ABSTRACT
Background: Worldwide, coronary artery disease (CAD) is a leading cause of mortality and morbidity and remains to be a top health priority in many countries. A non-invasive imaging modality for diagnosis of CAD such as single photon emission computed tomography-myocardial perfusion imaging (SPECT-MPI) is usually requested by cardiologists as it displays radiotracer distribution in the heart reflecting myocardial perfusion. The interpretation of SPECT-MPI is done visually by a nuclear medicine physician and is largely dependent on his clinical experience and showing significant inter-observer variability. Objective: The aim of the study is to apply a deep learning approach in the classification of SPECT-MPI for perfusion abnormalities using convolutional neural networks (CNN). Methods: A publicly available anonymized SPECT-MPI from a machine learning repository (https://www.kaggle.com/selcankaplan/spect-mpi) was used in this study involving 192 patients who underwent stress-test-rest Tc99m MPI. An exploratory approach of CNN hyperparameter selection to search for optimum neural network model was utilized with particular focus on various dropouts (0.2, 0.5, 0.7), batch sizes (8, 16, 32, 64), and number of dense nodes (32, 64, 128, 256). The base CNN model was also compared with the commonly used pre-trained CNNs in medical images such as VGG16, InceptionV3, DenseNet121 and ResNet50. All simulations experiments were performed in Kaggle using TensorFlow 2.6.0., Keras 2.6.0, and Python language 3.7.10. Results: The best performing base CNN model with parameters consisting of 0.7 dropout, batch size 8, and 32 dense nodes generated the highest normalized Matthews Correlation Coefficient at 0.909 and obtained 93.75% accuracy, 96.00% sensitivity, 96.00% precision, and 96.00% F1-score. It also obtained higher classification performance as compared to the pre-trained architectures. Conclusions: The results suggest that deep learning approaches through the use of CNN models can be deployed by nuclear medicine physicians in their clinical practice to further augment their decision skills in the interpretation of SPECT-MPI tests. These CNN models can also be used as a dependable and valid second opinion that can aid physicians as a decision-support tool as well as serve as teaching or learning materials for the less-experienced physicians particularly those still in their training career. These highlights the clinical utility of deep learning approaches through CNN models in the practice of nuclear cardiology.
ABSTRACT
BACKGROUND: The segment of the latest mechanical contraction (LMC) does not always overlap with the site of the latest electrical activation (LEA). By integrating both mechanical and electrical dyssynchrony, this proof-of-concept study aimed to propose a new method for recommending left ventricular (LV) lead placements, with the goal of enhancing response to cardiac resynchronization therapy (CRT). METHODS: The LMC segment was determined by single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) phase analysis. The LEA site was detected by vectorcardiogram. The recommended segments for LV lead placement were as follows: (1) the LMC viable segments that overlapped with the LEA site; (2) the LMC viable segments adjacent to the LEA site; (3) If no segment met either of the above, the LV lateral wall was recommended. The response was defined as ≥15% reduction in left ventricular end-systolic volume (LVESV) 6-months after CRT. Patients with LV lead located in the recommended site were assigned to the recommended group, and those located in the non-recommended site were assigned to the non-recommended group. RESULTS: The cohort comprised of 76 patients, including 54 (71.1%) in the recommended group and 22 (28.9%) in the non-recommended group. Among the recommended group, 74.1% of the patients responded to CRT, while 36.4% in the non-recommended group were responders (P = .002). Compared to pacing at the non-recommended segments, pacing at the recommended segments showed an independent association with an increased response by univariate and multivariable analysis (odds ratio 5.00, 95% confidence interval 1.73-14.44, P = .003; odds ratio 7.33, 95% confidence interval 1.53-35.14, P = .013). Kaplan-Meier curves showed that pacing at the recommended LV lead position demonstrated a better long-term prognosis. CONCLUSION: Our findings indicate that pacing at the recommended segments, by integrating of mechanical and electrical dyssynchrony, is significantly associated with an improved CRT response and better long-term prognosis.
Subject(s)
Cardiac Resynchronization Therapy , Heart Ventricles , Vectorcardiography , Humans , Cardiac Resynchronization Therapy/methods , Female , Male , Aged , Middle Aged , Vectorcardiography/methods , Treatment Outcome , Heart Ventricles/diagnostic imaging , Heart Failure/diagnostic imaging , Heart Failure/therapy , Cardiac-Gated Single-Photon Emission Computer-Assisted Tomography/methods , Myocardial Perfusion Imaging/methods , Proof of Concept Study , Tomography, Emission-Computed, Single-Photon , Cardiac Resynchronization Therapy DevicesABSTRACT
AIMS: Variation in diagnostic performance of single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) has been observed, yet the impact of cardiac size has not been well characterized. We assessed whether low left ventricular volume influences SPECT MPI's ability to detect obstructive coronary artery disease (CAD) and its interaction with age and sex. METHODS AND RESULTS: A total of 2066 patients without known CAD (67% male, 64.7 ± 11.2 years) across nine institutions underwent SPECT MPI with solid-state scanners followed by coronary angiography as part of the REgistry of Fast Myocardial Perfusion Imaging with NExt Generation SPECT. Area under receiver-operating characteristic curve (AUC) analyses evaluated the performance of quantitative and visual assessments according to cardiac size [end-diastolic volume (EDV); <20th vs. ≥20th population or sex-specific percentiles], age (<75 vs. ≥75 years), and sex. Significantly decreased performance was observed in patients with low EDV compared with those without (AUC: population 0.72 vs. 0.78, P = 0.03; sex-specific 0.72 vs. 0.79, P = 0.01) and elderly patients compared with younger patients (AUC 0.72 vs. 0.78, P = 0.03), whereas males and females demonstrated similar AUC (0.77 vs. 0.76, P = 0.67). The reduction in accuracy attributed to lower volumes was primarily observed in males (sex-specific threshold: EDV 0.69 vs. 0.79, P = 0.01). Accordingly, a significant decrease in AUC, sensitivity, specificity, and negative predictive value for quantitative and visual assessments was noted in patients with at least two characteristics of low EDV, elderly age, or male sex. CONCLUSION: Detection of CAD with SPECT MPI is negatively impacted by small cardiac size, most notably in elderly and male patients.
Subject(s)
Coronary Artery Disease , Myocardial Perfusion Imaging , Registries , Tomography, Emission-Computed, Single-Photon , Humans , Male , Female , Middle Aged , Myocardial Perfusion Imaging/methods , Aged , Tomography, Emission-Computed, Single-Photon/methods , Coronary Artery Disease/diagnostic imaging , Organ Size , Sex Factors , Coronary Angiography/methods , ROC Curve , Age Factors , Sensitivity and SpecificityABSTRACT
Aims: Current machine learning-based (ML) models usually attempt to utilize all available patient data to predict patient outcomes while ignoring the associated cost and time for data acquisition. The purpose of this study is to create a multi-stage machine learning model to predict cardiac resynchronization therapy (CRT) response for heart failure (HF) patients. This model exploits uncertainty quantification to recommend additional collection of single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) variables if baseline clinical variables and features from electrocardiogram (ECG) are not sufficient. Methods: 218 patients who underwent rest-gated SPECT MPI were enrolled in this study. CRT response was defined as an increase in left ventricular ejection fraction (LVEF) > 5% at a 6±1 month follow-up. A multi-stage ML model was created by combining two ensemble models: Ensemble 1 was trained with clinical variables and ECG; Ensemble 2 included Ensemble 1 plus SPECT MPI features. Uncertainty quantification from Ensemble 1 allowed for multi-stage decision-making to determine if the acquisition of SPECT data for a patient is necessary. The performance of the multi-stage model was compared with that of Ensemble models 1 and 2. Results: The response rate for CRT was 55.5% (n = 121) with overall male gender 61.0% (n = 133), an average age of 62.0±11.8, and LVEF of 27.7±11.0. The multi-stage model performed similarly to Ensemble 2 (which utilized the additional SPECT data) with AUC of 0.75 vs. 0.77, accuracy of 0.71 vs. 0.69, sensitivity of 0.70 vs. 0.72, and specificity 0.72 vs. 0.65, respectively. However, the multi-stage model only required SPECT MPI data for 52.7% of the patients across all folds. Conclusions: By using rule-based logic stemming from uncertainty quantification, the multi-stage model was able to reduce the need for additional SPECT MPI data acquisition without sacrificing performance.
ABSTRACT
Background@#Worldwide, coronary artery disease (CAD) is a leading cause of mortality and morbidity and remains to be a top health priority in many countries. A non-invasive imaging modality for diagnosis of CAD such as single photon emission computed tomography-myocardial perfusion imaging (SPECT-MPI) is usually requested by cardiologists as it displays radiotracer distribution in the heart reflecting myocardial perfusion. The interpretation of SPECT-MPI is done visually by a nuclear medicine physician and is largely dependent on his clinical experience and showing significant inter-observer variability.@*Objective@#The aim of the study is to apply a deep learning approach in the classification of SPECT-MPI for perfusion abnormalities using convolutional neural networks (CNN).@*Methods@#A publicly available anonymized SPECT-MPI from a machine learning repository (https://www.kaggle.com/ selcankaplan/spect-mpi) was used in this study involving 192 patients who underwent stress-test-rest Tc99m MPI. An exploratory approach of CNN hyperparameter selection to search for optimum neural network model was utilized with particular focus on various dropouts (0.2, 0.5, 0.7), batch sizes (8, 16, 32, 64), and number of dense nodes (32, 64, 128, 256). The base CNN model was also compared with the commonly used pre-trained CNNs in medical images such as VGG16, InceptionV3, DenseNet121 and ResNet50. All simulations experiments were performed in Kaggle using TensorFlow 2.6.0., Keras 2.6.0, and Python language 3.7.10.@*Results@#The best performing base CNN model with parameters consisting of 0.7 dropout, batch size 8, and 32 dense nodes generated the highest normalized Matthews Correlation Coefficient at 0.909 and obtained 93.75% accuracy, 96.00% sensitivity, 96.00% precision, and 96.00% F1-score. It also obtained higher classification performance as compared to the pre-trained architectures. @*Conclusions@#The results suggest that deep learning approaches through the use of CNN models can be deployed by nuclear medicine physicians in their clinical practice to further augment their decision skills in the interpretation of SPECT-MPI tests. These CNN models can also be used as a dependable and valid second opinion that can aid physicians as a decision-support tool as well as serve as teaching or learning materials for the less-experienced physicians particularly those still in their training career. These highlights the clinical utility of deep learning approaches through CNN models in the practice of nuclear cardiology.
Subject(s)
Coronary Artery Disease , Deep LearningABSTRACT
Objectives: We evaluated the impact of the COVID-19 pandemic on the number of referrals for SPECT myocardial perfusion imaging (SPECT-MPI) as well as changes in the clinical and imaging characteristics. Methods: We respectively reviewed 1042 SPECT-MPI cases performed in a 4-month period during the COVID-19 pandemic (PAN; n=423) and compared their findings with those acquired in the same months before the pandemic (PRE; n=619). Results: The number of stress SPECT-MPI studies performed during the PAN period significantly dropped compared to the number of studies carried out in the PRE period (p = 0.014). In the PRE period, the rates of patients presenting with non-anginal, atypical and typical chest pain were 31%, 25% and 19%, respectively. The figures significantly changed in the PAN period to 19%, 42%, and 11%, respectively (all p-values <0.001). Regarding the pretest probability of coronary artery disease (CAD), a significant decrease and increase were noticed in patients with high and intermediate pretest probability, respectively (PRE: 18% and 55%, PAN: 6% and 65%, p <0.001 and 0.008, respectively). Neither the rates of myocardial ischemia nor infarction differed significantly in the PRE vs. PAN study periods. Conclusion: The number of referrals dropped significantly in the PAN era. While the proportion of patients with intermediate risk for CAD being referred for SPECT-MPI increased, those with high pretest probability were less frequently referred. Image parameters were mostly comparable between the study groups in the PRE and PAN periods.
ABSTRACT
BACKGROUND: Single photon emission computed tomography (SPECT) myocardial perfusion images (MPI) can be displayed both in traditional short-axis (SA) cardiac planes and polar maps for interpretation and quantification. It is essential to reorient the reconstructed transaxial SPECT MPI into standard SA slices. This study is aimed to develop a deep-learning-based approach for automatic reorientation of MPI. METHODS: A total of 254 patients were enrolled, including 226 stress SPECT MPIs and 247 rest SPECT MPIs. Fivefold cross-validation with 180 stress and 201 rest MPIs was used for training and internal validation; the remaining images were used for testing. The rigid transformation parameters (translation and rotation) from manual reorientation were annotated by an experienced nuclear cardiologist and used as the reference standard. A convolutional neural network (CNN) was designed to predict the transformation parameters. Then, the derived transform was applied to the grid generator and sampler in spatial transformer network (STN) to generate the reoriented image. A loss function containing mean absolute errors for translation and mean square errors for rotation was employed. A three-stage optimization strategy was adopted for model optimization: (1) optimize the translation parameters while fixing the rotation parameters; (2) optimize rotation parameters while fixing the translation parameters; (3) optimize both translation and rotation parameters together. RESULTS: In the test set, the Spearman determination coefficients of the translation distances and rotation angles between the model prediction and the reference standard were 0.993 in X axis, 0.992 in Y axis, 0.994 in Z axis, 0.987 along X axis, 0.990 along Y axis and 0.996 along Z axis, respectively. For the 46 stress MPIs in the test set, the Spearman determination coefficients were 0.858 in percentage of profusion defect (PPD) and 0.858 in summed stress score (SSS); for the 46 rest MPIs in the test set, the Spearman determination coefficients were 0.9 in PPD and 0.9 in summed rest score (SRS). CONCLUSIONS: Our deep learning-based LV reorientation method is able to accurately generate the SA images. Technical validations and subsequent evaluations of measured clinical parameters show that it has great promise for clinical use.
Subject(s)
Deep Learning , Myocardial Perfusion Imaging , Humans , Tomography, Emission-Computed, Single-Photon/methods , Heart , Perfusion , Myocardial Perfusion Imaging/methodsABSTRACT
Coronary microvascular dysfunction is present in two-thirds of patients showing symptoms and signs of myocardial ischemia. Their microcirculation has abnormalities due to endothelial and smooth muscle cell dysfunction. Impairment of this mechanism causes a high risk of adverse cardiovascular event. Diagnosing coronary microvascular dysfunction is challenging. Guidelines recommend the use of nuclear medicine procedures in the above-mentioned indications. Myocardial perfusion imaging with positron emission tomography is a novel procedure with high diagnostic accuracy and quality of images. It has short acquisition, low effective radiation dose and prognostic factors. There are still unknowns about this procedure and all its benefits.
Subject(s)
Coronary Artery Disease , Myocardial Ischemia , Myocardial Perfusion Imaging , Coronary Angiography/methods , Coronary Artery Disease/complications , Coronary Circulation/physiology , Humans , Microcirculation/physiology , Myocardial Ischemia/diagnostic imaging , Myocardial Perfusion Imaging/adverse effects , Myocardial Perfusion Imaging/methodsABSTRACT
OBJECTIVE: The exploration and the implementation of a deep learning method using a state-of-the-art convolutional neural network for the classification of polar maps represent myocardial perfusion for the detection of coronary artery disease. SUBJECTS AND METHODS: In the proposed research, the dataset includes stress and rest polar maps in attenuation-corrected (AC) and non-corrected (NAC) format, counting specifically 144 normal and 170 pathological cases. Due to the small number of the dataset, the following methods were implemented: First, transfer learning was conducted using VGG16, which is applied broadly in medical industry. Furthermore, data augmentation was utilized, wherein the images are rotated and flipped for expanding the dataset. Secondly, we evaluated a custom convolutional neural network called RGB CNN, which utilizes fewer parameters and is more lightweight. In addition, we utilized the k-fold validation for evaluating variability and overall performance of the examined model. RESULTS: Our RGB CNN model achieved an agreement rating of 92.07% with a loss of 0.2519. The transfer learning technique (VGG16) attained 95.83% accuracy. CONCLUSIONS: The proposed model could be an effective tool for medical classification problems, in the case of polar map data acquired from myocardial perfusion images.
Subject(s)
Coronary Artery Disease , Deep Learning , Coronary Artery Disease/diagnostic imaging , Humans , Neural Networks, Computer , Tomography, Emission-Computed, Single-PhotonABSTRACT
BACKGROUND: Stress-only myocardial perfusion imaging (MPI) markedly reduces radiation dose, scanning time, and cost. We developed an automated clinical algorithm to safely cancel unnecessary rest imaging with high sensitivity for obstructive coronary artery disease (CAD). METHODS AND RESULTS: Patients without known CAD undergoing both MPI and invasive coronary angiography from REFINE SPECT were studied. A machine learning score (MLS) for prediction of obstructive CAD was generated using stress-only MPI and pre-test clinical variables. An MLS threshold with a pre-defined sensitivity of 95% was applied to the automated patient selection algorithm. Obstructive CAD was present in 1309/2079 (63%) patients. MLS had higher area under the receiver operator characteristic curve (AUC) for prediction of CAD than reader diagnosis and TPD (0.84 vs 0.70 vs 0.78, P < .01). An MLS threshold of 0.29 had superior sensitivity than reader diagnosis and TPD for obstructive CAD (95% vs 87% vs 87%, P < .01) and high-risk CAD, defined as stenosis of the left main, proximal left anterior descending, or triple-vessel CAD (sensitivity 96% vs 89% vs 90%, P < .01). CONCLUSIONS: The MLS is highly sensitive for prediction of both obstructive and high-risk CAD from stress-only MPI and can be applied to a stress-first protocol for automatic cancellation of unnecessary rest imaging.
Subject(s)
Coronary Artery Disease , Myocardial Perfusion Imaging , Algorithms , Coronary Angiography/methods , Coronary Artery Disease/diagnostic imaging , Humans , Machine Learning , Myocardial Perfusion Imaging/methods , Patient Selection , Perfusion , Tomography, Emission-Computed, Single-Photon/methodsABSTRACT
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.
Subject(s)
Deep Learning , Myocardial Perfusion Imaging , Humans , Image Processing, Computer-Assisted/methods , Myocardial Perfusion Imaging/methods , Perfusion , ROC Curve , Tomography, Emission-Computed, Single-Photon/methodsABSTRACT
The aim of this study was to employ phase analysis to diagnose left ventricular mechanical dyssynchrony (LVMD) in asymptomatic patients with diabetes mellitus type 2 and normal perfusion study which may help prevent diabetic cardiomyopathy. Ninety-three consecutive patients with known type 2 diabetes and 81 age- and gender- matched patients without diabetes who were candidates for SPECT-MPI were considered as the control group. The presence of LVMD as an possible risk factor for cardiomyopathy- was determined using phase analysis for each scan with quantitative gated SPECT (QGS) and corridor4DM (4DM) software. All outcomes such as phase bandwidth (PBW) and phase standard deviation (PSD) were compared between the two groups. A total of 174 patients were included in the study. There were no statistically significant difference regarding demographic factors between the two groups (P > 0.05). PBW showed statistically significant differences (increased in diabetics) between the control and diabetic patients (P < 0.05). Kruskal Wallis analysis revealed that as the duration of diabetes is prolonged, especially more than 15 years, the probability of LVMD is increased as well (P = 0.021). Fraction of asymptomatic diabetic patients with normal ejection fraction and gated SPECT MPI-especially those with prolonged diabetes- might have some degrees of LVMD. Phase analysis can detect this which in turn may prevent progress into heart failure.
Subject(s)
Cardiac-Gated Single-Photon Emission Computer-Assisted Tomography , Diabetes Mellitus, Type 2 , Myocardial Perfusion Imaging , Ventricular Dysfunction, Left , Diabetes Mellitus, Type 2/complications , Humans , Predictive Value of Tests , Tomography, Emission-Computed, Single-Photon , Ventricular Dysfunction, Left/diagnostic imaging , Ventricular Dysfunction, Left/etiologyABSTRACT
BACKGROUND: Epicardial fat volume (EFV) has been reported to be associated with coronary artery disease (CAD). CAD is the leading cause of myocardial ischemia and myocardial ischemia is closely related to major adverse cardiovascular events. We hypothesized that EFV could provide incremental value to traditional risk factors and coronary artery calcium score (CACS) in predicting myocardial ischemia in Chinese patients with suspected CAD. METHODS: We retrospectively studied 204 Chinese patients with suspected CAD who underwent single-photon emission computerized tomography-myocardial perfusion imaging (SPECT-MPI) combined with computed tomography (CT). Pericardial contours were manually defined, and EFV was automatically calculated. A reversible perfusion defect with summed difference score (SDS) ≥ 2 was defined as myocardial ischemia. RESULTS: The myocardial ischemia group had higher EFV than normal MPI group (137.80 ± 34.95cm3 vs. 106.63 ± 29.10 cm3, P < .001). In multivariable logistic regression analysis, high EFV was significantly associated with myocardial ischemia [odds ratio (OR): 8.30, 95% CI: 3.72-18.49, P < .001]. Addition of EFV to CACS and traditional risk factors could predict myocardial ischemia more effectively, with larger AUC .82 (P < .001), positive net reclassification index .14 (P = .04) and integrated discrimination improvement .14 (P < .001). The bootstrap resampling method (times = 500) was used to internally validation and calculate the 95% confidence interval (CI) of the AUC (95% CI .75-.87). The calibration curve for the probability of myocardial ischemia demonstrated good agreement between prediction and observation. CONCLUSIONS: In Chinese patients with suspected CAD, EFV was significantly associated with myocardial ischemia, and improved prediction of myocardial ischemia above traditional risk factors and CACS.
Subject(s)
Coronary Artery Disease , Myocardial Ischemia , Myocardial Perfusion Imaging , Calcium , Coronary Angiography/methods , Coronary Artery Disease/diagnostic imaging , Humans , Myocardial Ischemia/diagnostic imaging , Myocardial Perfusion Imaging/methods , Predictive Value of Tests , Retrospective Studies , Risk FactorsABSTRACT
Based on superior image quality, more accurate gated images, and lower radiation exposure to patients, Technetium-99m (Tc-99m) based tracers are preferred over Thallium-201 for SPECT myocardial perfusion imaging. The two Tc-99m tracers, sestamibi and tetrofosmin, have many similar characteristics but there are differences in blood and liver clearance rates, as well as the recommended time after injection for imaging to achieve optimal image quality. Because published peer-reviewed studies examining optimal times between injection and imaging are limited, it can be difficult to identify evidence-based opportunities to optimize imaging protocols. Using systematic literature review methods, this study was designed to identify and consolidate the available evidence on the use of sestamibi compared to tetrofosmin for variable injection to imaging times in regard to test efficiency, including test length and re-scan rates, and image quality, including overall quality and cardiac to extra-cardiac ratios. The composite of this data shows that earlier imaging with tetrofosmin is equivalent to later imaging with sestamibi when assessing subjective image quality or when quantifying heart-to-extra-cardiac ratios. Image quality and heart-to-extra-cardiac ratios comparing early versus later imaging with tetrofosmin were comparable if not equivalent to each other. The equivalency of the imaging quality occurs with 15 minutes (on average) earlier imaging compared to sestamibi and 30 minutes compared to standard time tetrofosmin. The subjective findings of equivalent image quality are also shown with objective measurements of heart-to-extra-cardiac ratios. In this review, the significantly shorter injection-to-acquisition times with tetrofosmin compared to sestamibi resulted in better efficiency and less waiting times for patients; in addition, significantly higher re-scan rates with sestamibi compared to tetrofosmin due to hepatic activity contributed to better throughput with tetrofosmin.
Subject(s)
Heart Diseases/diagnostic imaging , Myocardial Perfusion Imaging , Organophosphorus Compounds/administration & dosage , Organotechnetium Compounds/administration & dosage , Technetium Tc 99m Sestamibi/administration & dosage , Tomography, Emission-Computed, Single-Photon , Humans , Time FactorsABSTRACT
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.
Subject(s)
Deep Learning , Image Processing, Computer-Assisted , Myocardial Perfusion Imaging , Algorithms , Humans , Phantoms, Imaging , Signal-To-Noise Ratio , Tomography, Emission-Computed, Single-PhotonABSTRACT
PURPOSE: We assessed the effects of the COVID-19 pandemic on myocardial perfusion imaging (MPI) for ischemic heart disease during the lockdown imposed by the Italian Government. METHODS: We retrospectively reviewed the number and the findings of stress single-photon emission computed tomography (SPECT)-MPI performed between February and May 2020 during the COVID-19 pandemic at the University of Napoli Federico II. The number and the findings of stress SPECT-MPI studies acquired in the corresponding months of the years 2017, 2018, and 2019 were also evaluated for direct comparison. RESULTS: The number of stress SPECT-MPI studies performed during the COVID-19 pandemic (n = 123) was significantly lower (P < 0.0001) compared with the mean yearly number of procedures performed in the corresponding months of the years 2017, 2018, and 2019 (n = 413). Yet, the percentage of abnormal stress SPECT-MPI studies was similar (P = 0.65) during the pandemic (36%) compared with the mean percentage value of the corresponding period of the years 2017, 2018, and 2019 (34%). CONCLUSION: The number of stress SPECT-MPI studies was significantly reduced during the COVID-19 pandemic compared with the corresponding months of the previous 3 years. The lack of difference in the prevalence of abnormal SPECT-MPI studies between the two study periods strongly suggests that many patients with potentially abnormal imaging test have been missed during the pandemic.
Subject(s)
COVID-19/epidemiology , Myocardial Ischemia/diagnostic imaging , Myocardial Perfusion Imaging/statistics & numerical data , Tomography, Emission-Computed, Single-Photon/statistics & numerical data , Aged , Female , Humans , Italy , Male , Middle Aged , Quarantine/statistics & numerical dataABSTRACT
The performance of SPECT myocardial perfusion imaging (MPI) may deteriorate in smaller hearts, primarily because of the lower resolution of conventional Anger cameras. 18F-flurpiridaz is a novel PET MPI agent with superior image and defect resolution. We sought to determine the diagnostic performance of 99mTc-labeled SPECT MPI compared with 18F-flurpiridaz PET MPI according to left ventricle (LV) size. Methods: We conducted a substudy of the phase III clinical trial of flurpiridaz (n = 750) and stratified diagnostic performance according to the median PET LV end-diastolic volume (LVEDV), with smaller LVs defined as having an LVEDV of less than 113 mL (n = 369) and larger LVs defined as having an LVEDV of at least 113 mL (n = 381). Images were interpreted by the majority rule of 3 independent masked readers. The reference standard was quantitative invasive angiography, with at least 50% stenosis in at least 1 coronary artery considered significant. Results: SPECT performance decreased significantly from an area under the curve (AUC) of 0.75 in larger LVs to 0.67 in smaller LVs (P = 0.03), whereas PET performance was similar in larger and smaller LVs (AUC, 0.79 vs. 0.77, P = 0.49). Accordingly, in smaller LVs, PET had a higher AUC (0.77) than the SPECT AUC (0.67) (P < 0.0001), a phenomenon driven by female patients (P < 0.0001). In smaller LVs, there was a degradation of SPECT sensitivity that was highly significant (P < 0.001), whereas there was no significant change in PET sensitivity according to LV size (P = 0.07). Overall, PET had significantly higher sensitivity than SPECT in both smaller LVs (67% vs. 43%, P < 0.001) and larger LVs (76% vs. 61%, P < 0.001). The specificities of PET and SPECT were similar in larger LVs (76% vs. 83%, P = 0.11). Although SPECT specificity improved in smaller compared with larger LVs (90% vs. 83%, P = 0.03), the PET specificity did not change with LV size (76% vs. 76%, P = 0.9). Conclusion: The diagnostic performance of 18F-flurpiridaz PET MPI is not affected by LV size and is superior to SPECT MPI in patients with smaller LVs, highlighting the importance of appropriate test selection in these patients.
Subject(s)
Heart Ventricles/diagnostic imaging , Myocardial Perfusion Imaging , Positron-Emission Tomography , Pyridazines , Tomography, Emission-Computed, Single-Photon , Female , Heart Ventricles/pathology , Humans , Male , Middle Aged , Organ SizeABSTRACT
"A quick glance at selected topics in this issue" aims to highlight contents of the Journal and provide a quick review to the readers.
Subject(s)
Cardiac Imaging Techniques/trends , Cardiology/trends , Cardiovascular Diseases/diagnostic imaging , Image Enhancement/methods , Nuclear Medicine/trends , Tomography, Emission-Computed/trends , Cardiovascular Diseases/therapy , Humans , Positron Emission Tomography Computed TomographyABSTRACT
OBJECTIVE: To test whether phase analysis indices from SPECT-MPI for left ventricular mechanical dyssynchrony (LVMD) are predictors of major adverse cardiac events (MACEs) in long-standing diabetes mellitus (DM). METHODS: A total of 136 DM patients with normal perfusion and left ventricular systolic functions were followed up for about two years and divided into two groups according to the presence and the absence of MACEs. RESULT: Thirteen (9.5%) patients experienced MACEs during follow-up. Patients experiencing MACEs showed significantly higher phase standard deviation (PSD) and wider phase bandwidth (PBW) than those who did not. Moreover, both PSD and PBW showed significant correlations (r = 0.25 and 0.27; P < 0.05) with duration of DM. Logistic regression analysis revealed significant associations of DM duration, microvascular complications, and LVMD indices for predicting MACEs. Kaplan-Meier event-free survival analysis revealed significantly higher rate of MACEs (Logrank = 10.02; P = 0.001) in patients with high PSD and wide PBW. An overall fit model consisting of high-PSD and wide-PBW group was improved with the addition of microvascular complications (χ2 = 15.9; P = 0.03) and further by addition of DM duration of ≥ 15 years (χ2 = 24.3; P = 0.007) as variables. CONCLUSION: LVMD indices are novel prognostic markers in diabetic patients with normal perfusion and left ventricular systolic functions and their increases in magnitudes with DM-duration and in the presence of microvascular complications.
Subject(s)
Diabetes Mellitus, Type 2/complications , Diabetes Mellitus, Type 2/diagnostic imaging , Ventricular Dysfunction, Left/diagnostic imaging , Ventricular Dysfunction, Left/epidemiology , Adult , Aged , Aged, 80 and over , Cardiac-Gated Single-Photon Emission Computer-Assisted Tomography , Cohort Studies , Diabetes Mellitus, Type 2/mortality , Disease-Free Survival , Female , Humans , Male , Middle Aged , Myocardial Perfusion Imaging , Predictive Value of Tests , PrognosisABSTRACT
BACKGROUND: Stomach wall uptake (SWU) of tracer in 99mTc-MIBI myocardial perfusion imaging (MPI) occasionally leads to imaging artifacts, thereby lowering the diagnostic accuracy. It is less-studied phenomenon for possible link with proton pump inhibitors (PPIs) intake. This prospective work looked for association of SWU with PPI intake and compared its incidence with H2 antagonists (H2A) users and patients not on either gastroprotective medication. METHODS: One hundred fifty-six patients undergoing one day stress/rest 99mTc-MIBI SPECT-MPI were distributed into four groups: control group (n = 48, not on any gastroprotective medication), PPI group (n = 47, on PPI treatment), H2A group (n = 19, on H2A therapy), and intervention group (N = 42, PPI discontinued for 3 days before MPI). Poststress planar images were analyzed for clinically relevant SWU. RESULTS: Clinically relevant SWU was seen in 36% of PPI group patients compared to 8% in the control group, 10.5% in the H2A group, and 9.5% in the intervention group, respectively, with statistically significant difference. Only 1/40 patients undergoing exercise stress showed clinically relevant SWU compared to 26/116 patients undergoing adenosine stress (P = .020). CONCLUSION: Patients on PPIs scheduled for vasodilator stress MPI may discontinue PPIs for 3 days, or replace with H2A to reduce the incidence of clinically relevant SWU associated with PPI therapy.