Anti-correlated Noise Reduction Research Articles

Signal-difference-to-noise comparison of temporal subtraction, kV-switching dual-energy and photon-counting dual-energy x-ray angiography.

Dual-energy (DE) x-ray angiography with photon-counting detectors (PCDs) may enable single-exposure DE imaging of coronaryvasculature. To compare the iodine signal-difference-to-noise ratio (SDNR) of single-exposure DE angiography with digital subtraction angiography (DSA) and kV-switching DE angiography for matched patient x-rayexposure. In a phantom study, we determined the technique parameters that maximized the iodine SDNR per root entrance air kerma for DSA, kV-switching DE angiography and single-exposure DE angiography. We measured SDNR from images of a phantom consisting of an iodine step-wedge immersed in a water tank of either 20 or 30cm in thickness. We also imaged a phantom with simulated vessels embedded in background clutter and measured vessel SDNR. For this second phantom, we also applied anti-correlated noise reduction (ACNR) and calculated the resulting iodine SDNR. All images were acquired using a cadmium telluride PCD with two energy bins and analog charge summing for charge sharing suppression. The energy-discrimination capabilities were only used for the single-exposure DE approach. Optimized techniques were compared in terms of SDNR per root air kerma for two levels of x-rayscatter. For the same patient x-ray exposure, the SDNR of single-exposure DE imaging without ACNR was 75% to 85% of that of kV-switching DE imaging (also without ACNR) and DSA, the latter two of which had nearly equal SDNR. The single-exposure DE approach required ∼50% of the tube load of the kV-switching approach to achieve the same SDNR. For matched patient air kermas, the single exposure approach required only ∼25% of the tube load of the kV-switching approach. ACNR increased SDNR by 2.4 and 3.0 for kV-switching and single-exposure DE imaging,respectively. Photon-counting, single-exposure DE angiography can suppress soft tissues and provide iodine SDNR levels comparable to DSA and kV-switching DE angiography for matched patient radiation exposures. When ACNR is used to reduce DE image noise, the SDNR of single-exposure DE imaging and kV-switching DE imaging exceed that of DSA by more than a factor of two. Compared to kV-switching DE imaging, single-exposure DE imaging requires substantially lower tube loading to achieve the sameSDNR.

Effect of different noise reduction techniques and template matching parameters on markerless tumor tracking using dual-energy imaging.

To evaluate the impact of various noise reduction algorithms and template matching parameters on the accuracy of markerless tumor tracking (MTT) using dual-energy (DE) imaging. A Varian TrueBeam linear accelerator was used to acquire a series of alternating 60 and 120kVp images (over a 180° arc) using fast kV switching, on five early-stage lung cancer patients. Subsequently, DE logarithmic weighted subtraction was performed offline on sequential images to remove bone. Various noise reduction techniques-simple smoothing, anticorrelated noise reduction (ACNR), noise clipping (NC), and NC-ACNR-were applied to the resultant DE images. Separately, tumor templates were generated from the individual planning CT scans, and band-pass parameter settings for template matching were varied. Template tracking was performed for each combination of noise reduction techniques and templates (based on band-pass filter settings). The tracking success rate (TSR), root mean square error (RMSE), and missing frames (percent unable to track) were evaluated against the estimated ground truth, which was obtained using Bayesian inference. DE-ACNR, combined with template band-pass filter settings of σlow =0.4mm and σhigh =1.6mm resulted in the highest TSR (87.5%), RMSE (1.40mm), and a reasonable amount of missing frames (3.1%). In comparison to unprocessed DE images, with optimized band-pass filter settings of σlow =0.6mm and σhigh =1.2mm, the TSR, RMSE, and missing frames were 85.3%, 1.62mm, and 2.7%, respectively. Optimized band-pass filter settings resulted in improved TSR values and a lower missing frame rate for both unprocessed DE and DE-ACNR as compared to the use previously published band-pass parameters based on single energy kV images. Noise reduction strategies combined with the optimal selection of band-pass filter parameters can improve the accuracy and TSR of MTT for lung tumors when using DE imaging.

Adaptive dual-energy algorithm based on pre-calibrated weighting factors for chest radiography

Objective. To develop a dual-energy (DE) algorithm with spatially varying weighting factors for material selection and noise suppression. Approach. Calibration step-phantoms, with overlapping slabs of solid water and bone with different thicknesses, were used to obtain the pre-calibrated material selection and noise reduction weighting factors. The Material selection weighting factors were calculated by finding a zero of contrast-to-noise-ratio (CNR) between regions with two overlapping materials and regions of only target material, while noise suppression weighting factors were determined by maximizing signal-to-noise ratio for overlapping regions. The pre-calibrated weighting factors were fitted with low and high energy radiograph of two Rando phantoms to create maps of material selection and noise suppression weighting factors, which used with DE algorithm and anti-correlated noise reduction (ACNR) algorithm to generate DE images. Three different implementations, including two different sizes of Rando phantoms and two different orientations (oblique and anterior-posterior), were investigated. Soft-tissue and bone only images of Rando phantoms were obtained with five combinations of DE algorithms and CNR, contrast, and noise values of selected regions of interest were compared to evaluate the performance of the novel method: simple log subtraction (SLS), SLS with uniform ACNR, adaptive DE (aDE), aDE with uniform ACNR, and aDE and adaptive ACNR (aACNR). Main results. Compared to SLS, the aDE algorithm demonstrated improved image quality in all three orientations. CNR increased with better contrast for both soft-tissue and bone images. Implementation of aACNR algorithm resulted in further reduction of image noise and improvements in CNR at the cost of contrast. However, aACNR algorithm showed better contrast compared to ACNR method. Significance. A novel DE algorithm was proposed, which showed improved material selection and noise suppression as compared to the conventional DE techniques and can be easily implemented in a clinical environment for real-time DE image generation.

Adaptive noise reduction for dual-energy x-ray imaging based on spatial variations in beam attenuation

Purpose. The main goal of this work is to improve the previously proposed patient-specific pixel-based dual-energy (PP-DE) algorithm by developing an adaptive anti-correlated noise reduction (ACNR) method, resulting in reduced image noise. Methods. Theoretical models of contrast-to-noise (CNR) and signal-to-noise (SNR) ratio were developed as functions of weighting factors for DE bone ω Bn or soft tissue ω ST cancellation. These analytical expressions describe CNR and SNR properties of dual-energy (DE) images, obtained with both simple log subtraction (SLS) and ACNR algorithms, and allow for a direct comparison between experimental and theoretical results. The theoretical models demonstrate the importance of ACNR weighting factor (ω A ) optimization leading to the maximization of the SNR of the final image. A step phantom was constructed, which consisted of overlapping slabs of solid water (0-30 cm) and bone-mimicking material (0-6 cm), resulting in a total of 7 × 7 regions. High-energy (HE) and low-energy (LE) images were acquired at 140 kVp and 60 kVp with a clinical ExacTrac imaging system. The CNR and SNR were obtained for the DE images as functions of ω Bn,ST and noise reduction weighting factor ω A for different combinations of thicknesses. Weighting factors for bone cancellation were optimized for each region of interest (ROI) by finding zeros of CNR function for DE images between soft tissue only and soft tissue plus bone regions (and vice versa for soft tissue cancellation). The weighting factor for the ACNR algorithm ω A was then optimized by maximizing the SNR function for each ROI. HE and LE images for an anthropomorphic Rando phantom were obtained with the same acquisition parameters as for the step phantom. DE images for bone only and soft tissue only were obtained with three algorithms: SLS and PP-DE with conventional ACNR (uniform ω A ), and PP-DE with adaptive ACNR (region-varying ω A ). Weighting factor maps for PP-DE and adaptive ACNR methods were obtained for Rando phantom geometry (which was determined from its CT scans) by interpolation (or extrapolation) of weighting factors for the step phantom. CNR values were calculated for different regions. Results. The CNR and SNR characteristics as functions of material cancellation and noise reduction weighting factors were obtained from theoretical models and experimental data from the step phantom. This showed a good qualitative validation of the models. For the ANCR algorithm, both the theory and experiment demonstrated that the material cancellation weighting factors (ω Bn,ST ) can be optimized independently of the noise cancellation weighting factors (ω A ), which can be optimized by maximizing SNR. For each ROI (with different overlapping bone and soft tissue thicknesses) the weighting factors ω Bn,ST were determined as well as corresponding optimal weighting factors ω A for noise reduction. For the Rando phantom, CNR values for regions representing different anatomical structures (ribs, spine, and tumor) were evaluated. It was shown that the proposed adaptive ACNR further improves image quality, compared to the conventional ACNR algorithm. The improvement is maximized for regions with bones (ribs or spine), where the largest attenuation is observed. Conclusion. The ACNR weighting factors are dependent on the material thicknesses due to varying beam attenuation leading to different levels of quantum noise. This was shown with the derived theoretical expressions of the CNR and SNR functions and was validated by experimental data. The adaptive ANCR DE algorithm was developed, which allows for an increase in image quality by spatially varying weighting factors for noise reduction. This algorithm complements the previously developed PP-DE algorithm to obtain better quality DE images.

Development of a non-linear dual-energy technique in chest radiography

The dual-energy (DE) technique is useful in chest radiography because it can separate anatomical structures, such as bone and soft tissue. The standard log subtraction, simple smoothing of the high-energy image, anti-correlated noise reduction, and general linear noise reduction algorithms have been used as conventional DE techniques to separate bone and soft tissue. However, conventional DE techniques cannot accurately decompose anatomical structures because these techniques are based on the assumption that X-ray imaging proceeds in a linear relationship. This relationship can cause quantum noise, as well as anatomical loss of normal tissue and difficulty in detecting lesions. In this study, we propose a non-linear DE technique that requires the calculation of the coefficient in advance by using a calibration phantom. The calibration phantom is composed of aluminium and PMMA to calculate non-linear coefficients using the quadratic fitting model for soft tissue and bone. The results demonstrated that the non-linear DE technique exhibited a higher contrast-to-noise ratio, signal-to-noise ratio, and figure of merit at 60/70 kVp and 130 kVp. In addition, it demonstrated better performance and image quality than conventional DE techniques in terms of material decomposition capability. In conclusion, diagnostic accuracy in chest radiography is expected to increase through the non-linear DE technique.

A novel phantom for characterization of dual energy imaging using an on-board imaging system

Dual-energy (DE) imaging using planar imaging with an on-board imager (OBI) is being considered in radiotherapy. We describe here a custom phantom designed to optimize DE imaging parameters using the OBI of a commercial linear accelerator. The phantom was constructed of lung-, tissue- and bone-equivalent material slabs. Five simulated tumors located at two different depths were encased in the lung-equivalent materials. Two slabs with bone-equivalent material inserts were constructed to simulate ribs, which overlap the simulated tumors. DE bone suppression was performed using a weighted logarithmic subtraction based on an iterative method that minimized the contrast between simulated bone- and lung-equivalent materials. The phantom was subsequently used to evaluate different combinations of high-low kV x-ray pairs of images based on the signal-difference-to-noise ratio (SDNR) metric. The results show a strong correlation between tumor visibility and selected energy pairs, where higher energy separation leads to larger SDNR values. To evaluate the effect of image post-processing methods on tumor visibility, an anti-correlated noise reduction (ACNR) technique and adaptive kernel scatter correction method were applied to subsequent DE images. Application of the ACNR technique approximately doubled the SDNR values, hence increasing tumor visibility, while scatter correction had little effect on SDNR values. This phantom allows for quick image acquisition and optimization of imaging parameters and weighting factors. Optimized DE imaging increases soft tissue visibility and may allow for markerless motion tracking of lung tumors.

Noise-reduction approaches to single-shot dual-energy imaging with a multilayer detector

The single-shot dual-energy (DE) method using a sandwich-like multilayer detector generally suffers from noise in the resultant DE images due to the poor quantum noise in images obtained from the rear detector layer and its amplification due to the subtraction operation in DE reconstruction. The use of interlayer metal filter can further increase noise in DE images by increasing the rear-detector quantum noise. We have adopted several noise-reduction algorithms to the single-shot DE reconstruction, which included the Gaussian-filtering noise reduction (GNR), the median-filtering noise reduction (MNR), and the anti-correlated noise reduction (ANR). We assessed the effectiveness of noise-reduction methods by investigating noise and dose-normalized contrast-to-noise ratio for a mouse-mimicking phantom consisting of aluminum bone and polyurethane soft tissues. Noise-power spectrum was also measured to investigate correlation noise. Overall, the ANR showed the best performance. At some extreme imaging technique conditions, such as a lower tube voltages and a larger spectral energy separation using thick copper sheets between the front and rear detector layers, the MNR outperformed the ANR. The performance of GNR was nearly similar to that of MNR. The results were well reflected into demonstration bone images obtained for a postmortem mouse. Although the ANR was effective to reduce noise in the single-shot DE imaging using the sandwich detector, multivariate optimization of ANR parameters with respect to imaging tasks and imaging techniques is remained as a future study.

Comparison study of noise reduction algorithms in dual energy chest digital tomosynthesis

Dual energy chest digital tomosynthesis (CDT) is a recently developed medical technique that takes advantage of both tomosynthesis and dual energy X-ray images. However, quantum noise, which occurs in dual energy X-ray images, strongly interferes with diagnosis in various clinical situations. Therefore, noise reduction is necessary in dual energy CDT. In this study, noise-compensating algorithms, including a simple smoothing of high-energy images (SSH) and anti-correlated noise reduction (ACNR), were evaluated in a CDT system. We used a newly developed prototype CDT system and anthropomorphic chest phantom for experimental studies. The resulting images demonstrated that dual energy CDT can selectively image anatomical structures, such as bone and soft tissue. Among the resulting images, those acquired with ACNR showed the best image quality. Both coefficient of variation and contrast to noise ratio (CNR) were the highest in ACNR among the three different dual energy techniques, and the CNR of bone was significantly improved compared to the reconstructed images acquired at a single energy. This study demonstrated the clinical value of dual energy CDT and quantitatively showed that ACNR is the most suitable among the three developed dual energy techniques, including standard log subtraction, SSH, and ACNR.

Bias-variance tradeoff in anticorrelated noise reduction for spectral CT.

In spectral CT, basis material decomposition is commonly used to generate a set of basis images showing the material composition at each point in the field of view. The noise in these images typically contains anticorrelations between the different basis images, which leads to increased noise in each basis image. These anticorrelations can be removed by changing the basis functions used in the material decomposition, but the resulting basis images can then no longer be used for quantitative measurements. Recent studies have demonstrated that reconstruction methods which take the anticorrelations into account give reduced noise in the reconstructed image. The purpose of this work is to analyze an analytically solvable denoising model problem and investigate its effect on the noise level and bias in the image as a function of spatial frequency. A denoising problem with a quadratic regularization term is studied as a mathematically tractable model for such a reconstruction method. An analytic formula for the resulting image in the spatial frequency domain is presented, and this formula is applied to a simple mathematical phantom consisting of an iodinated contrast agent insert embedded in soft tissue. We study the effect of the denoising on the image in terms of its transfer function and the visual appearance, the noise power spectrum and the Fourier component correlation coefficient of the resulting image, and compare the result to a denoising problem which does not model the anticorrelations in the image. Including the anticorrelations in the noise model of the denoising method gives 3-40% lower noise standard deviation in the soft-tissue image while leaving the iodine standard deviation nearly unchanged (0-1% difference). It also gives a sharper edge-spread function. The studied denoising method preserves the noise level and the anticorrelated structure at low spatial frequencies but suppresses the noise and removes the anticorrelations at higher spatial frequencies. Cross-talk between images gives rise to artifacts at high spatial frequencies. Modeling anticorrelations in a denoising problem can decrease the noise level in the basis images by removing anticorrelations at high spatial frequencies while leaving low spatial frequencies unchanged. In this way, basis image cross-talk does not lead to low spatial frequency bias but it may cause artifacts at edges in the image. This theoretical insight will be useful for researchers analyzing and designing reconstruction algorithms for spectral CT.

Improvement of material decomposition and image quality in dual-energy radiography by reducing image noise

Although digital radiography has been widely used for screening human anatomical structures in clinical situations, it has several limitations due to anatomical overlapping. To resolve this problem, dual-energy imaging techniques, which provide a method for decomposing overlying anatomical structures, have been suggested as alternative imaging techniques. Previous studies have reported several dual-energy techniques, each resulting in different image qualities. In this study, we compared three dual-energy techniques: simple log subtraction (SLS), simple smoothing of a high-energy image (SSH), and anti-correlated noise reduction (ACNR) with respect to material thickness quantification and image quality. To evaluate dual-energy radiography, we conducted Monte Carlo simulation and experimental phantom studies. The Geant 4 Application for Tomographic Emission (GATE) v 6.0 and tungsten anode spectral model using interpolation polynomials (TASMIP) codes were used for simulation studies and digital radiography, and human chest phantoms were used for experimental studies. The results of the simulation study showed improved image contrast-to-noise ratio (CNR) and coefficient of variation (COV) values and bone thickness estimation accuracy by applying the ACNR and SSH methods. Furthermore, the chest phantom images showed better image quality with the SSH and ACNR methods compared to the SLS method. In particular, the bone texture characteristics were well-described by applying the SSH and ACNR methods. In conclusion, the SSH and ACNR methods improved the accuracy of material quantification and image quality in dual-energy radiography compared to SLS. Our results can contribute to better diagnostic capabilities of dual-energy images and accurate material quantification in various clinical situations.

Simultaneous Reduction in Noise and Cross-Contamination Artifacts for Dual-Energy X-Ray CT

Purpose. Dual-energy CT imaging tends to suffer from much lower signal-to-noise ratio than single-energy CT. In this paper, we propose an improved anticorrelated noise reduction (ACNR) method without causing cross-contamination artifacts. Methods. The proposed algorithm diffuses both basis material density images (e.g., water and iodine) at the same time using a novel correlated diffusion algorithm. The algorithm has been compared to the original ACNR algorithm in a contrast-enhanced, IRB-approved patient study. Material density accuracy and noise reduction are quantitatively evaluated by the percent density error and the percent noise reduction. Results. Both algorithms have significantly reduced the noises of basis material density images in all cases. The average percent noise reduction is 69.3% and 66.5% with the ACNR algorithm and the proposed algorithm, respectively. However, the ACNR algorithm alters the original material density by an average of 13% (or 2.18 mg/cc) with a maximum of 58.7% (or 8.97 mg/cc) in this study. This is evident in the water density images as massive cross-contaminations are seen in all five clinical cases. On the contrary, the proposed algorithm only changes the mean density by 2.4% (or 0.69 mg/cc) with a maximum of 7.6% (or 1.31 mg/cc). The cross-contamination artifacts are significantly minimized or absent with the proposed algorithm. Conclusion. The proposed algorithm can significantly reduce image noise present in basis material density images from dual-energy CT imaging, with minimized cross-contaminations compared to the ACNR algorithm.

Comparison of model and human observer performance for detection and discrimination tasks using dual‐energy x‐ray images

Model observer performance, computed theoretically using cascaded systems analysis (CSA), was compared to the performance of human observers in detection and discrimination tasks. Dual-energy (DE) imaging provided a wide range of acquisition and decomposition parameters for which observer performance could be predicted and measured. This work combined previously derived observer models (e.g., Fisher-Hotelling and non-prewhitening) with CSA modeling of the DE image noise-equivalent quanta (NEQ) and imaging task (e.g., sphere detection, shape discrimination, and texture discrimination) to yield theoretical predictions of detectability index (d') and area under the receiver operating characteristic (Az). Theoretical predictions were compared to human observer performance assessed using 9-alternative forced-choice tests to yield measurement of Az as a function of DE image acquisition parameters (viz., allocation of dose between the low- and high-energy images) and decomposition technique [viz., three DE image decomposition algorithms: standard log subtraction (SLS), simple-smoothing of the high-energy image (SSH), and anti-correlated noise reduction (ACNR)]. Results showed good agreement between theory and measurements over a broad range of imaging conditions. The incorporation of an eye filter and internal noise in the observer models demonstrated improved correspondence with human observer performance. Optimal acquisition and decomposition parameters were shown to depend on the imaging task; for example, ACNR and SSH yielded the greatest performance in the detection of soft-tissue and bony lesions, respectively. This study provides encouraging evidence that Fourier-based modeling of NEQ computed via CSA and imaging task provides a good approximation to human observer performance for simple imaging tasks, helping to bridge the gap between Fourier metrics of detector performance (e.g., NEQ) and human observer performance.

Cascaded systems analysis of noise reduction algorithms in dual-energy imaging

An important aspect of dual-energy (DE) x-ray image decomposition is the incorporation of noise reduction techniques to mitigate the amplification of quantum noise. This article extends cascaded systems analysis of imaging performance to DE imaging systems incorporating linear noise reduction algorithms. A general analytical formulation of linear DE decomposition is derived, with weighted log subtraction and several previously reported noise reduction algorithms emerging as special cases. The DE image noise-power spectrum (NPS) and modulation transfer function (MTF) demonstrate that noise reduction algorithms impart significant, nontrivial effects on the spatial-frequency-dependent transfer characteristics which do not cancel out of the noise-equivalent quanta (NEQ). Theoretical predictions were validated in comparison to the measured NPS and MTF. The resulting NEQ was integrated with spatial-frequency-dependent task functions to yield the detectability index, d', for evaluation of DE imaging performance using different decomposition algorithms. For a 3 mm lung nodule detection task, the detectability index varied from d' < 1 (i.e., nodule barely visible) in the absence of noise reduction to d' > 2.5 (i.e., nodule clearly visible) for "anti-correlated noise reduction" (ACNR) or "simple-smoothing of the high-energy image" (SSH) algorithms applied to soft-tissue or bone-only decompositions, respectively. Optimal dose allocation (A*, the fraction of total dose delivered in the low-energy projection) was also found to depend on the choice of noise reduction technique. At fixed total dose, multi-function optimization suggested a significant increase in optimal dose allocation from A* = 0.32 for conventional log subtraction to A* = 0.79 for ACNR and SSH in soft-tissue and bone-only decompositions, respectively. Cascaded systems analysis extended to the general formulation of DE image decomposition provided an objective means of investigating DE imaging performance across a broad range of acquisition and decomposition algorithms in a manner that accounts for the spatial-frequency-dependent imaging task.

TH‐C‐330A‐09: Cascaded Systems Analysis of Noise Reduction Algorithms for Dual‐Energy Imaging

Purpose: While dual‐energy (DE) imaging provides increased nodule conspicuity in soft‐tissue images and greater calcification visualization in bone‐only images, DE image decomposition amplifies noise present in the projection data. This paper extends task‐based cascaded systems analysis (CSA) to include a variety of DE noise reduction algorithms, offering a general analytical approach to optimizing DE imaging performance. Method and Materials: Two noise reduction algorithms [simple‐smoothing of the high‐energy image (SSH) and anti‐correlated noise reduction (ACNR)] were incorporated into CSA models for DE imaging to describe the DE modulation transfer function (MTFDE), noise‐power spectrum (NPSDE), and noise equivalent quanta (NEQDE). The MTFDE and NPSDE were measured using standard edge‐spread function and flood‐field techniques adapted to DE imaging (with noise‐reduction processing) and compared to theoretical results. The MTFDE and NPSDE were combined to yield the NEQDE and integrated with a spatial‐frequency‐dependent task function to provide a detectability index for evaluation of imaging performance using standard, SSH, and ACNR image decompositions. Results: The MTFDE and NPSDE calculated using CSA agreed well with measurements. Detectability index provided an objective performance metric for identifying superior noise reduction algorithms under conditions of varying kVp, dose, and imaging task. For example, the DE detectability index for a delta‐function detection task in the soft‐tissue image was by far greatest for the ACNR algorithm, whereas SSH performed best for the bone‐only image. A gaussian detection task, on the other hand, indicated superior performance for the ACNR algorithm for both soft‐tissue and bone‐only images. Conclusions: Extension of CSA to include the influence of DE noise‐reduction algorithms such as SSH and ACNR offers a powerful guide to system optimization. The general, analytical approach provides an objective means of selecting superior noise‐reduction algorithms and “tuning” the parameters therein in a manner that weighs spatial resolution and noise in relation to the imaging task.