconstruction accuracy. These minimization methods include the steepest descent (SD) method and the conjugate gradient (CG) method. These algorithms require a line search to be performed. Thus, for each minimization algorithm, two line searching algorithms are evaluated: a backtracking (BT) line search and a fast Newton-Raphson (NR) line search. The relative root mean square error is used to evaluate the reconstruction accuracy. The algorithm that offers the best convergence speed is used to study the performance of PICCS with respect to the prior image parameter α and the data consistency parameter λ. PICCS is studied in terms of reconstruction accuracy, low-contrast spatial resolution, and noise characteristics. A numerical phantom was simulated and an animal model was scanned using a multirow detector computed tomography (CT) scanner to yield the projection datasets used in this study. RESULTS For λ within a broad range, the CG method with Fletcher-Reeves formula and NR line search offers the fastest convergence for an equal level of reconstruction accuracy. Using this minimization method, the reconstruction accuracy of PICCS was studied with respect to variations in α and λ. When the number of view angles is varied between 107, 80, 64, 40, 20, and 16, the relative root mean square error reaches a minimum value for α ≈ 0.5. For values of α near the optimal value, the spatial resolution of the reconstructed image remains relatively constant and the noise texture is very similar to that of the prior image, which was reconstructed using the filtered backprojection (FBP) algorithm. CONCLUSIONS Regarding the performance of the minimization methods, the nonlinear CG method with NR line search yields the best convergence speed. Regarding the performance of the PICCS image reconstruction, three main conclusions can be reached. (1) The performance of PICCS is optimal when the weighting parameter of the prior image parameter is selected to be near α = 0.5. (2) The spatial resolution measured for static objects in images reconstructed using PICCS from undersampled datasets is not degraded with respect to the fully-sampled reconstruction for α near its optimal value. (3) The noise texture of PICCS reconstructions is similar to that of the prior image, which was reconstructed using the conventional FBP method.

Reduced Image Noise at Low-Dose Multidetector CT of the Abdomen with Prior Image Constrained Compressed Sensing Algorithm

PURPOSE To assess the effect of prior image constrained compressed sensing (PICCS) on noise reduction and image quality at low-dose computed tomography (CT). MATERIALS AND METHODS This HIPAA-compliant institutional review board-approved retrospective study was performed by using DICOM CT colonography data sets obtained in 20 adult patients. Informed consent was waived. Low-dose CT colonography was performed with 64-detector CT by using the standard protocol with mean effective dose per series of 3.06 mSv (range, 1.4-7.7 mSv). PICCS was applied to standard filtered back-projection (FBP) series. For FBP and PICCS series, mean and standard deviation (SD) of attenuation were obtained with 100-mm(2) circular region of interest (ROI) at six sites (240 soft-tissue, colonic gas, and subcutaneous fat measurements). Two abdominal radiologists reviewed two- and three-dimensional CT colonography displays and graded image quality with a five-point scale. Phantom studies were performed to compare spatial resolution and image quality between FBP and PICCS. Mean image noise and image quality scores were calculated and compared for clinical and phantom data sets. Bland-Altman, generalized estimating equation regression model, and Student t tests were used to obtain limits of agreement and to compare noise ratios and subjective image quality. RESULTS Mean SD of attenuation (image noise) for ROIs was 38.0 for FBP and 12.2 for PICCS, corresponding to a noise-reduction factor of 3.1 (P < .001). Average noise reduction was 3.3 for soft tissue, 2.8 for air, and 3.0 for fat attenuation. Attenuation did not substantially change between FBP and PICCS images. Average two-dimensional image quality was 2.45 for FBP and 3.4 for PICCS (P < .001). Average three-dimensional image quality at three sites in the colon was 3.5 for FBP and 3.7 for PICCS (P = .34). Phantom data sets revealed no loss of spatial resolution in a line phantom and reduced noise in a liver tumor phantom when PICCS was compared with FBP. CONCLUSION Application of PICCS to standard FBP low-dose multidetector CT abdominal images results in substantial noise reduction and improved image quality.

Radiation dose efficiency comparison between differential phase contrast CT and conventional absorption CT

PURPOSE The superior radiation dose efficiency of a newly implemented differential phase contrast CT imaging method compared to the conventional absorption CT method is demonstrated. METHODS A differential phase contrast CT imaging method has recently been implemented using conventional x-ray sources with a grating interferometer consisting of three gratings. This approach offers the possibility of simultaneous reconstruction of both attenuation contrast and phase contrast images from a single acquisition. This enables a direct comparison of radiation dose efficiency of both types of reconstructed images under identical conditions. Radiation dose efficiency was studied by measuring the change in contrast-to-noise ratio (CNR) with different exposure levels. A physical phantom of 28.5 mm diameter was constructed and used for measurement of CNR in both the absorption and phase contrast CT images, which were reconstructed from the same data set. RESULTS For three of the four materials studied, at any given exposure level, the CNR of the differential phase contrast CT images was superior to that of the corresponding absorption contrast CT images. The most dramatic improvement was noted in the contrast between PMMA and water, where the CNR was improved by a factor of approximately 5.5 in the differential phase contrast CT images. Additionally, the CNR of phase contrast CT is empirically shown to have the same square root dependence on exposure, as is the case for absorption contrast CT. CONCLUSIONS The differential phase contrast CT method provided higher CNR than conventional absorption CT at equivalent dose levels for most of the materials studied, and so may enable achievement of the same object visibility as conventional absorption CT methods at a lower exposure level.

An alternative derivation of Katsevich's cone-beam reconstruction formula

In this paper an alternative derivation of Katsevichs cone-beam image reconstruction algorithm is presented. The starting point is the classical Tuys inversion formula. After (i) using the hidden symmetries of the intermediate functions, (ii) handling the redundant data by weighting them, (iii) changing the weighted average into an integral over the source trajectory parameter, and (iv) imposing an additional constraint on the weighting function, a filtered backprojection reconstruction formula from cone beam projections is derived. The following features are emphasized in the present paper: First, the nontangential condition in Tuys original data sufficiency conditions has been relaxed. Second, a practical regularization scheme to handle the singularity is proposed. Third, the derivation in the cone beam case is in the same fashion as that in the fan-beam case. Our final cone-beam reconstruction formula is the same as the one discovered by Katsevich in his most recent paper. However, the data sufficiency conditions and the regularization scheme of singularities are different. A detailed comparison between these two methods is presented.

Small-angle scattering computed tomography (SAS-CT) using a Talbot-Lau interferometer and a rotating anode x-ray tube: theory and experiments

X-ray differential phase contrast imaging methods, including projection imaging and the corresponding computed tomography (CT), have been implemented using a Talbot interferometer and either a synchrotron beam line or a low brilliance x-ray source generated by a stationary-anode x-ray tube. From small-angle scattering events which occur as an x-ray propagates through a medium, a signal intensity loss can be recorded and analyzed for an understanding of the micro-structures in an image object. This has been demonstrated using a Talbot-Lau interferometer and a stationary-anode x-ray tube. In this paper, theoretical principles and an experimental implementation of the corresponding CT imaging method are presented. First, a line integral is derived from analyzing the cross section of the small-angle scattering events. This method is referred to as small-angle scattering computed tomography (SAS-CT). Next, a Talbot-Lau interferometer and a rotating-anode x-ray tube were used to implement SAS-CT. A physical phantom and human breast tissue sample were used to demonstrate the reconstructed SAS-CT image volumes.