Search

PURPOSE: Artificial intelligence (AI) has high diagnostic accuracy for coronary artery disease (CAD) from myocardial perfusion imaging (MPI). However, when trained using high-risk populations (such as patients with correlating invasive testing), the disease probability can be overestimated due to selection bias. We evaluated different strategies for training AI models to improve the calibration (accurate estimate of disease probability), using external testing. METHODS: Deep learning was trained using 828 patients from 3 sites, with MPI and invasive angiography within 6-months. Perfusion was assessed using upright (U-TPD) and supine total perfusion deficit (S-TPD). AI training without data augmentation (Model 1) was compared to training with augmentation (increased sampling) of patients without obstructive CAD (Model 2), and patients without CAD and TPD

BACKGROUND: Single photon emission computed tomography (SPECT) myocardial perfusion is frequently used to predict risk of major adverse cardiovascular events (MACE). We performed an external validation of the CRAX2MACE score, developed to estimate 2-year risk of MACE in patients with suspected coronary artery disease (CAD). METHODS: Patients who underwent clinically indicated SPECT with available follow-up for MACE were included (N=2,985). The prediction performance for MACE (revascularization, myocardial infarction, or death) within 2-years for CRAX2MACE was compared with stress and ischemic total perfusion deficit (TPD) using area under the receiver operating characteristic curve (AUC). Calibration was assessed with calibration plots, Brier score and the Hosmer-Lemeshow test. RESULTS: MACE occurred within 2 years in 243 (8.1%) patients. The AUC for CRAX2MACE (0.710, 95% CI 0.677 – 0.743) was significantly higher compared to stress TPD (AUC 0.669, 95% CI 0.632 – 0.706, p=0.010) and ischemic TPD (AUC 0.664, 95% CI 0.627 – 0.700, p

AimsTo assess processes of care and clinical outcomes in cancer patients with ST elevation myocardial infarction (STEMI) according to cancer type.Methods and resultsThis is a national population-based study of patients admitted with STEMI in the UK between January 2005 and March 2019. Data were obtained from the National Heart Attack Myocardial Infarction National Audit Project (MINAP) registry and the Hospital Episode Statistics registry. We identified 353 448 STEMI-indexed admissions between 2005 and 2019. Of those, 8581 (2.4%) had active cancer. Prostate cancer (29% of STEMI patients with cancer) was the most common cancer followed by haematologic malignancies (14%) and lung cancer (13%). Cancer patients were less likely to receive invasive coronary revascularization (60.0% vs. 71.6%, P ConclusionSTEMI patients with cancer have a higher risk of short- and long-term mortality, particularly lung cancer. Colon cancer is the main cancer associated with major bleeding. Cardiovascular disease was the main cause of death in the first month, whereas cancer was the main cause of death within 1 year.

BACKGROUND: The utility of cardiac stress testing depends on the prevalence of myocardial ischemia within candidate populations. However, a comprehensive assessment of the factors influencing frequency of myocardial ischemia within contemporary populations referred for stress testing has not been performed. METHODS: We assessed 19,690 patients undergoing nuclear stress testing from a multicenter registry. The chi-square test was used to assess the relative importance of features for predicting myocardial ischemia. RESULTS: In the overall cohort, LVEF, male gender, and rest total perfusion deficit (TPD) were the top three predictors of ischemia, followed by CAD status, age, typical angina, and CAD risk factors. Myocardial ischemia was observed in 13.6% of patients with LVEF>55%, in 26.2% of patients with LVEF 45-54%, and in 48.3% among patients with LVEF

BACKGROUND: Accurately predicting which patients will have abnormal perfusion on MPI based on pre-test clinical information may help physicians make test selection decisions. We developed and validated a machine learning (ML) model for predicting abnormal perfusion using pre-test features. METHODS: We included consecutive patients who underwent SPECT MPI, with 20,418 patients from a multi-center (5 sites) international registry in the training population and 9,019 patients (from 2 separate sites) in the external testing population. The ML (extreme gradient boosting) model utilized 30 pre-test features to predict the presence of abnormal myocardial perfusion by expert visual interpretation. RESULTS: In external testing, the ML model had higher prediction performance for abnormal perfusion (area under receiver-operating characteristic curve [AUC] 0.762, 95% CI 0.750 – 0.774) compared to the clinical CAD consortium (AUC 0.689) basic CAD consortium (AUC 0.657), and updated Diamond-Forrester models (AUC 0.658, p

Severe acute respiratory syndrome coronavirus (SARS-CoV2) and associated COVID-19 infection continue to impact patients globally. Patients with underlying health conditions are at heightened risk of adverse outcomes from COVID-19; however, research involving patients with rare health conditions remains scarce. The amyloidoses are a rare grouping of protein deposition diseases. Light-chain and transthyretin amyloidosis are the most common disease forms, often present with systemic involvement of vital organs including the heart, nerves, kidneys, and GI tracts of affected individuals. The Amyloidosis Program of Calgary examined 152 ATTR patients and 103 AL patients analyzing rates of vaccination, COVID-19 testing, infection outcomes, influence referrals, and excess deaths. Results showed 15 total PCR-confirmed COVID-19 infections in the tested population of amyloid patients, with a higher frequency of infections among patient with AL compared to the ATTR cohort (26.2% vs 5.1%). Four patients (26.6%) required hospital admission for COVID-19 infection, 2 ATTR, and 2 AL patients. Of the confirmed cases, 1 (0.07%) unvaccinated ATTR patient died of a COVID-19 infection. An excess of deaths was found in both the ATTR and AL cohorts when comparing pre-pandemic years 2018 and 2019 to the pandemic years of 2020 and 2021. The finding suggests that amyloidosis patients are likely at a high risk for severe COVID-19 infection and mortality, especially those of advanced age, those on an active treatment with chemotherapy, and those with concomitant B-cell or plasma cell disorder. The impact of virtual healthcare visits and pandemic measures on the excess of deaths observed requires further research.

Aims We report disease-specific cardiovascular causes of mortality among cancer patients in the USA between 1999 and 2019, considering temporal trends by age, sex, and cancer site. Methods and results We used the Multiple Cause of Death database, accessed through the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research resource. We included 629 308 decedents with cardiovascular disease (CVD) recorded as the primary cause of death and active malignancy listed as a contributing cause of death. We created disease-specific CVD categories and grouped cancers by site. We calculated the proportion of CVD deaths attributed to each disease category stratified by sex, age, and cancer site. We also examined disease-specific temporal trends by cancer site. Ischaemic heart disease (IHD) was the most common cardiovascular cause of death across all cancer types (55.6%), being more common in men (59.8%), older ages, and in those with lung (67.8%) and prostate (58.3%) cancers. Cerebrovascular disease (12.9%) and hypertensive diseases (7.6%) were other common causes of death. The proportion of deaths due to heart failure was greatest in haematological (7.7%) and breast (6.3%) cancers. There was a decreasing temporal trend in the proportion of cardiovascular deaths attributed to IHD across all cancer types. The proportion of deaths due to hypertensive diseases showed the greatest percentage increase, with the largest change in breast cancer patients (+191.1%). Conclusion We demonstrate differential cardiovascular mortality risk by cancer site and demographics, providing insight into the evolving healthcare needs of this growing high-cardiovascular risk population.

Artificial intelligence (AI) techniques have emerged as a highly efficient approach to accurately and rapidly interpret diagnostic imaging and may play a vital role in nuclear cardiology. In nuclear cardiology, there are many clinical, stress, and imaging variables potentially available, which need to be optimally integrated to predict the presence of obstructive coronary artery disease (CAD) or predict the risk of cardiovascular events. In spite of clinical awareness of a large number of potential variables, it is difficult for physicians to integrate multiple features consistently and objectively. Machine learning (ML) is particularly well suited to integrating this vast array of information to provide patient-specific predictions. Deep learning (DL), a branch of ML characterized by a multi-layered convolutional model architecture, can extract information directly from images and identify latent image features associated with a specific prediction. This review will discuss the latest AI applications to disease diagnosis and risk prediction in nuclear cardiology with a focus on potential clinical applications.

Background Transthyretin amyloidosis cardiomyopathy (ATTR-CM) patients are often older and may be at risk for obstructive epicardial coronary artery disease (oeCAD). While ATTR-CM may cause small vessel coronary disease, the prevalence and clinical significance of oeCAD is not well described. Methods and results The prevalence and incidence of oeCAD and its association with all-cause mortality and hospitalization among 133 ATTR-CM patients with ≥ 1-year follow-up was evaluated. The mean age was 78 ± 9 years, 119 (89%) were male, 116 (87%) had wild-type and 17 (13%) had hereditary subtypes. Seventy-two (54%) patients underwent oeCAD investigations, with 30 (42%) receiving a positive diagnosis. Among patients with a positive oeCAD diagnosis, 23 (77%) were diagnosed prior to ATTR-CM diagnosis, 6 (20%) at the time of ATTR-CM diagnosis, and 1 (3%) after ATTR-CM diagnosis. Baseline characteristics between patients with and without oeCAD were similar. Among patients with oeCAD, only 2 (7%) required additional investigations, intervention or hospitalization after ATTR-CM diagnosis. After a median follow-up of 27 months there were 37 (28%) deaths in the study population, including 5 patients with oeCAD (17%). Fifty-six (42%) patients in the study population required hospitalization, including 10 patients with oeCAD (33%). There was no significant difference in the rates of death or hospitalization among ATTR-CM patients with and without oeCAD, and oeCAD was not significantly associated with either outcome by univariable regression analysis. Conclusions While oeCAD is prevalent in ATTR-CM patients, this diagnosis is frequently known at time of ATTR-CM diagnosis and characteristics are similar to patients without oeCAD.

Aims Positron emission tomography (PET) myocardial perfusion imaging (MPI) is often combined with coronary artery calcium (CAC) scanning, allowing for a combined anatomic and functional assessment. We evaluated the independent prognostic value of quantitative assessment of myocardial perfusion and CAC scores in patients undergoing PET. Methods and results Consecutive patients who underwent Rb-82 PET with CAC scoring between 2010 and 2018, with follow-up for major adverse cardiovascular events (MACE), were identified. Perfusion was quantified automatically with total perfusion deficit (TPD). Our primary outcome was MACE including all-cause mortality, myocardial infarction (MI), admission for unstable angina, and late revascularization. Associations with MACE were assessed using multivariable Cox models adjusted for age, sex, medical history, and MPI findings including myocardial flow reserve. In total, 2507 patients were included with median age 70. During median follow-up of 3.9 years (interquartile range 2.1–6.1), 594 patients experienced at least one MACE. Increasing CAC and ischaemic TPD were associated with increased MACE, with the highest risk associated with CAC > 1000 [adjusted hazard ratio (HR) 1.67, 95% CI 1.24–2.26] and ischaemic TPD > 10% (adjusted HR 1.80, 95% CI 1.40–2.32). Ischaemic TPD and CAC improved overall patient classification, but ischaemic TPD improved classification of patients who experienced MACE while CAC mostly improved classification of low-risk patients. Conclusions Ischaemic TPD and CAC were independently associated with MACE. Combining extent of atherosclerosis and functional measures improves the prediction of MACE risk, with CAC 0 identifying low-risk patients and regional ischaemia identifying high-risk patients in those with CAC > 0.

A decade of unprecedented progress in artificial intelligence (AI) has demonstrated a lot of interest in medical imaging research including nuclear cardiology. AI has a potential to reduce cost, save time and improve image acquisition, interpretation, and decision-making. This review summarizes recent researches and potential applications of AI in nuclear cardiology and discusses the pitfall of AI.

Quantitation of myocardialsup99m/supTc-pyrophosphate activity may have high diagnostic accuracy, but its correlation with disease burden is unknown. We examined the relationship betweensup99m/supTc-pyrophosphate quantitation and cardiac magnetic resonance (CMR) measures in patients with suspected transthyretin cardiac amyloidosis (ATTR-CM) or light chain cardiac amyloidosis (AL-CM).Consecutive patients who underwentsup99m/supTc-pyrophosphate imaging and CMR were included. ATTR-CM and AL-CM were diagnosed using standard criteria.sup99m/supTc-pyrophosphate images were assessed with standard parameters and quantified with cardiac pyrophosphate activity (CPA) and volume of involvement (VOI). We assessed the association betweensup99m/supTc-pyrophosphate image interpretation and CMR tissue characteristics.Seventy patients were identified, mean age 70.4 ± 11.4 years, with ATTR-CM and AL-CM diagnosed in 22 (31%) and 11 (16%) patients, respectively. In patients with ATTR-CM, there were significant correlations between CPA (rsup2/sup= 0.509, Plt; 0.001) and VOI (rsup2/sup= 0.586, Plt; 0.001) with native myocardial T1 mapping values. Additionally, CPA (adjusted hazard ratio (aHR) 1.04, P = 0.016), VOI (aHR 1.12, P = 0.034), and average myocardial T1 (aHR 1.12, P = 0.025) were associated with incidence of heart failure hospitalization or death.CPA and VOI were correlated with CMR measures of myocardial fibrosis in patients with ATTR-CM.sup99m/supTc-pyrophosphate quantitation may have a role in ATTR-CM disease staging, guiding treatment, or following response to therapy.

The frequency of inducible myocardial ischemia has declined in contemporary stress test cohorts, suggesting a need to re-evaluate its optimal use. To-date, however, a comprehensive analysis of the most potent predictors of myocardial ischemia among cardiac stress test patients has not been conducted.We assessed 27,615 patients referred for stress-rest SPECT myocardial perfusion imaging between January 1, 2004 and December 31, 2017. Chi-square analysis was used to ascertain the most potent predictors of ischemia.Among our cohort, CAD status (presence/absence of known CAD), rest left ventricular ejection fraction (LVEF), and typical angina were the most potent predictors of ischemia. The frequency of ischemia was only 6.6% among patients with an LVEF55% but 38.1% for patients with LVEF45% (P0.001). The frequency of myocardial ischemia was fourfold higher among patients with known CAD vs no known CAD (28.0% vs 6.5%, P0.001) and approximately threefold higher among patients with typical angina vs patients with atypical symptoms (P0.001).The frequency of myocardial ischemia varies markedly according to the common clinical parameters and is particularly high among patients with known CAD, low LVEF, and typical angina. These observations may be used to develop more cost-effective strategies for referring patients for cardiac stress testing.

The aim of this meta-analysis was to assess the diagnostic performance of various CMR imaging parameters for evaluating acute cardiac transplant rejection.Endomyocardial biopsy is the current gold standard for detection of acute cardiac transplant rejection. Cardiac magnetic resonance (CMR) is uniquely capable of myocardial tissue characterization and may be useful as a noninvasive alternative for the diagnosis of graft rejection.PubMed and Web of Science were searched for relevant publications reporting on the use of CMR myocardial tissue characterization for detection of acute cardiac transplant rejection with endomyocardial biopsy as the reference standard. Pooled sensitivity, specificity, and hierarchical modeling-based summary receiver-operating characteristic curves were calculated.Of 478 papers, 10 studies comprising 564 patients were included. The sensitivity and specificity for the detection of acute cardiac transplant rejection were 84.6 (95% CI: 65.6-94.0) and 70.1 (95% CI: 54.2-82.2) for T1, 86.5 (95% CI: 72.1-94.1) and 85.9 (95% CI: 65.2-94.6) for T2, 91.3 (95% CI: 63.9-98.4) and 67.6 (95% CI: 56.1-77.4) for extracellular volume fraction (ECV), and 50.1 (95% CI: 31.2-68.9) and 60.2 (95% CI: 36.7-79.7) for late gadolinium enhancement (LGE). The areas under the hierarchical modeling-based summary receiver-operating characteristic curve were 0.84 (95% CI: 0.81-0.87) for T1, 0.92 (95% CI: 0.89-94) for T2, 0.78 (95% CI: 0.74-0.81) for ECV, and 0.56 (95% CI: 0.51-0.60) for LGE. T2 values demonstrated the highest diagnostic accuracy, followed by native T1, ECV, and LGE (all P values 0.001 for T1, ECV, and LGE vs T2).T2 mapping demonstrated higher diagnostic accuracy than other CMR techniques. Native T1 and ECV provide high diagnostic use but lower diagnostic accuracy compared with T2, which was related primarily to lower specificity. LGE showed poor diagnostic performance for detection of rejection.

We hypothesized early post-stress left ventricular ejection fraction reserve (EFR) on solid-state-SPECT is associated with major cardiac adverse events (MACE).151 patients (70 ± 12 years, male 50%) undergoing same-day rest/regadenoson stressDuring the follow-up (mean 3.2 years), 28 MACE occurred (19%). In Kaplan-Meier analysis, there was a significantly decreased event-free survival in patients with early EFR 0% (P = 0.004). Similarly, there was a decreased event-free survival in patients with ∆ESV ≥ 5 ml at early stress (P = 0.003). However, EFR, ∆EDV, and ∆ESV at late stress were not associated with MACE-free survival. Cox proportional hazards model adjusting for clinical information and stress TPD demonstrated that EFR, ∆EDV, and ∆ESV at early stress were significantly associated with MACE (P 0.05 for all).Reduced early post-stress EFR on vasodilator stress solid-state SPECT is associated with MACE.

Quantification of myocardial perfusion and myocardial blood flow using

The photon sensitivity and spatial resolution of single-photon emission-computed tomography (SPECT) has been significantly improved by solid-state camera systems using cadmium zinc telluride (CZT) detectors. While the diagnostic accuracy of these systems is well established, there is little evidence directly comparing the prognostic utility to conventional NaI cameras.Retrospective analysis of patients undergoing SPECT between 2008 and 2012. Visual SPECT assessment was performed utilizing the 17-segment model to determine summed stress scores (SSS). We identified 12,830 consecutive patients, mean age 63.2 ± 13.7 and 56.1% male, 5072 of whom underwent CZT and 7758 NaI imaging. During a median follow-up duration of 7.0 years (IQR 5.5-8.2), a total of 2788 (21.7%) patients died. Compared to SSS 0, minimal perfusion abnormality (SSS 1-3) was associated with increased all-cause mortality with CZT camera (adjusted HR 1.32, P = .017) and NaI camera (adjusted HR 1.29, P = .001, interaction P = .803). Increasing stress abnormality was associated with a similar increase in risk with CZT or NaI imaging (interaction P .500). In a propensity matched analysis, patients with normal perfusion stress perfusion assessed with a CZT was associated with decreased mortality compared to normal perfusion assessed by a NaI camera system (hazard ratio .88, 95% CI .78-.99, P = .040).Increasing stress perfusion abnormality was associated with similar increase in all-cause mortality with CZT or NaI cameras. CZT and NaI camera systems provide similar risk stratification, however, normal myocardial perfusion may be associated with a more benign prognosis when assessed with a CZT camera system.

BACKGROUND: We compared quantification of MBF and myocardial flow reserve (MFR) with a 99mTc-sestamibi CZT-SPECT to (15)O-water PET. METHODS: SPECT-MBF for thirty patients in the WATERDAY study were re-analyzed by QPET software with motion correction and optimal placement of the arterial input function. (15)O–water PET MBF was re-quantified using dedicated software. Inter-operator variability was assessed using repeatability coefficients (RPC). RESULTS: Significant correlations were observed between global (r=0.91, p < 0.001) and regional MBF (r=0.86, p < 0.001) with SPECT compared to PET. Global MBF (rest 0.95 vs. 1.05 ml/min/g, p=0.07; stress 2.62 vs. 2.68 ml/min/g, p=0.17) and MFR (2.65 vs. 2.75, p=0.86) were similar between SPECT and PET. Rest (0.81 vs 0.98 ml/min/g, p=0.03) and stress MBF (1.98 vs. 2.61 ml/min/g, p=0.01) in right coronary artery (RCA) were lower with SPECT compared to PET. However, MFR in the RCA territory was similar (2.54 vs. 2.77, p=0.21). The SPECT-PET RPC for global MBFs and MFR were 0.95 ml/min/g and 0.94, with inter-observer RPC of 0.59 ml/min/g and 0.74 respectively. CONCLUSIONS: MBF and MFR derived from CZT-SPECT with motion correction and optimal placement of the arterial input function showed good agreement with (15)O-water PET, as well as low inter-operator variability.