Observer studies of image quality of denoising reduced-count cardiac single photon emission computed tomography myocardial perfusion imaging by three-dimensional Gaussian post-reconstruction filtering and deep learning.

BACKGROUND: The aim of this research was to asses perfusion-defect detection-accuracy by human observers as a function of reduced-counts for 3D Gaussian post-reconstruction filtering vs deep learning (DL) denoising to determine if there was improved performance with DL. METHODS: SPECT projection data of 156 normally interpreted patients were used for these studies. Half were altered to include hybrid perfusion defects with defect presence and location known. Ordered-subset expectation-maximization (OSEM) reconstruction was employed with the optional correction of attenuation (AC) and scatter (SC) in addition to distance-dependent resolution (RC). Count levels varied from full-counts (100%) to 6.25% of full-counts. The denoising strategies were previously optimized for defect detection using total perfusion deficit (TPD). Four medical physicist (PhD) and six physician (MD) observers rated the slices using a graphical user interface. Observer ratings were analyzed using the LABMRMC multi-reader, multi-case receiver-operating-characteristic (ROC) software to calculate and compare statistically the area-under-the-ROC-curves (AUCs). RESULTS: For the same count-level no statistically significant increase in AUCs for DL over Gaussian denoising was determined when counts were reduced to either the 25% or 12.5% of full-counts. The average AUC for full-count OSEM with solely RC and Gaussian filtering was lower than for the strategies with AC and SC, except for a reduction to 6.25% of full-counts, thus verifying the utility of employing AC and SC with RC. CONCLUSION: We did not find any indication that at the dose levels investigated and with the DL network employed, that DL denoising was superior in AUC to optimized 3D post-reconstruction Gaussian filtering.

Improving detection accuracy of perfusion defect in standard dose SPECT-myocardial perfusion imaging by deep-learning denoising.

BACKGROUND: We previously developed a deep-learning (DL) network for image denoising in SPECT-myocardial perfusion imaging (MPI). Here we investigate whether this DL network can be utilized for improving detection of perfusion defects in standard-dose clinical acquisitions. METHODS: To quantify perfusion-defect detection accuracy, we conducted a receiver-operating characteristic (ROC) analysis on reconstructed images with and without processing by the DL network using a set of clinical SPECT-MPI data from 190 subjects. For perfusion-defect detection hybrid studies were used as ground truth, which were created from clinically normal studies with simulated realistic lesions inserted. We considered ordered-subset expectation-maximization (OSEM) reconstruction with corrections for attenuation, resolution, and scatter and with 3D Gaussian post-filtering. Total perfusion deficit (TPD) scores, computed by Quantitative Perfusion SPECT (QPS) software, were used to evaluate the reconstructed images. RESULTS: Compared to reconstruction with optimal Gaussian post-filtering (sigma = 1.2 voxels), further DL denoising increased the area under the ROC curve (AUC) from 0.80 to 0.88 (P-value < 10-4). For reconstruction with less Gaussian post-filtering (sigma = 0.8 voxels), thus better spatial resolution, DL denoising increased the AUC value from 0.78 to 0.86 (P-value < 10-4) and achieved better spatial resolution in reconstruction. CONCLUSIONS: DL denoising can effectively improve the detection of abnormal defects in standard-dose SPECT-MPI images over conventional reconstruction.

Retrospective fractional dose reduction in Tc-99m cardiac perfusion SPECT/CT patients: A human and model observer study.

BACKGROUND: In the ongoing efforts to reduce cardiac perfusion dose (injected radioactivity) for conventional SPECT/CT systems, we performed a human observer study to confirm our clinical model observer findings that iterative reconstruction employing OSEM (ordered-subset expectation-maximization) at 25% of the full dose (quarter-dose) has a similar performance for detection of hybrid cardiac perfusion defects as FBP at full dose. METHODS: One hundred and sixty-six patients, who underwent routine rest-stress Tc-99m sestamibi cardiac perfusion SPECT/CT imaging and clinically read as normally perfused, were included in the study. Ground truth was established by the normal read and the insertion of hybrid defects. In addition to the reconstruction of the 25% of full-dose data using OSEM with attenuation (AC), scatter (SC), and spatial resolution correction (RC), FBP and OSEM (with AC, SC, and RC) both at full dose (100%) were done. Both human observer and clinical model observer confidence scores were obtained to generate receiver operating characteristics (ROC) curves in a task-based image quality assessment. RESULTS: Average human observer AUC (area under the ROC curve) values of 0.725, 0.876, and 0.890 were obtained for FBP at full dose, OSEM at 25% of full dose, and OSEM at full dose, respectively. Both OSEM strategies were significantly better than FBP with P values of 0.003 and 0.01 respectively, while no significant difference was recorded between OSEM methods (P = 0.48). The clinical model observer results were 0.791, 0.822, and 0.879, respectively, for the same patient cases and processing strategies used in the human observer study. CONCLUSIONS: Cardiac perfusion SPECT/CT using OSEM reconstruction at 25% of full dose has AUCs larger than FBP and closer to those of full-dose OSEM when read by human observers, potentially replacing the higher dose studies during clinical reading.

Evaluation of the effect of reducing administered activity on assessment of function in cardiac gated SPECT.

BACKGROUND: We previously optimized several reconstruction strategies in SPECT myocardial perfusion imaging (MPI) with low dose for perfusion-defect detection. Here we investigate whether reducing the administered activity can also maintain the diagnostic accuracy in evaluating cardiac function. METHODS: We quantified the myocardial motion in cardiac-gated stress 99m-Tc-sestamibi SPECT studies from 163 subjects acquired with full dose (29.8 ± 3.6 mCi), and evaluated the agreement of the obtained motion/thickening and ejection fraction (EF) measures at various reduced dose levels (uniform reduction or personalized dose) with that at full dose. We also quantified the detectability of abnormal motion via a receiver-operating characteristics (ROC) study. For reconstruction we considered both filtered backprojection (FBP) without correction for degradations, and iterative ordered-subsets expectation-maximization (OS-EM) with resolution, attenuation and scatter corrections. RESULTS: With dose level lowered to 25% of full dose, the obtained results on motion/thickening, EF and abnormal motion detection were statistically comparable to full dose in both reconstruction strategies, with Pearson's r > 0.9 for global motion measures between low dose and full dose. CONCLUSIONS: The administered activity could be reduced to 25% of full dose without degrading the function assessment performance. Low dose reconstruction optimized for perfusion-defect detection can be reasonable for function assessment in gated SPECT.

Prediction of cardiac death after adenosine myocardial perfusion SPECT based on machine learning.

BACKGROUND: We developed machine-learning (ML) models to estimate a patient's risk of cardiac death based on adenosine myocardial perfusion SPECT (MPS) and associated clinical data, and compared their performance to baseline logistic regression (LR). We demonstrated an approach to visually convey the reasoning behind a patient's risk to provide insight to clinicians beyond that of a "black box." METHODS: We trained multiple models using 122 potential clinical predictors (features) for 8321 patients, including 551 cases of subsequent cardiac death. Accuracy was measured by area under the ROC curve (AUC), computed within a cross-validation framework. We developed a method to display the model's rationale to facilitate clinical interpretation. RESULTS: The baseline LR (AUC = 0.76; 14 features) was outperformed by all other methods. A least absolute shrinkage and selection operator (LASSO) model (AUC = 0.77; p = .045; 6 features) required the fewest features. A support vector machine (SVM) model (AUC = 0.83; p < .0001; 49 features) provided the highest accuracy. CONCLUSIONS: LASSO outperformed LR in both accuracy and simplicity (number of features), with SVM yielding best AUC for prediction of cardiac death in patients undergoing MPS. Combined with presenting the reasoning behind the risk scores, our results suggest that ML can be more effective than LR for this application.

Improving perfusion defect detection with respiratory motion correction in cardiac SPECT at standard and reduced doses.

BACKGROUND: In cardiac SPECT perfusion imaging, respiratory motion can cause non-uniform blurring in the reconstructed myocardium. We investigate the potential benefit of respiratory correction with respiratory-binned acquisitions, both at standard dose and at reduced dose, for defect detection and for left ventricular (LV) wall resolution. METHODS: We applied two reconstruction methods for respiratory motion correction: post-reconstruction motion correction (PMC) and motion-compensated reconstruction (MCR), and compared with reconstruction without motion correction (Non-MC). We quantified the presence of perfusion defects in reconstructed images by using the total perfusion deficit (TPD) scores and conducted receiver-operating-characteristic (ROC) studies using TPD. We quantified the LV spatial resolution by using the FWHM of its cross-sectional intensity profile. RESULTS: The values in the area-under-the-ROC-curve (AUC) achieved by MCR, PMC, and Non-MC at standard dose were 0.835, 0.830, and 0.798, respectively. Similar AUC improvements were also obtained by MCR and PMC over Non-MC at 50%, 25%, and 12.5% of full dose. Improvements in LV resolution were also observed with motion correction. CONCLUSIONS: Respiratory-binned acquisitions can improve perfusion-defect detection accuracy over traditional reconstruction both at standard dose and at reduced dose. Motion correction may contribute to achieving further dose reduction while maintaining the diagnostic accuracy of traditional acquisitions.

Investigation of dose reduction in cardiac perfusion SPECT via optimization and choice of the image reconstruction strategy.

BACKGROUND: We investigated the extent to which the administered dose (activity) level can be reduced without sacrificing diagnostic accuracy for three reconstruction strategies for SPECT-myocardial perfusion imaging (MPI). METHODS: We optimized the parameters of the three reconstruction strategies for perfusion-defect detection over a range of simulated administered dose levels using a set of hybrid studies (derived from 190 subjects) consisting of clinical SPECT-MPI data modified to contain realistic simulated lesions. The optimized strategies we considered are filtered backprojection (FBP) with no correction for degradations, ordered-subsets expectation-maximization (OS-EM) with attenuation correction (AC), scatter correction (SC), and resolution correction (RC), and OS-EM with scatter and resolution correction only. Each study was evaluated using a total perfusion deficit (TPD) score computed by the Quantitative Perfusion SPECT (QPS) software package. We conducted a receiver operating characteristics (ROC) study based on the TPD scores for each dose level and reconstruction strategy. RESULTS: For FBP, the achieved optimum values of the area under the ROC curve (AUC) at 100%, 50%, 25%, and 12.5% of standard dose were 0.75, 0.74, 0.72, and 0.70, respectively, compared to 0.81, 0.79, 0.76, and 0.74 for OS-EM with AC-SC-RC and 0.78, 0.77, 0.74, 0.72 for OS-EM with SC-RC. CONCLUSIONS: Our results suggest that studies reconstructed by OS-EM with AC-SC-RC could possibly be reduced, on average, to 25% of the originally administered dose without causing diagnostic accuracy (AUC) to decrease below that of FBP.

A comparison study of image features between FFDM and film mammogram images.

PURPOSE: This work is to provide a direct, quantitative comparison of image features measured by film and full-field digital mammography (FFDM). The purpose is to investigate whether there is any systematic difference between film and FFDM in terms of quantitative image features and their influence on the performance of a computer-aided diagnosis (CAD) system. METHODS: The authors make use of a set of matched film-FFDM image pairs acquired from cadaver breast specimens with simulated microcalcifications consisting of bone and teeth fragments using both a GE digital mammography system and a screen-film system. To quantify the image features, the authors consider a set of 12 textural features of lesion regions and six image features of individual microcalcifications (MCs). The authors first conduct a direct comparison on these quantitative features extracted from film and FFDM images. The authors then study the performance of a CAD classifier for discriminating between MCs and false positives (FPs) when the classifier is trained on images of different types (film, FFDM, or both). RESULTS: For all the features considered, the quantitative results show a high degree of correlation between features extracted from film and FFDM, with the correlation coefficients ranging from 0.7326 to 0.9602 for the different features. Based on a Fisher sign rank test, there was no significant difference observed between the features extracted from film and those from FFDM. For both MC detection and discrimination of FPs from MCs, FFDM had a slight but statistically significant advantage in performance; however, when the classifiers were trained on different types of images (acquired with FFDM or SFM) for discriminating MCs from FPs, there was little difference. CONCLUSIONS: The results indicate good agreement between film and FFDM in quantitative image features. While FFDM images provide better detection performance in MCs, FFDM and film images may be interchangeable for the purposes of training CAD algorithms, and a single CAD algorithm may be applied to either type of images.

Adaptation of a clustered lumpy background model for task-based image quality assessment in x-ray phase-contrast mammography.

PURPOSE: Since the introduction of clinical x-ray phase-contrast mammography (PCM), a technique that exploits refractive-index variations to create edge enhancement at tissue boundaries, a number of optimization studies employing physical image-quality metrics have been performed. Ideally, task-based assessment of PCM would have been conducted with human readers. These studies have been limited, however, in part due to the large parameter-space of PCM system configurations and the difficulty of employing expert readers for large-scale studies. It has been proposed that numerical observers can be used to approximate the statistical performance of human readers, thus enabling the study of task-based performance over a large parameter-space. METHODS: Methods are presented for task-based image quality assessment of PCM images with a numerical observer, the most significant of which is an adapted lumpy background from the conventional mammography literature that accounts for the unique wavefield propagation physics of PCM image formation and will be used with a numerical observer to assess image quality. These methods are demonstrated by performing a PCM task-based image quality study using a numerical observer. This study employs a signal-known-exactly, background-known-statistically Bayesian ideal observer method to assess the detectability of a calcification object in PCM images when the anode spot size and calcification diameter are varied. RESULTS: The first realistic model for the structured background in PCM images has been introduced. A numerical study demonstrating the use of this background model has compared PCM and conventional mammography detection of calcification objects. The study data confirm the strong PCM calcification detectability dependence on anode spot size. These data can be used to balance the trade-off between enhanced image quality and the potential for motion artifacts that comes with use of a reduced spot size and increased exposure time. CONCLUSIONS: A method has been presented for the incorporation of structured breast background data into task-based numerical observer assessment of PCM images. The method adapts conventional background simulation techniques to the wavefield propagation physics necessary for PCM imaging. This method is demonstrated with a simple detection task.

Measurement of neurodegeneration using a multivariate early frame amyloid PET classifier.

Introduction: Amyloid measurement provides important confirmation of pathology for Alzheimer's disease (AD) clinical trials. However, many amyloid positive (Am+) early-stage subjects do not worsen clinically during a clinical trial, and a neurodegenerative measure predictive of decline could provide critical information. Studies have shown correspondence between perfusion measured by early amyloid frames post-tracer injection and fluorodeoxyglucose (FDG) positron emission tomography (PET), but with limitations in sensitivity. Multivariate machine learning approaches may offer a more sensitive means for detection of disease related changes as we have demonstrated with FDG. Methods: Using summed dynamic florbetapir image frames acquired during the first 6 minutes post-injection for 107 Alzheimer's Disease Neuroimaging Initiative subjects, we applied optimized machine learning to develop and test image classifiers aimed at measuring AD progression. Early frame amyloid (EFA) classification was compared to that of an independently developed FDG PET AD progression classifier by scoring the FDG scans of the same subjects at the same time point. Score distributions and correlation with clinical endpoints were compared to those obtained from FDG. Region of interest measures were compared between EFA and FDG to further understand discrimination performance. Results: The EFA classifier produced a primary pattern similar to that of the FDG classifier whose expression correlated highly with the FDG pattern (R-squared 0.71), discriminated cognitively normal (NL) amyloid negative (Am-) subjects from all Am+ groups, and that correlated in Am+ subjects with Mini-Mental State Examination, Clinical Dementia Rating Sum of Boxes, and Alzheimer's Disease Assessment Scale-13-item Cognitive subscale (R = 0.59, 0.63, 0.73) and with subsequent 24-month changes in these measures (R = 0.67, 0.73, 0.50). Discussion: Our results support the ability to use EFA with a multivariate machine learning-derived classifier to obtain a sensitive measure of AD-related loss in neuronal function that correlates with FDG PET in preclinical and early prodromal stages as well as in late mild cognitive impairment and dementia. Highlights: The summed initial post-injection minutes of florbetapir positron emission tomography  correlate with fluorodeoxyglucose.A machine learning classifier enabled sensitive detection of early prodromal Alzheimer's disease.Early frame amyloid (EFA) classifier scores correlate with subsequent change in Mini-Mental State Examination, Clinical Dementia Rating Sum of Boxes, and Alzheimer's Disease Assessment Scale-13-item Cognitive subscale.EFA classifier effect sizes and clinical prediction outperformed region of interest standardized uptake value ratio.EFA classification may aid in stratifying patients to assess treatment effect.

Dimensionality estimation for optimal detection of functional networks in BOLD fMRI data.

Estimation of the intrinsic dimensionality of fMRI data is an important part of data analysis that helps to separate the signal of interest from noise. We have studied multiple methods of dimensionality estimation proposed in the literature and used these estimates to select a subset of principal components that was subsequently processed by linear discriminant analysis (LDA). Using simulated multivariate Gaussian data, we show that the dimensionality that optimizes signal detection (in terms of the receiver operating characteristic (ROC) metric) goes through a transition from many dimensions to a single dimension as a function of the signal-to-noise ratio. This transition happens when the loci of activation are organized into a spatial network and the variance of the networked, task-related signals is high enough for the signal to be easily detected in the data. We show that reproducibility of activation maps is a metric that captures this switch in intrinsic dimensionality. Except for reproducibility, all of the methods of dimensionality estimation we considered failed to capture this transition: optimization of Bayesian evidence, minimum description length, supervised and unsupervised LDA prediction, and Stein's unbiased risk estimator. This failure results in sub-optimal ROC performance of LDA in the presence of a spatially distributed network, and may have caused LDA to underperform in many of the reported comparisons in the literature. Using real fMRI data sets, including multi-subject group and within-subject longitudinal analysis we demonstrate the existence of these dimensionality transitions in real data.

Effects of motion, attenuation, and scatter corrections on gated cardiac SPECT reconstruction.

PURPOSE: In gated cardiac single photon emission computed tomography (SPECT), image reconstruction is often hampered by various degrading factors including depth-dependent spatial blurring, attenuation, scatter, motion blurring, and low data counts. Consequently, there has been significant development in image reconstruction methods for improving the quality of reconstructed images. The goal of this work is to investigate how these degrading factors will impact the reconstructed myocardium when different reconstruction methods are used. METHODS: The authors conduct a comparative study of the effects of these degrading factors on the accuracy of myocardium by several reconstruction algorithms, including (1) a clinical spatiotemporal processing method, (2) maximum likelihood (ML) estimation, (3) 3D maximum a posteriori (MAP) estimation, (4) 3D MAP with posttemporal filtering, and (5) motion-compensated spatiotemporal (4D) reconstruction. To quantify the reconstruction results, the authors use the following measures on different aspects of the myocardium: (1) overall error level in the myocardium, (2) regional accuracy of the left ventricle (LV) wall, (3) uniformity of the LV, (4) accuracy of regional time activity curves by normalized cross-correlation coefficient, and (5) perfusion defect detectability. The authors also assess the effectiveness of degrading corrections in reconstruction by considering an upper bound for each reconstruction method, which represents what would be achieved by each method if the acquired data were free from attenuation and scatter degradations. In the experiments the authors use Monte Carlo simulated cardiac gated SPECT imaging based on the 4D NURBS-based cardiac-torso (NCAT) phantom with different patient geometry and lesion settings, in which the simulated ground truth is known for the purpose of quantitative evaluation. RESULTS: The results demonstrate that use of temporal processing in reconstruction (Methods 1, 4, and 5 above) can greatly improve the reconstructed myocardium in terms of both error level and perfusion defect detection. In low-count gated studies, it can have even greater impact than other degrading factors. Both attenuation and scatter corrections can lead to reduced error levels in the myocardium in all methods; in particular, with 4D the bias can be reduced by as much as four-fold compared to no correction. There is a slight increase in noise level observed with scatter correction. A significant improvement in heart wall appearance is demonstrated in reconstruction results from three sets of clinical acquisitions as correction for degradations is combined with refinement of temporal filtering. CONCLUSIONS: Correction for degrading factors such as resolution, attenuation, scatter, and motion blur can all lead to improved image quality in cardiac gated SPECT reconstruction. However, their effectiveness could also vary with the reconstruction algorithms used. Both attenuation and scatter corrections can effectively reduce the bias level of the reconstructed LV wall, though scatter correction is also observed to increase the variance level. Use of temporal processing in reconstruction can have greater impact on the accuracy of the myocardium than correction of other degrading factors. Overall, use of degrading corrections in 4D reconstruction is shown to be most effective for improving both reconstruction accuracy of the myocardium and detectability of perfusion defects in gated images.

Deep learning with noise-to-noise training for denoising in SPECT myocardial perfusion imaging.

PURPOSE: Post-reconstruction filtering is often applied for noise suppression due to limited data counts in myocardial perfusion imaging (MPI) with single-photon emission computed tomography (SPECT). We study a deep learning (DL) approach for denoising in conventional SPECT-MPI acquisitions, and investigate whether it can be more effective for improving the detectability of perfusion defects compared to traditional postfiltering. METHODS: Owing to the lack of ground truth in clinical studies, we adopt a noise-to-noise (N2N) training approach for denoising in SPECT-MPI images. We consider a coupled U-Net (CU-Net) structure which is designed to improve learning efficiency through feature map reuse. For network training we employ a bootstrap procedure to generate multiple noise realizations from list-mode clinical acquisitions. In the experiments we demonstrated the proposed approach on a set of 895 clinical studies, where the iterative OSEM algorithm with three-dimensional (3D) Gaussian postfiltering was used to reconstruct the images. We investigated the detection performance of perfusion defects in the reconstructed images using the non-prewhitening matched filter (NPWMF), evaluated the uniformity of left ventricular (LV) wall in terms of image intensity, and quantified the effect of smoothing on the spatial resolution of the reconstructed LV wall by using its full-width at half-maximum (FWHM). RESULTS: Compared to OSEM with Gaussian postfiltering, the DL denoised images with CU-Net significantly improved the detection performance of perfusion defects at all contrast levels (65%, 50%, 35%, and 20%). The signal-to-noise ratio (SNRD ) in the NPWMF output was increased on average by 8% over optimal Gaussian smoothing (P < 10-4 , paired t-test), while the inter-subject variability was greatly reduced. The CU-Net also outperformed a 3D nonlocal means (NLM) filter and a convolutional autoencoder (CAE) denoising network in terms of SNRD . In addition, the FWHM of the LV wall in the reconstructed images was varied by less than 1%. Furthermore, CU-Net also improved the detection performance when the images were processed with less post-reconstruction smoothing (a trade-off of increased noise for better LV resolution), with SNRD improved on average by 23%. CONCLUSIONS: The proposed DL with N2N training approach can yield additional noise suppression in SPECT-MPI images over conventional postfiltering. For perfusion defect detection, DL with CU-Net could outperform conventional 3D Gaussian filtering with optimal setting as well as NLM and CAE.

Detectability of perfusion defect in five-dimensional gated-dynamic cardiac SPECT images.

PURPOSE: In previous work, the authors developed and demonstrated the concept of an image reconstruction procedure aimed to unify gated and dynamic nuclear cardiac imaging, which the authors have termed five-dimensional (5D) SPECT. Gated imaging permits the clinician to evaluate wall motion and, through the use of stress and rest scans, allows perfusion defects to be observed. Dynamic imaging depicts kinetics in the myocardium, which can be used to evaluate perfusion, but traditional dynamic images are motionless and do not depict wall motion. In this article, the authors investigate the degree to which perfusion defects can be detected from the dynamic information conveyed by 5D images, a problem that is particularly challenging in the absence of multiple fast camera rotations. METHODS: The authors first demonstrate the usefulness of dynamic reconstructed images for perfusion detection by using linear discriminant analyses (Fisher linear discriminant analysis and principal component analysis) and a numerical channelized Hotelling observer. The authors then derive three types of discriminant metrics for characterizing the temporal kinetic information in reconstructed dynamic images for differentiating perfusion defects from normal cardiac perfusion, which are the Fisher linear discriminant map, temporal derivative map, and kinetic parametric images. RESULTS: Results are based on the NURBS-based cardiac-torso phantom with simulation of Tc99m-teboroxime as the imaging agent. The derived metric maps and quantitative contrast-to-noise ratio results demonstrate that the reconstructed dynamic images could yield higher detectability of the perfusion defect than conventional gated reconstruction while providing wall motion information simultaneously. CONCLUSIONS: The proposed metrics can be used to produce new types of visualizations, showing wall motion and perfusion information, that may potentially be useful for clinical evaluation. Since 5D imaging permits wall motion and kinetics to be observed simultaneously, it may ultimately obviate the need for separate stress and rest scans.

Supervised and unsupervised methods for prostate cancer segmentation with multispectral MRI.

PURPOSE: Magnetic resonance imaging (MRI) has been proposed as a promising alternative to transrectal ultrasound for the detection and localization of prostate cancer and fusing the information from multispectral MR images is currently an active research area. In this study, the goal is to develop automated methods that combine the pharmacokinetic parameters derived from dynamic contrast enhanced (DCE) MRI with quantitative T2 MRI and diffusion weighted imaging (DWI) in contrast to most of the studies which were performed with human readers. The main advantages of the automated methods are that the observer variability is removed and easily reproducible results can be efficiently obtained when the methods are applied to a test data. The goal is also to compare the performance of automated supervised and unsupervised methods for prostate cancer localization with multispectral MRI. METHODS: The authors use multispectral MRI data from 20 patients with biopsy-confirmed prostate cancer patients, and the image set consists of parameters derived from T2, DWI, and DCE-MRI. The authors utilize large margin classifiers for prostate cancer segmentation and compare them to an unsupervised method the authors have previously developed. The authors also develop thresholding schemes to tune support vector machines (SVMs) and their probabilistic counterparts, relevance vector machines (RVMs), for an improved performance with respect to a selected criterion. Moreover, the authors apply a thresholding method to make the unsupervised fuzzy Markov random fields method fully automatic. RESULTS: The authors have developed a supervised machine learning method that performs better than the previously developed unsupervised method and, additionally, have found that there is no significant difference between the SVM and RVM segmentation results. The results also show that the proposed methods for threshold selection can be used to tune the automated segmentation methods to optimize results for certain criteria such as accuracy or sensitivity. The test results of the automated algorithms indicate that using multispectral MRI improves prostate cancer segmentation performance when compared to single MR images, a result similar to the human reader studies that were performed before. CONCLUSIONS: The automated methods presented here can help diagnose and detect prostate cancer, and improve segmentation results. For that purpose, multispectral MRI provides better information about cancer and normal regions in the prostate when compared to methods that use single MRI techniques; thus, the different MRI measurements provide complementary information in the automated methods. Moreover, the use of supervised algorithms in such automated methods remain a good alternative to the use of unsupervised algorithms.

Regularized Fully 5D Reconstruction of Cardiac Gated Dynamic SPECT Images.

In our recent work, we proposed an image reconstruction procedure aimed to unify gated imaging and dynamic imaging in nuclear cardiac imaging. With this procedure the goal is to obtain an image sequence from a single acquisition which shows simultaneously both cardiac motion and tracer distribution change over the course of imaging. In this work, we further develop and demonstrate this procedure for fully 5D (3D space plus time plus gate) reconstruction in gated, dynamic cardiac SPECT imaging, where the challenge is even greater without the use of multiple fast camera rotations. For 5D reconstruction, we develop and compare two iterative algorithms: one is based on the modified block sequential regularized EM (BSREM-II) algorithm, and the other is based on the one-step late (OSL) algorithm. In our experiments, we simulated gated cardiac imaging with the NURBS-based cardiac-torso (NCAT) phantom and Tc99m-Teboroxime as the imaging agent, where acquisition with the equivalent of only three full camera rotations was used during the course of a 12-minute postinjection period. We conducted a thorough evaluation of the reconstruction results using a number of quantitative measures. Our results demonstrate that the 5D reconstruction procedure can yield gated dynamic images which show quantitative information for both perfusion defect detection and cardiac motion.

Improving Diagnostic Accuracy in Low-Dose SPECT Myocardial Perfusion Imaging With Convolutional Denoising Networks.

Lowering the administered dose in SPECT myocardial perfusion imaging (MPI) has become an important clinical problem. In this study we investigate the potential benefit of applying a deep learning (DL) approach for suppressing the elevated imaging noise in low-dose SPECT-MPI studies. We adopt a supervised learning approach to train a neural network by using image pairs obtained from full-dose (target) and low-dose (input) acquisitions of the same patients. In the experiments, we made use of acquisitions from 1,052 subjects and demonstrated the approach for two commonly used reconstruction methods in clinical SPECT-MPI: 1) filtered backprojection (FBP), and 2) ordered-subsets expectation-maximization (OSEM) with corrections for attenuation, scatter and resolution. We evaluated the DL output for the clinical task of perfusion-defect detection at a number of successively reduced dose levels (1/2, 1/4, 1/8, 1/16 of full dose). The results indicate that the proposed DL approach can achieve substantial noise reduction and lead to improvement in the diagnostic accuracy of low-dose data. In particular, at 1/2 dose, DL yielded an area-under-the-ROC-curve (AUC) of 0.799, which is nearly identical to the AUC = 0.801 obtained by OSEM at full-dose ( p -value = 0.73); similar results were also obtained for FBP reconstruction. Moreover, even at 1/8 dose, DL achieved AUC = 0.770 for OSEM, which is above the AUC = 0.755 obtained at full-dose by FBP. These results indicate that, compared to conventional reconstruction filtering, DL denoising can allow for additional dose reduction without sacrificing the diagnostic accuracy in SPECT-MPI.

A quantitative evaluation study of four-dimensional gated cardiac SPECT reconstruction.

In practice, gated cardiac SPECT images suffer from a number of degrading factors, including distance-dependent blur, attenuation, scatter and increased noise due to gating. Recently, we proposed a motion-compensated approach for four-dimensional (4D) reconstruction for gated cardiac SPECT and demonstrated that use of motion-compensated temporal smoothing could be effective for suppressing the increased noise due to lowered counts in individual gates. In this work, we further develop this motion-compensated 4D approach by also taking into account attenuation and scatter in the reconstruction process, which are two major degrading factors in SPECT data. In our experiments, we conducted a thorough quantitative evaluation of the proposed 4D method using Monte Carlo simulated SPECT imaging based on the 4D NURBS-based cardiac-torso (NCAT) phantom. In particular, we evaluated the accuracy of the reconstructed left ventricular myocardium using a number of quantitative measures including regional bias-variance analyses and wall intensity uniformity. The quantitative results demonstrate that use of motion-compensated 4D reconstruction can improve the accuracy of the reconstructed myocardium, which in turn can improve the detectability of perfusion defects. Moreover, our results reveal that while traditional spatial smoothing could be beneficial, its merit would become diminished with the use of motion-compensated temporal regularization. As a preliminary demonstration, we also tested our 4D approach on patient data. The reconstructed images from both simulated and patient data demonstrated that our 4D method can improve the definition of the LV wall.

Novel observations and potential applications using digital infrared iris imaging.

Digital infrared iris photography using a modified digital camera system was performed on approximately 300 subjects seen during routine clinical care and research at one facility. Because this image database offered an opportunity to gain new insight into the potential utility of infrared iris imaging, it was surveyed for unique image patterns. Then, a selection of photographs was compiled that would illustrate the spectrum of this imaging experience. Potentially informative image patterns were observed in subjects with cataracts, diabetic retinopathy, Posner-Schlossman syndrome, iridociliary cysts, long anterior lens zonules, nevi, oculocutaneous albinism, pigment dispersion syndrome, pseudophakia, suspected vascular anomaly, and trauma. Image patterns were often unanticipated regardless of preexisting information and suggest that infrared iris imaging may have numerous potential clinical and research applications, some of which may still not be recognized. These observations suggest further development and study of this technology.

Cardiac motion correction for improving perfusion defect detection in cardiac SPECT at standard and reduced doses of activity.

In cardiac SPECT perfusion imaging, cardiac motion can lead to motion blurring of anatomical detail and perfusion defects in the reconstructed myocardium. In this study, we investigated the potential benefit of cardiac motion correction for improving the detectability of perfusion defects. We considered a post-reconstruction motion correction (PMC) approach in which the image motion between two cardiac gates is obtained with optical flow estimation. In the experiments, we demonstrated the proposed post-reconstruction motion correction with optical flow estimation (PMC-OFE) approach on a set of clinical acquisitions from 194 subjects. We quantified the detectability of perfusion defects in the reconstructed images by using the total perfusion deficit scores, calculated by the clinical software tool QPS, and conducted a receiver-operating-characteristic (ROC) study to obtain the detection performance. Besides imaging with conventional standard dose, we also evaluated the approach for reduced dose SPECT imaging where the imaging dose was retrospectively reduced to 50%, 25%, and 12.5% of the standard dose. The proposed PMC-OFE approach achieved at each dose level higher area-under-the-ROC-curve (AUC) for perfusion defect detection than the traditional approach of using ungated data (Non-MC) (p -value < 0.05); in particular, with half dose, PMC-OFE achieved AUC = 0.813, which is comparable to Non-MC with standard dose (AUC = 0.795). Moreover, the proposed PMC-OFE approach also outperformed the 'Motion Frozen' (MF) method implemented in the clinical quantitative gated SPECT (QGS) software. In particular, at 25% and 12.5% of standard dose, the AUC values obtained by PMC-OFE are 0.788 and 0.779, respectively, compared to 0.758 and 0.731 for MF (p -value < 0.05).