Quantitative statistical methods for image quality assessment

Investigating the role of imaging factors in the variability of CT-based texture analysis metrics

OBJECTIVE

This study assesses the robustness of first-order radiomic texture features namely interquartile range (IQR), coefficient of variation (CV) and standard deviation (SD) derived from computed tomography (CT) images by varying dose, reconstruction algorithms and slice thickness using scans of a uniform water phantom, a commercial anthropomorphic liver phantom, and a human liver in-vivo.

MATERIALS AND METHODS

Scans were acquired on a 16 cm detector GE Revolution Apex Edition CT scanner with variations across three different nominal slice thicknesses: 0.625, 1.25, and 2.5 mm, three different dose levels: CTDIvol of 13.86 mGy for the standard dose, 40% reduced dose and 60% reduced dose and two different reconstruction algorithms: a deep learning image reconstruction (DLIR-high) algorithm and a hybrid iterative reconstruction (IR) algorithm ASiR-V50% (AV50) were explored, varying one at a time. To assess the effect of non-linear modifications of images by AV50 and DLIR-high, images of the water phantom were also reconstructed using filtered back projection (FBP). Quantitative measures of IQR, CV and SD were extracted from twelve pre-selected, circular (1 cm diameter) regions of interest (ROIs) capturing different texture patterns across all scans.

RESULTS

Across all scans, imaging, and reconstruction settings, CV, IQR and SD were observed to increase with reduction in dose and slice thickness. An exception to this observation was found when using FBP reconstruction. Lower values of CV, IQR and SD were observed in DLIR-high reconstructions compared to AV50 and FBP. The Poisson statistics were more stringently noted in FBP than DLIR-high and AV50, due to the non-linear nature of the latter two algorithms.

CONCLUSION

Variation in image noise due to dose reduction algorithms, tube current, and slice thickness show a consistent trend across phantom and patient scans. Prospective evaluation across multiple centers, scanners and imaging protocols is needed for establishing quality assurance standards of radiomics.

Image quality assessment for machine learning tasks using meta-reinforcement learning

In this paper, we consider image quality assessment (IQA) as a measure of how images are amenable with respect to a given downstream task, or task amenability. When the task is performed using machine learning algorithms, such as a neural-network-based task predictor for image classification or segmentation, the performance of the task predictor provides an objective estimate of task amenability. In this work, we use an IQA controller to predict the task amenability which, itself being parameterised by neural networks, can be trained simultaneously with the task predictor. We further develop a meta-reinforcement learning framework to improve the adaptability for both IQA controllers and task predictors, such that they can be fine-tuned efficiently on new datasets or meta-tasks. We demonstrate the efficacy of the proposed task-specific, adaptable IQA approach, using two clinical applications for ultrasound-guided prostate intervention and pneumonia detection on X-ray images.

A probabilistic approach to tomography and adjoint state methods, with an application to full waveform inversion in medical ultrasound

Bayesian methods are a popular research direction for inverse problems. There are a variety of techniques available to solve Bayes' equation, each with their own strengths and limitations. Here, we discuss stochastic variational inference (SVI), which solves Bayes' equation using gradient-based methods. This is important for applications which are time-limited (e.g. medical tomography) or where solving the forward problem is expensive (e.g. adjoint methods). To evaluate the use of SVI in both these contexts, we apply it to ultrasound tomography of the brain using full-waveform inversion (FWI). FWI is a computationally expensive adjoint method for solving the ultrasound tomography inverse problem, and we demonstrate that SVI can be used to find a no-cost estimate of the pixel-wise variance of the sound-speed distribution using a mean-field Gaussian approximation. In other words, we show experimentally that it is possible to estimate the pixel-wise uncertainty of the sound-speed reconstruction using SVI and a common approximation which is already implicit in other types of iterative reconstruction. Uncertainty estimates have a variety of uses in adjoint methods and tomography. As an illustrative example, we focus on the use of uncertainty for image quality assessment. This application is not limiting; our variance estimator has effectively no computational cost and we expect that it will have applications in fields such as non-destructive testing or aircraft component design where uncertainties may not be routinely estimated.

Noise2Void: unsupervised denoising of PET images

Objective:Elevated noise levels in positron emission tomography (PET) images lower image quality and quantitative accuracy and are a confounding factor for clinical interpretation. The objective of this paper is to develop a PET image denoising technique based on unsupervised deep learning.Significance:Recent advances in deep learning have ushered in a wide array of novel denoising techniques, several of which have been successfully adapted for PET image reconstruction and post-processing. The bulk of the deep learning research so far has focused on supervised learning schemes, which, for the image denoising problem, require paired noisy and noiseless/low-noise images. This requirement tends to limit the utility of these methods for medical applications as paired training datasets are not always available. Furthermore, to achieve the best-case performance of these methods, it is essential that the datasets for training and subsequent real-world application have consistent image characteristics (e.g. noise, resolution, etc), which is rarely the case for clinical data. To circumvent these challenges, it is critical to develop unsupervised techniques that obviate the need for paired training data.Approach:In this paper, we have adapted Noise2Void, a technique that relies on corrupt images alone for model training, for PET image denoising and assessed its performance using PET neuroimaging data. Noise2Void is an unsupervised approach that uses a blind-spot network design. It requires only a single noisy image as its input, and, therefore, is well-suited for clinical settings. During the training phase, a single noisy PET image serves as both the input and the target. Here we present a modified version of Noise2Void based on a transfer learning paradigm that involves group-level pretraining followed by individual fine-tuning. Furthermore, we investigate the impact of incorporating an anatomical image as a second input to the network.Main Results:We validated our denoising technique using simulation data based on the BrainWeb digital phantom. We show that Noise2Void with pretraining and/or anatomical guidance leads to higher peak signal-to-noise ratios than traditional denoising schemes such as Gaussian filtering, anatomically guided non-local means filtering, and block-matching and 4D filtering. We used the Noise2Noise denoising technique as an additional benchmark. For clinical validation, we applied this method to human brain imaging datasets. The clinical findings were consistent with the simulation results confirming the translational value of Noise2Void as a denoising tool.

Artificial Intelligence-Based Image Enhancement in PET Imaging: Noise Reduction and Resolution Enhancement

High noise and low spatial resolution are two key confounding factors that limit the qualitative and quantitative accuracy of PET images. Artificial intelligence models for image denoising and deblurring are becoming increasingly popular for the postreconstruction enhancement of PET images. We present a detailed review of recent efforts for artificial intelligence-based PET image enhancement with a focus on network architectures, data types, loss functions, and evaluation metrics. We also highlight emerging areas in this field that are quickly gaining popularity, identify barriers to large-scale adoption of artificial intelligence models for PET image enhancement, and discuss future directions.

Myocardial blood flow quantitation in patients with congestive heart failure: head-to-head comparison between rapid-rotating gantry SPECT and CZT SPECT

PURPOSE

Recently, the feasibility of myocardial blood flow (MBF) quantitation using rapid-rotating gantry (RRG) and cadmium-zinc-telluride (CZT) SPECT cameras has been demonstrated. We compared MBF quantitation using these two camera systems.

METHODS

Twenty patients with congestive heart failure (CHF) and 20 patients without CHF (non-CHF) were included. On two consecutive days, dynamic SPECT imaging was performed after a bolus injection of 20 mCi of 99mTc-Sestamibi (MIBI) with RRG-SPECT and list-mode acquisition with CZT-SPECT. All dynamic SPECT images were reconstructed with full physical corrections, corrections for ventricular spillover and partial volume effect, using one-tissue kinetic modeling. Resting MBF converted from K1 was then corrected for MIBI extraction fraction and adjusted for rate-pressure product.

RESULTS

In both patient groups, there was no significant difference between resting MBF values measured with RRG-SPECT and CZT-SPECT systems (P = 0.06, P = 0.2 respectively). For CHF patients, linear regression (LR) was y(RRG) = 1.0412x (CZT) (r = 0.97) with a small systemic difference (Δ = 0.03 mL·min-1·g-1, 95% CI - 0.11 to 0.20) by Bland-Altman analysis. For non-CHF patients, LR was y(RRG) = 1.025x (CZT) (r = 0.89) with also small systemic difference (ΔT= 0.02 mL·min-1·g-1, 95% CI - 0.14 to 0.19) in BA analysis.

CONCLUSION

Physical corrections along with other image corrections can provide comparable MBF quantitations in both CHF and non-CHF patients, regardless of the type of SPECT systems used.

A method to measure the extent of myocardial ischemia and steal with SPECT myocardial blood flow quantitation

OBJECTIVE

We introduced a method to measure the extent of myocardial ischemia and steal with SPECT MBF quantitation.

METHODS

Eighty-seven patients who received rest/Dipyridamole-stress ^99mTc-Sestamibi (MIBI) dynamic SPECT scans and coronary angiography were included. Dynamic SPECT images were reconstructed with full physical corrections. The one-tissue kinetic model was utilized to quantify K1 and further converted to MBF with required corrections. Rest MBF, stress MBF and myocardial flow reserve (MFR) were converted to a flow status polar map by a flow diagram. Extents of 7 flow statuses were verified their cutoff points for detecting stenoses. The diagnostic performance (DP) was compared to that of MFR.

RESULTS

Cutoff point of the extent to detect ≥ 50% stenosis was 3.01% for ischemia-steal status and 20.3% for the combined status of ischemia-steal and moderate. Using these criteria, sensitivity, specificity and accuracy to detect ≥ 50% stenosis were (80%, 75%, 79%) and (86%, 68%, 80%) for ≥ 70% stenosis. The DP was superior to that of MFR < 2.0 criterion (≥ 50%: 70%, 63%, 69%; ≥ 70%: 73%, 61%, 69%) (all p < 0.015).

CONCLUSION

SPECT MBF quantitation integrated with the flow diagram can measure the extent of myocardial ischemia and steal which appeared more accurate to detect angiographic stenoses than the single MFR parameter.

Application of Radiomics for Personalized Treatment of Cancer Patients

Radiomics is a novel concept that relies on obtaining image data from examinations such as computed tomography (CT), magnetic resonance imaging (MRI), or positron emission tomography (PET). With the appropriate algorithm, the extracted results have broad applicability and potential for a massive positive impact in radiology. For example, clinicians can verify treatment efficiency, predict the location of tumor metastasis, correlate results with a histopathological examination, or more accurately define the type of cancer. Combining radiomics with other testing techniques allows every patient to have a personalized treatment plan that is essential for advanced examination and treatment. This article explains the process of radiomics, including data collection mechanisms, combined use with genomics, and artificial intelligence and immunology techniques, which may solve many of the challenges faced by doctors in diagnosing and treating their patients.

3D Tensor Based Nonlocal Low Rank Approximation in Dynamic PET Reconstruction

Reconstructing images from multi-view projections is a crucial task both in the computer vision community and in the medical imaging community, and dynamic positron emission tomography (PET) is no exception. Unfortunately, image quality is inevitably degraded by the limitations of photon emissions and the trade-off between temporal and spatial resolution. In this paper, we develop a novel tensor based nonlocal low-rank framework for dynamic PET reconstruction. Spatial structures are effectively enhanced not only by nonlocal and sparse features, but momentarily by tensor-formed low-rank approximations in the temporal realm. Moreover, the total variation is well regularized as a complementation for denoising. These regularizations are efficiently combined into a Poisson PET model and jointly solved by distributed optimization. The experiments demonstrated in this paper validate the excellent performance of the proposed method in dynamic PET.

PET Image Deblurring and Super-Resolution with an MR-Based Joint Entropy Prior

The intrinsically limited spatial resolution of PET confounds image quantitation. This paper presents an image deblurring and super-resolution framework for PET using anatomical guidance provided by high-resolution MR images. The framework relies on image-domain post-processing of already-reconstructed PET images by means of spatially-variant deconvolution stabilized by an MR-based joint entropy penalty function. The method is validated through simulation studies based on the BrainWeb digital phantom, experimental studies based on the Hoffman phantom, and clinical neuroimaging studies pertaining to aging and Alzheimer's disease. The developed technique was compared with direct deconvolution and deconvolution stabilized by a quadratic difference penalty, a total variation penalty, and a Bowsher penalty. The BrainWeb simulation study showed improved image quality and quantitative accuracy measured by contrast-to-noise ratio, structural similarity index, root-mean-square error, and peak signal-to-noise ratio generated by this technique. The Hoffman phantom study indicated noticeable improvement in the structural similarity index (relative to the MR image) and gray-to-white contrast-to-noise ratio. Finally, clinical amyloid and tau imaging studies for Alzheimer's disease showed lowering of the coefficient of variation in several key brain regions associated with two target pathologies.

Artificial Neural Network Enhanced Bayesian PET Image Reconstruction

In positron emission tomography (PET) image reconstruction, the Bayesian framework with various regularization terms has been implemented to constrain the radio tracer distribution. Varying the regularizing weight of a maximum a posteriori (MAP) algorithm specifies a lower bound of the tradeoff between variance and spatial resolution measured from the reconstructed images. The purpose of this paper is to build a patch-based image enhancement scheme to reduce the size of the unachievable region below the bound and thus to quantitatively improve the Bayesian PET imaging. We cast the proposed enhancement as a regression problem which models a highly nonlinear and spatial-varying mapping between the reconstructed image patches and an enhanced image patch. An artificial neural network model named multilayer perceptron (MLP) with backpropagation was used to solve this regression problem through learning from examples. Using the BrainWeb phantoms, we simulated brain PET data at different count levels of different subjects with and without lesions. The MLP was trained using the image patches reconstructed with a MAP algorithm of different regularization parameters for one normal subject at a certain count level. To evaluate the performance of the trained MLP, reconstructed images from other simulations and two patient brain PET imaging data sets were processed. In every testing cases, we demonstrate that the MLP enhancement technique improves the noise and bias tradeoff compared with the MAP reconstruction using different regularizing weights thus decreasing the size of the unachievable region defined by the MAP algorithm in the variance/resolution plane.

Fast Quasi-Newton Algorithms for Penalized Reconstruction in Emission Tomography and Further Improvements via Preconditioning

This paper reports on the feasibility of using a quasi-Newton optimization algorithm, limited-memory Broyden-Fletcher-Goldfarb-Shanno with boundary constraints (L-BFGS-B), for penalized image reconstruction problems in emission tomography (ET). For further acceleration, an additional preconditioning technique based on a diagonal approximation of the Hessian was introduced. The convergence rate of L-BFGS-B and the proposed preconditioned algorithm (L-BFGS-B-PC) was evaluated with simulated data with various factors, such as the noise level, penalty type, penalty strength and background level. Data of three ¹⁸F-FDG patient acquisitions were also reconstructed. Results showed that the proposed L-BFGS-B-PC outperforms L-BFGS-B in convergence rate for all simulated conditions and the patient data. Based on these results, L-BFGS-B-PC shows promise for clinical application.

Can visual analogue scale be used in radiologic subjective image quality assessment?

BACKGROUND

Assessment of qualitative or subjective image quality in radiology is traditionally performed using a fixed-point scale even though reproducibility has proved challenging.

OBJECTIVE

Image quality of 3-T coronary magnetic resonance (MR) angiography was evaluated using three scoring methods, hypothesizing that a continuous scoring scale like visual analogue scale would improve the assessment.

MATERIALS AND METHODS

Adolescents corrected for transposition of the great arteries with arterial switch operation, ages 9-15 years (n=12), and healthy, age-matched controls (n=12), were examined with 3-D steady-state free precession magnetic resonance imaging. Image quality of the coronary artery origin was evaluated by using a fixed-point scale (1-4), visual analogue scale of 10 cm and a visual analogue scale with reference points (figurative visual analogue scale). Satisfactory image quality was set to a fixed-point scale 3=visual analogue scale/figurative visual analogue scale 6.6 cm. Statistical analysis was performed using Cohen kappa coefficient and agreement index.

RESULTS

The mean interobserver scores for the fixed-point scale, visual analogue scale and figurative visual analogue scale were, respectively, in the left main stem 2.8, 5.7, 7.0; left anterior descending artery 2.8, 4.7, 6.6; circumflex artery 2.5, 4.5, 6.2, and right coronary artery 3.2, 6.3, 7.7. Scoring with a fixed-point scale gave an intraobserver κ of 0.52-0.77 while interobserver κ was lacking. For visual analogue scale and figurative visual analogue scale, intraobserver agreement indices were, respectively, 0.08-0.58 and 0.43-0.71 and interobserver agreement indices were up to 0.5 and 0.65, respectively.

CONCLUSION

Qualitative image quality evaluation with coronary 3-D steady-state free precession MR angiography, using a visual analogue scale with reference points, had better reproducibility compared to a fixed-point scale and visual analogue scale. Image quality, being a continuum, may be better determined by this method.

Avoiding full corrections in dynamic SPECT images impacts the performance of SPECT myocardial blood flow quantitation

INTRODUCTION

This study investigated the performance of SPECT myocardial blood flow (MBF) quantitation lacking full physical corrections (All Corr) in dynamic SPECT (DySPECT) images.

METHODS

Eleven healthy normal volunteers (HVT) and twenty-four patients with angiography-documented CAD were assessed. All Corr in 99mTc-sestamibi DySPECT encompassed noise reduction (NR), resolution recovery (RR), and corrections for scatter (SC) and attenuation (AC), otherwise no correction (NC) or only partial corrections. The performance was evaluated by quality index (R 2) and blood-pool spillover index (FBV) in kinetic modeling, and by rest flow (RMBF) and stress flow (SMBF) compared with those of All Corr.

RESULTS

In HVT group, NC diminished 2-fold flow uniformity with the most degraded quality (15%-18% reduced R 2) and elevated spillover effect (45%-50% increased FBV). Consistently higher RMBF and SMBF were discovered in both groups (HVT 1.54/2.31 higher; CAD 1.60/1.72; all P < .0001). Bland-Altman analysis revealed positive flow bias (HVT 0.9-2.6 mL/min/g; CAD 0.7-1.3) with wide ranges of 95% CI of agreement (HVT NC -1.9-7.1; NR -0.4-4.4; NR + SC -1.1-4.3; NR + SC + RR -0.7-2.5) (CAD NC -1.2-3.8; NR -1.0-2.8; NR + SC -1.0-2.5; NR + SC + RR -1.1-2.6).

CONCLUSIONS

Uncorrected physical interference in DySPECT images can extensively impact the performance of MBF quantitation. Full physical corrections should be considered to warrant this tool for clinical utilization.

Evaluation of lesion detectability in positron emission tomography when using a convergent penalized likelihood image reconstruction method

We have previously developed a convergent penalized likelihood (PL) image reconstruction algorithm using the relative difference prior (RDP) and showed that it achieves more accurate lesion quantitation compared to ordered subsets expectation maximization (OSEM). We evaluated the detectability of low-contrast liver and lung lesions using the PL-RDP algorithm compared to OSEM. We performed a two-alternative forced choice study using a channelized Hotelling observer model that was previously validated against human observers. Lesion detectability showed a stronger dependence on lesion size for PL-RDP than OSEM. Lesion detectability was improved using time-of-flight (TOF) reconstruction, with greater benefit for the liver compared to the lung and with increasing benefit for decreasing lesion size and contrast. PL detectability was statistically significantly higher than OSEM for 20 mm liver lesions when contrast was [Formula: see text] ([Formula: see text]), and TOF PL detectability was statistically significantly higher than TOF OSEM for 15 and 20 mm liver lesions with contrast [Formula: see text] and [Formula: see text], respectively. For all other cases, there was no statistically significant difference between PL and OSEM ([Formula: see text]). For the range of studied lesion properties, lesion detectability using PL-RDP was equivalent or improved compared to using OSEM.

Pulmonary imaging using respiratory motion compensated simultaneous PET/MR

PURPOSE

Pulmonary positron emission tomography (PET) imaging is confounded by blurring artifacts caused by respiratory motion. These artifacts degrade both image quality and quantitative accuracy. In this paper, the authors present a complete data acquisition and processing framework for respiratory motion compensated image reconstruction (MCIR) using simultaneous whole body PET/magnetic resonance (MR) and validate it through simulation and clinical patient studies.

METHODS

The authors have developed an MCIR framework based on maximum a posteriori or MAP estimation. For fast acquisition of high quality 4D MR images, the authors developed a novel Golden-angle RAdial Navigated Gradient Echo (GRANGE) pulse sequence and used it in conjunction with sparsity-enforcing k-t FOCUSS reconstruction. The authors use a 1D slice-projection navigator signal encapsulated within this pulse sequence along with a histogram-based gate assignment technique to retrospectively sort the MR and PET data into individual gates. The authors compute deformation fields for each gate via nonrigid registration. The deformation fields are incorporated into the PET data model as well as utilized for generating dynamic attenuation maps. The framework was validated using simulation studies on the 4D XCAT phantom and three clinical patient studies that were performed on the Biograph mMR, a simultaneous whole body PET/MR scanner.

RESULTS

The authors compared MCIR (MC) results with ungated (UG) and one-gate (OG) reconstruction results. The XCAT study revealed contrast-to-noise ratio (CNR) improvements for MC relative to UG in the range of 21%-107% for 14 mm diameter lung lesions and 39%-120% for 10 mm diameter lung lesions. A strategy for regularization parameter selection was proposed, validated using XCAT simulations, and applied to the clinical studies. The authors' results show that the MC image yields 19%-190% increase in the CNR of high-intensity features of interest affected by respiratory motion relative to UG and a 6%-51% increase relative to OG.

CONCLUSIONS

Standalone MR is not the traditional choice for lung scans due to the low proton density, high magnetic susceptibility, and low T2 (∗) relaxation time in the lungs. By developing and validating this PET/MR pulmonary imaging framework, the authors show that simultaneous PET/MR, unique in its capability of combining structural information from MR with functional information from PET, shows promise in pulmonary imaging.

A novel approach to assess the treatment response using Gaussian random field in PET

PURPOSE

The assessment of early therapeutic response to anticancer therapy is vital for treatment planning and patient management in clinic. With the development of personal treatment plan, the early treatment response, especially before any anatomically apparent changes after treatment, becomes urgent need in clinic. Positron emission tomography (PET) imaging serves an important role in clinical oncology for tumor detection, staging, and therapy response assessment. Many studies on therapy response involve interpretation of differences between two PET images, usually in terms of standardized uptake values (SUVs). However, the quantitative accuracy of this measurement is limited. This work proposes a statistically robust approach for therapy response assessment based on Gaussian random field (GRF) to provide a statistically more meaningful scale to evaluate therapy effects.

METHODS

The authors propose a new criterion for therapeutic assessment by incorporating image noise into traditional SUV method. An analytical method based on the approximate expressions of the Fisher information matrix was applied to model the variance of individual pixels in reconstructed images. A zero mean unit variance GRF under the null hypothesis (no response to therapy) was obtained by normalizing each pixel of the post-therapy image with the mean and standard deviation of the pretherapy image. The performance of the proposed method was evaluated by Monte Carlo simulation, where XCAT phantoms (128(2) pixels) with lesions of various diameters (2-6 mm), multiple tumor-to-background contrasts (3-10), and different changes in intensity (6.25%-30%) were used. The receiver operating characteristic curves and the corresponding areas under the curve were computed for both the proposed method and the traditional methods whose figure of merit is the percentage change of SUVs. The formula for the false positive rate (FPR) estimation was developed for the proposed therapy response assessment utilizing local average method based on random field. The accuracy of the estimation was validated in terms of Euler distance and correlation coefficient.

RESULTS

It is shown that the performance of therapy response assessment is significantly improved by the introduction of variance with a higher area under the curve (97.3%) than SUVmean (91.4%) and SUVmax (82.0%). In addition, the FPR estimation serves as a good prediction for the specificity of the proposed method, consistent with simulation outcome with ∼1 correlation coefficient.

CONCLUSIONS

In this work, the authors developed a method to evaluate therapy response from PET images, which were modeled as Gaussian random field. The digital phantom simulations demonstrated that the proposed method achieved a large reduction in statistical variability through incorporating knowledge of the variance of the original Gaussian random field. The proposed method has the potential to enable prediction of early treatment response and shows promise for application to clinical practice. In future work, the authors will report on the robustness of the estimation theory for application to clinical practice of therapy response evaluation, which pertains to binary discrimination tasks at a fixed location in the image such as detection of small and weak lesion.

Evaluation of reconstruction errors and identification of artefacts for JET gamma and neutron tomography

The Joint European Torus (JET) neutron profile monitor ensures 2D coverage of the gamma and neutron emissive region that enables tomographic reconstruction. Due to the availability of only two projection angles and to the coarse sampling, tomographic inversion is a limited data set problem. Several techniques have been developed for tomographic reconstruction of the 2-D gamma and neutron emissivity on JET, but the problem of evaluating the errors associated with the reconstructed emissivity profile is still open. The reconstruction technique based on the maximum likelihood principle, that proved already to be a powerful tool for JET tomography, has been used to develop a method for the numerical evaluation of the statistical properties of the uncertainties in gamma and neutron emissivity reconstructions. The image covariance calculation takes into account the additional techniques introduced in the reconstruction process for tackling with the limited data set (projection resampling, smoothness regularization depending on magnetic field). The method has been validated by numerically simulations and applied to JET data. Different sources of artefacts that may significantly influence the quality of reconstructions and the accuracy of variance calculation have been identified.

Estimation of noise properties for TV-regularized image reconstruction in computed tomography

A method for predicting the image covariance resulting from total-variation-penalized iterative image reconstruction (TV-penalized IIR) is presented and demonstrated in a variety of contexts. The method is validated against the sample covariance from statistical noise realizations for a small image using a variety of comparison metrics. Potential applications for the covariance approximation include investigation of image properties such as object- and signal-dependence of noise, and noise stationarity. These applications are demonstrated, along with the construction of image pixel variance maps for two-dimensional 128 × 128 pixel images. Methods for extending the proposed covariance approximation to larger images and improving computational efficiency are discussed. Future work will apply the developed methodology to the construction of task-based image quality metrics such as the Hotelling observer detectability for TV-based IIR.

Monte Carlo modeling of a conventional X-ray computed tomography scanner for gel dosimetry purposes

Our purpose in the current study was to model an X-ray CT scanner with the Monte Carlo (MC) method for gel dosimetry. In this study, a conventional CT scanner with one array detector was modeled with use of the MCNPX MC code. The MC calculated photon fluence in detector arrays was used for image reconstruction of a simple water phantom as well as polyacrylamide polymer gel (PAG) used for radiation therapy. Image reconstruction was performed with the filtered back-projection method with a Hann filter and the Spline interpolation method. Using MC results, we obtained the dose-response curve for images of irradiated gel at different absorbed doses. A spatial resolution of about 2 mm was found for our simulated MC model. The MC-based CT images of the PAG gel showed a reliable increase in the CT number with increasing absorbed dose for the studied gel. Also, our results showed that the current MC model of a CT scanner can be used for further studies on the parameters that influence the usability and reliability of results, such as the photon energy spectra and exposure techniques in X-ray CT gel dosimetry.

Quantitative comparison of OSEM and penalized likelihood image reconstruction using relative difference penalties for clinical PET

Ordered subset expectation maximization (OSEM) is the most widely used algorithm for clinical PET image reconstruction. OSEM is usually stopped early and post-filtered to control image noise and does not necessarily achieve optimal quantitation accuracy. As an alternative to OSEM, we have recently implemented a penalized likelihood (PL) image reconstruction algorithm for clinical PET using the relative difference penalty with the aim of improving quantitation accuracy without compromising visual image quality. Preliminary clinical studies have demonstrated visual image quality including lesion conspicuity in images reconstructed by the PL algorithm is better than or at least as good as that in OSEM images. In this paper we evaluate lesion quantitation accuracy of the PL algorithm with the relative difference penalty compared to OSEM by using various data sets including phantom data acquired with an anthropomorphic torso phantom, an extended oval phantom and the NEMA image quality phantom; clinical data; and hybrid clinical data generated by adding simulated lesion data to clinical data. We focus on mean standardized uptake values and compare them for PL and OSEM using both time-of-flight (TOF) and non-TOF data. The results demonstrate improvements of PL in lesion quantitation accuracy compared to OSEM with a particular improvement in cold background regions such as lungs.

Quantitation of myocardial blood flow and myocardial flow reserve with 99mTc-sestamibi dynamic SPECT/CT to enhance detection of coronary artery disease

PURPOSE

Conventional dual-head single photon emission computed tomography (SPECT)/CT systems capable of fast dynamic SPECT (DySPECT) imaging have a potential for flow quantitation. This study introduced a new method to quantify myocardial blood flow (MBF) and myocardial flow reserve (MFR) with DySPECT scan and evaluated the diagnostic performance of detecting coronary artery disease (CAD) compared with perfusion using invasive coronary angiography (CAG) as the reference standard.

METHODS

This study included 21 patients with suspected or known CAD who had received DySPECT, ECG-gated SPECT (GSPECT), and CAG (13 with ≥ 50% stenosis in any vessel; non-CAD group: 8 with patent arteries or < 50% stenosis). DySPECT and GSPECT scans were performed on a widely used dual-head SPECT/CT scanner. The DySPECT imaging protocol utilized 12-min multiple back-and-forth gantry rotations during injections of (99m)Tc-sestamibi (MIBI) tracer at rest or dipyridamole-stress stages. DySPECT images were reconstructed with full physical corrections and converted to the physical unit of becquerels per milliliter. Stress MBF (SMBF), rest MBF (RMBF), and MFR were quantified by a one-tissue compartment flow model using time-activity curves derived from DySPECT images. Perfusion images were processed for GSPECT scan and interpreted to obtain summed stress score (SSS) and summed difference score (SDS). Receiver-operating characteristic (ROC) analyses were conducted to evaluate the diagnostic performance of flow and perfusion.

RESULTS

Using the criteria of ≥ 50% stenosis as positive CAD, areas under the ROC curve (AUCs) of flow assessment were overall significantly greater than those of perfusion. For patient-based analysis, AUCs for MFR, SMBF, SSS, and SDS were 0.91 ± 0.07, 0.86 ± 0.09, 0.64 ± 0.12, and 0.59 ± 0.13. For vessel-based analysis, AUCs for MFR, SMBF, SSS, and SDS were 0.81 ± 0.05, 0.76 ± 0.06, 0.62 ± 0.07, and 0.56 ± 0.08, respectively.

CONCLUSION

The preliminary data suggest that MBF quantitation with a conventional SPECT/CT system and the flow quantitation method is a clinically effective approach to enhance CAD detection.

Task-based detectability in CT image reconstruction by filtered backprojection and penalized likelihood estimation

PURPOSE

Nonstationarity is an important aspect of imaging performance in CT and cone-beam CT (CBCT), especially for systems employing iterative reconstruction. This work presents a theoretical framework for both filtered-backprojection (FBP) and penalized-likelihood (PL) reconstruction that includes explicit descriptions of nonstationary noise, spatial resolution, and task-based detectability index. Potential utility of the model was demonstrated in the optimal selection of regularization parameters in PL reconstruction.

METHODS

Analytical models for local modulation transfer function (MTF) and noise-power spectrum (NPS) were investigated for both FBP and PL reconstruction, including explicit dependence on the object and spatial location. For FBP, a cascaded systems analysis framework was adapted to account for nonstationarity by separately calculating fluence and system gains for each ray passing through any given voxel. For PL, the point-spread function and covariance were derived using the implicit function theorem and first-order Taylor expansion according to Fessler ["Mean and variance of implicitly defined biased estimators (such as penalized maximum likelihood): Applications to tomography," IEEE Trans. Image Process. 5(3), 493-506 (1996)]. Detectability index was calculated for a variety of simple tasks. The model for PL was used in selecting the regularization strength parameter to optimize task-based performance, with both a constant and a spatially varying regularization map.

RESULTS

Theoretical models of FBP and PL were validated in 2D simulated fan-beam data and found to yield accurate predictions of local MTF and NPS as a function of the object and the spatial location. The NPS for both FBP and PL exhibit similar anisotropic nature depending on the pathlength (and therefore, the object and spatial location within the object) traversed by each ray, with the PL NPS experiencing greater smoothing along directions with higher noise. The MTF of FBP is isotropic and independent of location to a first order approximation, whereas the MTF of PL is anisotropic in a manner complementary to the NPS. Task-based detectability demonstrates dependence on the task, object, spatial location, and smoothing parameters. A spatially varying regularization "map" designed from locally optimal regularization can improve overall detectability beyond that achievable with the commonly used constant regularization parameter.

CONCLUSIONS

Analytical models for task-based FBP and PL reconstruction are predictive of nonstationary noise and resolution characteristics, providing a valuable framework for understanding and optimizing system performance in CT and CBCT.

Non-local means denoising of dynamic PET images

Objective

Dynamic positron emission tomography (PET), which reveals information about both the spatial distribution and temporal kinetics of a radiotracer, enables quantitative interpretation of PET data. Model-based interpretation of dynamic PET images by means of parametric fitting, however, is often a challenging task due to high levels of noise, thus necessitating a denoising step. The objective of this paper is to develop and characterize a denoising framework for dynamic PET based on non-local means (NLM).

Theory

NLM denoising computes weighted averages of voxel intensities assigning larger weights to voxels that are similar to a given voxel in terms of their local neighborhoods or patches. We introduce three key modifications to tailor the original NLM framework to dynamic PET. Firstly, we derive similarities from less noisy later time points in a typical PET acquisition to denoise the entire time series. Secondly, we use spatiotemporal patches for robust similarity computation. Finally, we use a spatially varying smoothing parameter based on a local variance approximation over each spatiotemporal patch.

Methods

To assess the performance of our denoising technique, we performed a realistic simulation on a dynamic digital phantom based on the Digimouse atlas. For experimental validation, we denoised PET images from a mouse study and a hepatocellular carcinoma patient study. We compared the performance of NLM denoising with four other denoising approaches – Gaussian filtering, PCA, HYPR, and conventional NLM based on spatial patches.

Results

The simulation study revealed significant improvement in bias-variance performance achieved using our NLM technique relative to all the other methods. The experimental data analysis revealed that our technique leads to clear improvement in contrast-to-noise ratio in Patlak parametric images generated from denoised preclinical and clinical dynamic images, indicating its ability to preserve image contrast and high intensity details while lowering the background noise variance.

Quantitative methods for molecular diagnostic and therapeutic imaging

This theme issue provides an overview on the basic quantitative methods, an in-depth discussion on the cutting-edge quantitative analysis approaches as well as their applications for both static and dynamic molecular diagnostic and therapeutic imaging.