Zhengrong Liang

Penalized weighted least-squares approach to sinogram noise reduction and image reconstruction for low-dose X-ray computed tomography

Reconstructing low-dose X-ray computed tomography (CT) images is a noise problem. This work investigated a penalized weighted least-squares (PWLS) approach to address this problem in two dimensions, where the WLS considers first- and second-order noise moments and the penalty models signal spatial correlations. Three different implementations were studied for the PWLS minimization. One utilizes a Markov random field (MRF) Gibbs functional to consider spatial correlations among nearby detector bins and projection views in sinogram space and minimizes the PWLS cost function by iterative Gauss-Seidel algorithm. Another employs Karhunen-Loeve (KL) transform to de-correlate data signals among nearby views and minimizes the PWLS adaptively to each KL component by analytical calculation, where the spatial correlation among nearby bins is modeled by the same Gibbs functional. The third one models the spatial correlations among image pixels in image domain also by a MRF Gibbs functional and minimizes the PWLS by iterative successive over-relaxation algorithm. In these three implementations, a quadratic functional regularization was chosen for the MRF model. Phantom experiments showed a comparable performance of these three PWLS-based methods in terms of suppressing noise-induced streak artifacts and preserving resolution in the reconstructed images. Computer simulations concurred with the phantom experiments in terms of noise-resolution tradeoff and detectability in low contrast environment. The KL-PWLS implementation may have the advantage in terms of computation for high-resolution dynamic low-dose CT imaging

Nonlinear sinogram smoothing for low-dose X-ray CT

When excessive quantum noise is present in extremely low dose X-ray CT imaging, statistical properties of the data has to be considered to achieve a satisfactory image reconstruction. Statistical iterative reconstruction with accurate modeling of the noise, rather than a filtered back-projection (FBP) with low-pass filtering, is one way to deal with the problem. Estimating a noise-free sinogram to satisfy the FBP reconstruction for the Radon transform is another way. The benefits of the latter include a higher computation efficiency, more uniform spatial resolution in the reconstructed image, and less modification of the current machine configurations. In a clinic X-ray CT system, the acquired raw data must be calibrated, in addition to the logarithmic transform, to achieve the high diagnostic image quality. The calibrated projection data or sinogram no longer follow a compound Poisson distribution in general, but are close to a Gaussian distribution with signal-dependent variance. In this paper, we first investigated a relatively accurate statistical model for the sinogram data, based on several phantom experiments. Then we developed a penalized likelihood method to smooth the sinogram, which led to a set of nonlinear equations that can be solved by iterated conditional mode (ICM) algorithm within a reasonable computing time. The method was applied to several experimental datasets acquired at 120 kVp, 10 mA/20 mA/50 mA protocols with a GE HiSpeed multi-slice detector CT scanner and demonstrated a significant noise suppression without noticeable sacrifice of the spatial resolution.

Parameter estimation and tissue segmentation from multispectral MR images

A statistical method is developed to classify tissue types and to segment the corresponding tissue regions from relaxation time T(1 ), T(2), and proton density P(D) weighted magnetic resonance images. The method assumes that the distribution of image intensities associated with each tissue type can be expressed as a multivariate likelihood function of three weighted signal intensity values (T(1), T(2), P(D)) at each location within that tissue regions. The method further assumes that the underlying tissue regions are piecewise contiguous and can be characterized by a Markov random field prior. In classifying the tissue types, the method models the likelihood of realizing the images as a finite multivariate-mixture function. The class parameters associated with the tissue types (i.e. the weighted intensity means, variances and correlation coefficients of the multivariate function, as well as the number of voxels within regions of the tissue types of are estimated by maximum likelihood. The estimation fits the class parameters to the image data via the expectation-maximization algorithm. The number of classes associated with the tissue types is determined by the information criterion of minimum description length. The method segments the tissue regions, given the estimated class parameters, by maximum a posteriori probability. The prior is constructed by the tissue-region membership of the first- and second-order neighborhood. The method is tested by a few sets of T(1), T(2), and P(D) weighted images of the brain acquired with a 1.5 Tesla whole body scanner. The number of classes and the associated class parameters are automatically estimated. The regions of different brain tissues are satisfactorily segmented.

3D virtual colonoscopy

The authors present here a method called 3D virtual colonoscopy, which is an alternative method to existing procedures of imaging the mucosal surface of the colon. Using 3D reconstruction of helical CT data and volume visualization techniques, the authors generate images of the inner surface of the colon as if the viewers eyes were inside the colon. They also create interactive flythroughs and off-line automatically-produced animations through the inside of the colon. The visualization is accomplished with VolVis, which is a comprehensive system for interactive volume visualization. The authors are specifically interested in visualizing colonic polyps larger than one cm since these have a high probability of containing carcinoma. The authors present testing results of their method as applied to two plastic pipe simulations and to the Visible Human data set.

Cognitive performance and MR markers of cerebral injury in cognitively impaired MS patients

Objective: To relate neuropsychological performance to measures of cerebral injury in persons with MS selected for cognitive impairment. Methods: Participants were 37 individuals with relapsing–remitting (59.5%) and secondary progressive (40.5%) MS. They were tested at baseline as part of a clinical trial to enhance cognition with an acetylcholinesterase inhibitor. Eligibility criteria included at least mild cognitive impairment on a verbal learning and memory task. A modified Brief Repeatable Battery of Neuropsychological Tests formed the core of the behavioral protocol. Neuroimaging measures were central (ventricular) cerebral atrophy, lesion volume, and ratios of N -acetyl aspartate (NAA) to both creatine and choline. Results: A clear, consistent relation was found between cognitive and MR measures. Among neuroimaging measures, central atrophy displayed the highest correlations with cognition, accounting for approximately half the variance in overall cognitive performance. NAA ratios in right hemisphere sites displayed larger correlations than those on the left. Multiple regression models combining the MR measures accounted for well over half the variance in overall cognitive performance. The Symbol Digit Modalities Test was the neuropsychological task most strongly associated with the neuroimaging variables. Conclusions: If a strong and stable association can be firmly established between cognitive and MR variables in appropriate subsets of MS patients, it might aid in the investigation of interventions to enhance cognition and modify the course of the disease.

Automatic centerline extraction for virtual colonoscopy

In this paper, we introduce a concise and concrete definition of an accurate colon centerline and provide an efficient automatic means to extract the centerline and its associated branches (caused by a forceful touching of colon and small bowel or a deep fold in twisted colon lumen). We further discuss its applications on fly-through path planning and endoscopic simulation, as well as its potential to solve the challenging touching and colon collapse problems in virtual colonoscopy. Experimental results demonstrated its centeredness, robustness, and efficiency.

A novel approach to extract colon lumen from CT images for virtual colonoscopy

An automatic method has been developed for segmentation of abdominal computed tomography (CT) images for virtual colonoscopy obtained after a bowel preparation of a low-residue diet with ingested contrast solutions to enhance the image intensities of residual colonic materials. Removal of the enhanced materials was performed electronically by a computer algorithm. The method is a multistage approach that employs a modified self-adaptive on-line, vector quantization technique for a low-level image classification and utilizes a region-growing strategy for a high-level feature extraction. The low-level classification labels each voxel based on statistical analysis of its three-dimensional intensity vectors consisting of nearby voxels. The high-level processing extracts the labeled stool, fluid and air voxels within the colon, and eliminates bone and lung voxels which have similar image intensities as the enhanced materials and air, but are physically separated from the colon. This method was evaluated by volunteer studies based on both objective and subjective criteria. The validation demonstrated that the method has a high reproducibility and repeatability and a small error due to partial volume effect. As a result of this electronic colon cleansing, routine physical bowel cleansing prior to virtual colonoscopy may not be necessary.

An experimental study on the noise properties of x-ray CT sinogram data in Radon space

Computed tomography (CT) has been well established as a diagnostic tool through hardware optimization and sophisticated data calibration. For screening purposes, the associated x-ray exposure risk must be minimized. An effective way to minimize the risk is to deliver fewer x-rays to the subject or lower the mAs parameter in data acquisition. This will increase the data noise. This work aims to study the noise property of the calibrated or preprocessed sinogram data in Radon space as the mAs level decreases. An anthropomorphic torso phantom was scanned repeatedly by a commercial CT imager at five different mAs levels from 100 down to 17 (the lowest value provided by the scanner). The preprocessed sinogram datasets were extracted from the CT scanner to a laboratory computer for noise analysis. The repeated measurements at each mAs level were used to test the normality of the repeatedly measured samples for each data channel using the Shapiro-Wilk statistical test merit. We further studied the probability distribution of the repeated measures. Most importantly, we validated a theoretical relationship between the sample mean and variance at each channel. It is our intention that the statistical test and particularly the relationship between the first and second statistical moments will improve low-dose CT image reconstruction for screening applications.

Reconstruction for proton computed tomography by tracing proton trajectories: A Monte Carlo study

Proton computed tomography (pCT) has been explored in the past decades because of its unique imaging characteristics, low radiation dose, and its possible use for treatment planning and on-line target localization in proton therapy. However, reconstruction of pCT images is challenging because the proton path within the object to be imaged is statistically affected by multiple Coulomb scattering. In this paper, we employ GEANT4-based Monte Carlo simulations of the two-dimensional pCT reconstruction of an elliptical phantom to investigate the possible use of the algebraic reconstruction technique (ART) with three different path-estimation methods for pCT reconstruction. The first method assumes a straight-line path (SLP) connecting the proton entry and exit positions, the second method adapts the most-likely path (MLP) theoretically determined for a uniform medium, and the third method employs a cubic spline path (CSP). The ART reconstructions showed progressive improvement of spatial resolution when going from the SLP [2 line pairs (lp) cm(-1)] to the curved CSP and MLP path estimates (5 lp cm(-1)). The MLP-based ART algorithm had the fastest convergence and smallest residual error of all three estimates. This work demonstrates the advantage of tracking curved proton paths in conjunction with the ART algorithm and curved path estimates.

Single- and dual-energy CT with monochromatic synchrotron x-rays

We explored the potential for clinical research of computed tomography (CT) with monochromatic x-rays using the preclinical multiple energy computed tomography (MECT) system at the National Synchrotron Light Source. MECT has a fixed, horizontal fan beam with a subject apparatus rotating about a vertical axis; it will be used for imaging the human head and neck. Two CdWO4-photodiode array detectors with different spatial resolutions were used. A 10.5 cm diameter acrylic phantom was imaged with MECT at 43 keV and with a conventional CT (CCT) at 80 kVp: spatial resolution approximately equal to 6.5 line pairs (lp)/cm for both; slice height, 2.6 mm for MECT against 3.0 mm for CCT; surface dose, 3.1 cGy for MECT against 2.0 cGy for CCT. The resultant image noise was 1.5 HU for MECT against 3 HU for CCT. Computer simulations of the same images with more precisely matched spatial resolution, slice height and dose indicated an image-noise ratio of 1.4:1.0 for CCT against MECT. A 13.5 cm diameter acrylic phantom imaged with MECT at approximately 0.1 keV above the iodine K edge and with CCT showed, for a 240 micrograms I ml-1 solution, an image contrast of 26 HU for MECT and 13 and 9 HU for the 80 and 100 kVp CCT, respectively. The corresponding numbers from computer simulation of the same images were 26, 12, and 9 HU, respectively. MECTs potential for use in clinical research is discussed.