Interpretation of SPECT wall motion with deep learning.

OBJECTIVES: We sought to develop a novel deep learning (DL) workflow to interpret single-photon emission computed tomography (SPECT) wall motion. BACKGROUND: Wall motion assessment with SPECT is limited by image temporal and spatial resolution. Visual interpretation of wall motion can be subjective and prone to error. Artificial intelligence (AI) may improve accuracy of wall motion assessment. METHODS: A total of 1038 patients undergoing rest electrocardiogram (ECG)-gated SPECT and echocardiography were included. Using echocardiography as truth, a DL-model (DL-model 1) was trained to predict the probability of abnormal wall motion. Of the 1038 patients, 317 were used to train a DL-model (DL-model 2) to assess regional wall motion. A 10-fold cross-validation was adopted. Diagnostic performance of DL was compared with human readers and quantitative parameters. RESULTS: The area under the receiver operating characteristic curve (AUC) and accuracy (ACC) of DL model (AUC: .82 [95% CI: .79-.85]; ACC: .88) were higher than human (AUC: .77 [95% CI: .73-.81]; ACC: .82; P < .001) and quantitative parameter (AUC: .74 [95% CI: .66-.81]; ACC: .78; P < .05). The net reclassification index (NRI) was 7.7%. The AUC and accuracy of DL model for per-segment and per-vessel territory diagnosis were also higher than human reader. The DL model generated results within 30 seconds with operable guided user interface (GUI) and therefore could provide preliminary interpretation. CONCLUSIONS: DL can be used to improve interpretation of rest SPECT wall motion as compared with current human readers and quantitative parameter diagnosis.

Myocardial flow reserve recovery in patients with Takotsubo syndrome: Insights from positron emission tomography.

BACKGROUND: Coronary microvascular dysfunction (CMD) has been implicated in the pathogenesis of Takotsubo syndrome (TTS). Positron emission tomography (PET) plays a key role in the assessment of CMD through myocardial flow reserve (MFR). However, there is limited information on the temporal progression of MFR and its relationship to coronary artery disease (CAD) in TTS patients. METHODS: This study evaluated patients with TTS who underwent cardiac catheterization and PET within one year of hospitalization. Patients were categorized into acute (≤10 days), subacute (11-30 days), and chronic (≥31 days) stages based on post-onset time of PET assessment. MFR values and prevalence of abnormal MFR (<2.0) were compared between stages. Temporal MFR changes in patients with obstructive CAD (≥70% stenosis by coronary angiography), non-obstructive CAD, and normal coronaries were compared. RESULTS: Of the 88 patients studied (mean age 70; 96% female), 52 (59%) were in the acute, 17 (19%) in the subacute, and 19 (22%) in the chronic stage. Median MFR in the acute stage was 2.0 (1.5-2.3), with 58% of patients showing abnormal MFR. A significant time-dependent improvement in MFR was observed (P = 0.002), accompanied by a decreased prevalence of abnormal MFR (P = 0.016). While patients with normal coronaries showed significant MFR improvement over time (P = 0.045), patients with obstructive or non-obstructive CAD demonstrated no improvement across three stages (P = 0.346 and 0.174, respectively). CONCLUSION: PET-derived MFR was impaired in TTS patients during the acute phase, with improvement suggesting potential recovery from CMD over time. The concurrent presence of obstructive CAD might impede this recovery process.

Myocardial blood flow quantification with SPECT.

INTRODUCTION: The addition of absolute myocardial blood flow (MBF) data improves the diagnostic and prognostic accuracy of relative perfusion imaging with nuclear medicine. Cardiac-specific gamma cameras allow measurement of MBF with SPECT. METHODS: This paper reviews the evidence supporting the use of SPECT to measure myocardial blood flow (MBF). Studies have evaluated SPECT MBF in large animal models and compared it in humans with invasive angiographic measurements and against the clinical standard of PET MBF. The repeatability of SPECT MBF has been determined in both single-site and multi-center trials. RESULTS: SPECT MBF has excellent correlation with microspheres in an animal model, with the number of stenoses and fractional flow reserve, and with PET-derived MBF. The inter-user coefficient of variability is ∼20% while the COV of test-retest MBF is ∼30%. SPECT MBF improves the sensitivity and specificity of the detection of multi-vessel disease over relative perfusion imaging and provides incremental value in predicting adverse cardiac events. CONCLUSION: SPECT MBF is a promising technique for providing clinically valuable information in the assessment of coronary artery disease.

Competency based medical education in nuclear cardiology: A tale of two axes.

BACKGROUND: Across medical specialties, including nuclear cardiology, competency based medical education (CBME) changes the emphasis of learning from a time or experiential emphasis to a proficiency focused approached. Plotted on a learning-curve graph the emphasis on learning has shifted from the duration/ volume-based x-axis to the performance-based y-axis. CURRENT STATUS: It has proven difficult to establish y-axis-based standards within nuclear cardiology to assess learning. As such there is a paucity of data to verify current experiential training targets and only recently is data emerging that seeks to find CBME targets by which proficiency (y-axis units) can be evaluated. Initial reports from such CBME-oriented studies indicate that in current nuclear cardiology practice, the number of studies required to achieve competency is dependent upon the chosen measure of competency that is assessed (summed stress score versus % LV ischemia), the case mix, and the modality being learnt (PET versus SPECT). Recent findings have also suggested that prior levels of experiential training may be an underestimation of the number of supervised studies learners need to interpret before they achieve competency. SUMMARY: Nuclear cardiology training has adopted the concept of CBME and is progressing toward a more modern approach to trainee assessment. This brief review provides the background, current requirements and insights into new developments in nuclear cardiology training.

Identifying left ventricular dysfunction using prospective electrocardiogram-triggered coronary computed tomography angiography.

PURPOSE: Coronary computed tomography angiography (CCTA) is an important non-invasive tool for the assessment of coronary artery disease and the delivery of information incremental to coronary anatomy. CCTA measured left ventricular (LV) mid-diastolic volume (LVMDV) and LV mass (LVMass) have important prognostic information but the utility of prospectively ECG-triggered CCTA to predict reduced left ventricular ejection fraction (LVEF) is unknown. The objective of this study was to determine if indexed LVMDV (LVMDVi) and the LVMDV:LVMass ratio on CCTA can identify patients with reduced LVEF. MATERIALS/METHODS: 8179 patients with prospectively ECG-triggered CCTA between November 2014 and December 2019 were reviewed. A subset derivation cohort of 4352 healthy patients was used to define normal LVMDVi and LVMDV:LVMass. Sex-specific thresholds were tested in a validation cohort of 1783 patients, excluded from the derivation cohort, with cardiac disease and known LVEF. The operating characteristics for 1 SD above the mean were tested for the identification of abnormal LVEF, LVEF≤35 â€‹% and ≤30 â€‹%. RESULTS: The derivation cohort had a mean LVMDVi of 61.0 â€‹± â€‹13.7 â€‹mL/m2 and LVMDV:LVMass of 1.11 â€‹± â€‹0.24 â€‹mL/g. LVMDVi and LVMDV:LVMass were both higher in patients with reduced LVEF than those with normal LVEF (98.8 â€‹± â€‹40.8 â€‹mL/m2 vs. 63.3 â€‹± â€‹19.7 â€‹mL/m2, p â€‹< â€‹0.001, and 1.32 â€‹± â€‹0.44 â€‹mL/g vs. 1.05 â€‹± â€‹0.28 â€‹mL/g, p â€‹< â€‹0.001). Both mean LVMDVi and LVMDV:LVMass increased with the severity of LVEF reduction. Sex-specific LVMDVi thresholds were 79 â€‹% and 80 â€‹% specific for identifying abnormal LVEF in females (LVMDVi â€‹≥ â€‹69.9 â€‹mL/m2) and males (LVMDVi â€‹≥ â€‹78.8 â€‹mL/m2), respectively. LVMDV:LVMass thresholds had high specificity (87 â€‹%) in both females (LVMDVi:LVMass â€‹≥ â€‹1.39 â€‹mL/g) and males (LVMDVi:LVMass â€‹≥ â€‹1.30 â€‹mL/g). CONCLUSION: Our study provides reference thresholds for LVMDVi and LVMDV:LVMass on prospectively ECG-triggered CCTA, which may identify patients who require further LV function assessment.

Application of Cardiovascular Computed Tomography to the Assessment of Patients With Hypertrophic Cardiomyopathy.

Hypertrophic cardiomyopathy is a common inherited cardiac condition in which regional myocardial thickening and scarring can lead to a range of symptoms including breathlessness, dizziness, chest pain, and collapse with loss of consciousness. It is vital to be able to understand the mechanisms behind these epiphenomena and to be able to distinguish, for example, between syncope because of arrhythmia versus syncope because of mechanical outflow tract obstruction. Therefore, we require a technique that can characterize anatomy, physiology, and myocardial substrate. Traditionally, this role has been the preserve of cardiac magnetic resonance (CMR) imaging. This review makes the case for cardiac computed tomography (CT) as an alternative imaging method. We review the use of functional CT to identify the components of outflow tract obstruction (and obstruction at other levels, which may be simultaneous), and as an aid to interventional and surgical planning. We demonstrate the added value of multiplanar isotropic reformats in this condition, particularly in cases where the diagnosis may be more challenging or where complications (such as early apical aneurysm) may be difficult to recognize with 2-dimensional techniques. In conclusion, our aim is to convince readers that cardiac CT is a highly valuable and versatile tool, which deserves wider usage and greater recognition in those caring for patients with hypertrophic cardiomyopathy.

Deep learning model to predict exercise stress test results: Optimizing the diagnostic test selection strategy and reduce wastage in suspected coronary artery disease patients.

BACKGROUND: Cardiac exercise stress testing (EST) offers a non-invasive way in the management of patients with suspected coronary artery disease (CAD). However, up to 30% EST results are either inconclusive or non-diagnostic, which results in significant resource wastage. Our aim was to build machine learning (ML) based models, using patients demographic (age, sex) and pre-test clinical information (reason for performing test, medications, blood pressure, heart rate, and resting electrocardiogram), capable of predicting EST results beforehand including those with inconclusive or non-diagnostic results. METHODS: A total of 30,710 patients (mean age 54.0 years, 69% male) were included in the study with 25% randomly sampled in the test set, and the remaining samples were split into a train and validation set with a ratio of 9:1. We constructed different ML models from pre-test variables and compared their discriminant power using the area under the receiver operating characteristic curve (AUC). RESULTS: A network of Oblivious Decision Trees provided the best discriminant power (AUC=0.83, sensitivity=69%, specificity=0.78%) for predicting inconclusive EST results. A total of 2010 inconclusive ESTs were correctly identified in the testing set. CONCLUSIONS: Our ML model, developed using demographic and pre-test clinical information, can accurately predict EST results and could be used to identify patients with inconclusive or non-diagnostic results beforehand. Our system could thus be used as a personalised decision support tool by clinicians for optimizing the diagnostic test selection strategy for CAD patients and to reduce healthcare expenditure by reducing nondiagnostic or inconclusive ESTs.

Prognostic Value of Coronary Computed Tomography Angiography in Coronary Artery Bypass Graft Patients Systematic Review and Meta-Analysis.

We sought to assess the prognostic value of coronary computed tomographic angiography (CCTA) in patients with coronary artery bypass graft (CABG) by meta-analysis. MEDLINE, Embase, Cochrane Central Register of Controlled Trials, and Scopus were searched for relevant original articles published up to July 2021. CCTA prognostic studies enrolling patients with CABG were screened and included if outcomes included all-cause mortality or major adverse cardiac events. Maximally adjusted hazard ratios (HRs) were extracted for CCTA-derived prognostic factors. HRs were log-transformed and pooled across studies using the DerSimonian-Laird random-effects model and statistical heterogeneity was assessed using the I2 statistic. Of 1,576 screened articles, 4 retrospective studies fulfilled all inclusion criteria. Collectively, a total of 1,809 patients with CABG underwent CCTA (mean [SD] age 67.0 [8.5] years across 3 studies, 81.5% male across 4 studies). Coronary artery disease severity and revascularization were categorized using 2 models: unprotected coronary territories and coronary artery protection score. The pooled HRs from the random-effects models using the most highly adjusted study estimate were 3.64 (95% confidence interval 2.48 to 5.34, I2 = 57.8%, p <0.001; 4 studies) and 4.85 (95% confidence interval 3.17 to 7.43, I2 = 39.9%, p <0.001; 2 studies) for unprotected coronary territories and coronary artery protection score, respectively. In conclusion, in a limited number of studies, CCTA is an independent predictor of adverse events in patients with CABG. Larger studies using uniform models and endpoints are needed.

Utility of FFRCT in Patients with Chest Pain.

PURPOSE OF REVIEW: The goal of this article is to review the data supporting the use of fractional flow reserve derived from coronary computed tomography angiography (FFRCT) in patients with chest pain. REVIEW FINDINGS: Numerous clinical trials have demonstrated that the diagnostic accuracy of coronary computed tomography angiography (CCTA) can be improved with the use of FFRCT, primarily due to its superior specificity when compared to CCTA alone. This promising development may help reduce the need for invasive angiography in patients presenting with chest pain. Furthermore, some studies have indicated that incorporating FFRCT into decision-making is safe, with an FFRCT value of ≥ 0.8 being associated with favorable outcomes. While FFRCT has been shown to be feasible in patients with acute chest pain, further large-scale studies are warranted to confirm its utility. The emergence of FFRCT as a tool for the management of patients with chest pain is promising. However, potential limitations require the interpretation of FFRCT in conjunction with clinical context.

Advances in Single-Photon Emission Computed Tomography: Hardware, Software, and Myocardial Flow Reserve.

The clinical presentation of coronary artery disease (CAD) has changed during the last 20 years with less ischemia on stress testing and more nonobstructive CAD on coronary angiography. Single-photon emission computed tomography (SPECT) myocardial perfusion imaging should include the measurement of myocardial flow reserve and assessment of coronary calcium for the diagnosis of nonobstructive CAD and coronary microvascular disease. SPECT/CT systems provide reliable attenuation correction for better specificity and low-dose CT for coronary calcium evaluation. SPECT MFR measurement is accurate, well validated, and repeatable.

Dual versus single energy cardiac CT to measure extra cellular volume in cardiac amyloidosis: Correlations with cardiac MRI.

Rationale and objectives: Determine in cardiac amyloid (CA) patients, whether cardiac CT derived extracellular volume (ECV) correlates with that obtained by MRI. Perform this correlation with single (SECT) versus dual energy (DECT) CT and evaluate whether a single sample volume ECV-measure was as reliable as a global (16 segment) assessment. Materials and methods: CA patients who had undergone a clinical cardiac MRI (CMR) were recruited prospectively. SECT and DECT cardiac scans were performed. Three ECG-triggered prospective SECT scans were acquired: non-contrast, arterial-phase contrast and 5-minute delayed images. A DECT scan was performed at 7 min. Post processing was used to determine ECV. Analyses of SECT or DECT global ECV versus CMR were performed using the Pearson correlation coefficient, Bland Altman analysis and Intraclass correlation coefficient (ICC). Similar analyses were performed to examine the performance of single-segment sampling by SECT or DECT versus CMR. Results: 25 patients were recruited, mean age was 80.0 ± 7.1 years, 80 % were male, 21 patients had transthyretin- CA, 4 had light chain- CA. Correlations were close with both SECT or DECT global ECV versus CMR (r = 0.79 and 0.88 respectively, p < 0.001 for both). Reliability of both SECT and DECT to assess global ECV in comparison to CMR was good: ICC for SECT was 0.88 (95 % CI 0.73-0.95) and 0.93 (95 % CI 0.82-0.97) for DECT. For single volume sampling techniques: correlations were close with both SECT or DECT versus CMR (r = 0.60 and 0.72 respectively, p < 0.01 for both) There was no difference in ICC for SECT (0.74, 95 %CI 0.41-0.88) versus DECT (0.84, 95 % CI 0.63-0.93). Wider confidence intervals were noted for ICC with single versus global CT derived ECV assessment. Mean effective radiation dose was for SECT was 5.49 ± 8.04 mSv and 6.90 ± 3.01 mSv for DECT dual energy CT (p = 0.75). Conclusions: Global ECV values derived by both DECT or SECT correlated with those obtained by CMR and demonstrated good reliability by ICC in a population of CA patients. DECT and SECT single sampling derived ECV values also demonstrated close correlation and good reliability but the ICCs for single sampling had wider confidence intervals than global ECV assessment.

An endogenous inhibitor of angiogenesis downregulated by hypoxia in human aortic valve stenosis promotes disease pathogenesis.

Aortic valve stenosis is the most common valve disease in the western world. Central to the pathogenesis of this disease is the growth of new blood vessels (angiogenesis) within the aortic valve allowing infiltration of immune cells and development of intra-valve inflammation. Identifying the cellular mediators involved in this angiogenesis is important as this may reveal new therapeutic targets which could ultimately prevent the progression of aortic valve stenosis. Aortic valves from patients undergoing surgery for aortic valve replacement or dilation of the aortic arch were examined both ex vivo and in vitro. We now demonstrate that the anti-angiogenic protein, soluble fms-like tyrosine kinase 1 (sFlt1), a non-signalling soluble receptor for vascular endothelial growth factor, is constitutively expressed in non-diseased valves. sFlt-1 expression was, however, significantly reduced in aortic valve tissue from patients with aortic valve stenosis while protein markers of hypoxia were simultaneously increased. Exposure of primary-cultured valve interstitial cells to hypoxia resulted in a decrease in the expression of sFlt-1. We further reveal using a bioassay that siRNA knock-down of sFlt1 in valve interstitial cells directly results in a pro-angiogenic environment. Finally, incubation of aortic valves with sphingosine 1-phosphate, a bioactive lipid-mediator, increased sFlt-1 expression and inhibited angiogenesis within valve tissue. In conclusion, this study demonstrates that sFlt1 expression is directly correlated with angiogenesis in aortic valves and the observed decrease in sFlt-1 expression in aortic valve stenosis could increase valve inflammation, promoting disease progression. This could be a viable therapeutic target in treating this disease.

Adiponectin serum level is an independent and incremental predictor of all-cause mortality after transcatheter aortic valve replacement.

BACKGROUND: Quantifiable biomarkers may be useful for a better risk and frailty assessment of patients referred for transcatheter aortic valve implantation (TAVI). HYPOTHESIS: To determine if adiponectin serum concentration predicts all-cause mortality in patients undergoing TAVI. METHODS: 77 consecutive patients, undergoing TAVI, were analyzed. The CT axial slices at the level of the fourth lumbar vertebra were used to measure the psoas muscle area, and its low-density muscle fraction (LDM (%)). To assess the operative risk, the STS (Society of Thoracic Surgeons Predicted Risk of Mortality) score, Log. Euroscore, and Euroscore II were determined. A clinical frailty assessment was performed. ELISA kits were used to measure adiponectin serum levels. We searched for a correlation between serum adiponectin concentration and all-cause mortality after TAVI. RESULTS: The mean age was 80.8 ± 7.4 years. All-cause mortality occurred in 22 patients. The mean follow-up was 1779 days (range: 1572-1825 days). Compared with patients with the lowest adiponectin level, patients in the third tertile had a hazards ratio of all-cause mortality after TAVI of 4.155 (95% CI: 1.364-12.655) (p = .004). In the multivariable model, including STS score, vascular access of TAVI procedure, LDM (%), and adiponectin serum concentration, serum adiponectin level, and LDM(%) were independent predictors of all-cause mortality after TAVI (p = .178, .303, .042, and .017, respectively). Adiponectin level was a predictor of all-cause mortality in females and males (p = .012 and 0.024, respectively). CONCLUSION: Adiponectin serum level is an independent and incremental predictor of all-cause mortality in patients undergoing TAVI.

Comparison of mechanisms of angiostasis caused by the anti-inflammatory steroid 5α-tetrahydrocorticosterone versus conventional glucocorticoids.

5α-Tetrahydrocorticosterone (5αTHB) is an effective topical anti-inflammatory agent in mouse, with less propensity to cause skin thinning and impede new blood vessel growth compared with corticosterone. Its anti-inflammatory effects were not prevented by RU38486, a glucocorticoid receptor antagonist, suggesting alternative mechanisms. The hypothesis that 5αTHB directly inhibits angiogenesis to a lesser extent than hydrocortisone was tested, focussing on glucocorticoid receptor mediated actions. New vessel growth from aortae from C57BL/6 male mice was monitored in culture, in the presence of 5αTHB, hydrocortisone (mixed glucocorticoid/mineralocorticoid receptor agonist) or the selective glucocorticoid receptor agonist dexamethasone. Transcript profiles were studied, as was the role of the glucocorticoid receptor, using the antagonist, RU38486. Ex vivo, 5αTHB suppressed vessel growth from aortic rings, but was less potent than hydrocortisone (EC50 2512 nM 5αTHB, versus 762 nM hydrocortisone). In contrast to conventional glucocorticoids, 5αTHB did not alter expression of genes related to extracellular matrix integrity or inflammatory signalling, but caused a small increase in Per1 transcript, and decreased transcript abundance of Pecam1 gene. RU38486 did not antagonise the residual effects of 5αTHB to suppress vessel growth or regulate gene expression, but modified effects of dexamethasone. 5αTHB did not alter expression of glucocorticoid-regulated genes Fkbp51 and Hsd11b1, unlike hydrocortisone and dexamethasone. In conclusion, compared with hydrocortisone, 5αTHB exhibits limited suppression of angiogenesis, at least directly in blood vessels and probably independent of the glucocorticoid receptor. Discriminating the mechanisms employed by 5αTHB may provide the basis for the development of novel safer anti-inflammatory drugs for topical use.

Anti-inflammatory effect of rosuvastatin in patients with HIV infection: An FDG-PET pilot study.

AIMS: This study aimed to evaluate markers of systemic as well as imaging markers of inflammation in the ascending aorta, bone marrow, and spleen measured by 18F-FDG PET/CT, in HIV+ patients at baseline and following therapy with rosuvastatin. METHODS AND RESULTS: Of the 35 HIV+ patients enrolled, 17 were randomized to treatment with 10 mg/day rosuvastatin and 18 to usual care for 6 months. An HIV- control cohort was selected for baseline comparison of serum inflammatory markers and monocyte markers of inflammation. 18F-FDG-PET/CT imaging of bone marrow, spleen, and thoracic aorta was performed in the HIV+ cohort at baseline and 6 months. While CD14++CD16- and CCR2 expressions were reduced, serum levels of IL-7, IL-8, and MCP-1 were elevated in the HIV+ population compared to the controls. There was a significant drop in FDG uptake in the bone marrow (TBRmax), spleen (SUVmax) and thoracic aortic (TBRmax) in the statin-treated group compared to the control group (bone marrow: - 10.3 ± 16.9% versus 5.0 ± 18.9%, p = .0262; spleen: - 9.8 ± 20.3% versus 11.3 ± 28.8%, p = .0497; thoracic aorta: - 19.1 ± 24.2% versus 4.3 ± 15.4%, p = .003). CONCLUSIONS: HIV+ patients had significantly markers of systemic inflammation including monocyte activation. Treatment with low-dose rosuvastatin in the HIV+ cohort significantly reduced bone marrow, spleen and thoracic aortic FDG uptake.