AGMN: Association Graph-based Graph Matching Network for Coronary Artery Semantic Labeling on Invasive Coronary Angiograms.

Semantic labeling of coronary arterial segments in invasive coronary angiography (ICA) is important for automated assessment and report generation of coronary artery stenosis in computer-aided coronary artery disease (CAD) diagnosis. However, separating and identifying individual coronary arterial segments is challenging because morphological similarities of different branches on the coronary arterial tree and human-to-human variabilities exist. Inspired by the training procedure of interventional cardiologists for interpreting the structure of coronary arteries, we propose an association graph-based graph matching network (AGMN) for coronary arterial semantic labeling. We first extract the vascular tree from invasive coronary angiography (ICA) and convert it into multiple individual graphs. Then, an association graph is constructed from two individual graphs where each vertex represents the relationship between two arterial segments. Thus, we convert the arterial segment labeling task into a vertex classification task; ultimately, the semantic artery labeling becomes equivalent to identifying the artery-to-artery correspondence on graphs. More specifically, the AGMN extracts the vertex features by the embedding module using the association graph, aggregates the features from adjacent vertices and edges by graph convolution network, and decodes the features to generate the semantic mappings between arteries. By learning the mapping of arterial branches between two individual graphs, the unlabeled arterial segments are classified by the labeled segments to achieve semantic labeling. A dataset containing 263 ICAs was employed to train and validate the proposed model, and a five-fold cross-validation scheme was performed. Our AGMN model achieved an average accuracy of 0.8264, an average precision of 0.8276, an average recall of 0.8264, and an average F1-score of 0.8262, which significantly outperformed existing coronary artery semantic labeling methods. In conclusion, we have developed and validated a new algorithm with high accuracy, interpretability, and robustness for coronary artery semantic labeling on ICAs.

First-in-human experience with occlusion of the superior vena cava to reduce cardiac filling pressures in congestive heart failure.

BACKGROUND: Acutely decompensated heart failure remains a major clinical problem. Volume overload promotes cardiac and renal dysfunction and is associated with increased morbidity and mortality in heart failure. We hypothesized that transient occlusion of the superior vena cava (SVC) will reduce cardiac filling pressures without reducing cardiac output or systemic blood pressure. The objective of this proof of concept study was to provide initial evidence of safety and feasibility of transient SVC occlusion in patients with acutely decompensated heart failure and reduced ejection fraction. METHODS AND RESULTS: In eight patients with systolic heart failure, SVC occlusion was performed using a commercially available occlusion balloon. Five minutes of SVC occlusion reduced biventricular filling pressures without decreasing systemic blood pressure or total cardiac output. In three of the eight patients, a second 10-minutes occlusion had similar hemodynamic effects. SVC occlusion was well-tolerated without development of new symptoms, new neurologic deficits, or any adverse events including stroke, heart attack, or reported SVC injury or thrombosis at 7 days of follow up. CONCLUSION: We report the first clinical experience with transient SVC occlusion as a potentially new therapeutic approach to rapidly reduce cardiac filling pressures in heart failure. No prohibitive safety signal was identified and further testing to establish the clinical utility of transient SVC occlusion for acute decompensated heart failure is justified.

Increased Plasma-Free Hemoglobin Levels Identify Hemolysis in Patients With Cardiogenic Shock and a Trans valvular Micro-Axial Flow Pump.

Hemolysis is a potential limitation of percutaneously delivered left-sided mechanical circulatory support pumps, including trans valvular micro-axial flow pumps (TVP). Hemolytic biomarkers among durable left ventricular assist devices include lactate dehydrogenase (LDH) >2.5 times the upper limit of normal (ULN) and plasma-free hemoglobin (pf-Hb) >20 mg/dL. We examined the predictive value of these markers among patients with cardiogenic shock (CS) receiving a TVP. We retrospectively studied records of 116 consecutive patients receiving an Impella TVP at our institution between 2012 and 2017 for CS. Twenty-three met inclusion/exclusion criteria, and had sufficient pf-Hb data for analysis. Area under receiver-operator characteristic (ROC) curve for diagnosing hemolysis were calculated. Mean age was 62 ± 14 years and ejection fraction was 15 ± 5%. Mean duration of support was 5.4 ± 3.5 days. Pre-device LDH levels were >2.5x ULN in 71% (n = 5/7) of 5.0 and 29% of CP patients, while pre-device pf-Hb levels were >20 mg/dL in 14% (n = 1/7) of 5.0 and 25% (n = 4/16) of CP patients. Given elevated baseline LDH and pf-Hb levels, we defined hemolysis as a pf-Hb level >40 mg/dL within 72 h post-implant plus clinical evidence of device-related hemolysis. We identified that 30% (n = 7/23) had device-related hemolysis. Using ROC curve-derived cut-points, an increase in delta pf-Hb by >27mg/dL, not delta LDH, within 24 h after TVP implant (delta pf-Hb: C-statistic = 0.79, sensitivity: 57%, specificity: 93%, p <0.05) was highly predictive of hemolysis. In conclusion, we identified a change in pf-Hb, not LDH, levels is highly sensitive and specific for hemolysis in patients treated with a TVP for CS.

Mechanical Circulatory Support Devices for Acute Right Ventricular Failure.

Right ventricular (RV) failure remains a major cause of global morbidity and mortality for patients with advanced heart failure, pulmonary hypertension, or acute myocardial infarction and after major cardiac surgery. Over the past 2 decades, percutaneously delivered acute mechanical circulatory support pumps specifically designed to support RV failure have been introduced into clinical practice. RV acute mechanical circulatory support now represents an important step in the management of RV failure and provides an opportunity to rapidly stabilize patients with cardiogenic shock involving the RV. As experience with RV devices grows, their role as mechanical therapies for RV failure will depend less on the technical ability to place the device and more on improved algorithms for identifying RV failure, patient monitoring, and weaning protocols for both isolated RV failure and biventricular failure. In this review, we discuss the pathophysiology of acute RV failure and both the mechanism of action and clinical data exploring the utility of existing RV acute mechanical circulatory support devices.

The Impella Microaxial Flow Catheter Is Safe and Effective for Treatment of Myocarditis Complicated by Cardiogenic Shock: An Analysis From the Global cVAD Registry.

BACKGROUND: Myocarditis complicated by cardiogenic shock remains a complex problem. The use of acute mechanical circulatory support devices for cardiogenic shock is growing. We explored the utility of Impella transvalvular microaxial flow catheters in the setting of myocarditis with cardiogenic shock. METHODS AND RESULTS: We retrospectively analyzed data from 21 sites within the cVAD registry, an ongoing multicenter voluntary registry at sites in North America and Europe that have used Impella in patients with myocarditis. Myocarditis was defined by endomyocardial biopsy (n = 11) or by clinical history without angiographic evidence of coronary disease (n = 23). A total of 34 patients received an Impella 2.5, CP, 5.0, or RP device for cardiogenic shock complicating myocarditis. Baseline characteristics included age 42 ± 17 years, left ventricular ejection fraction (LVEF) 18% ± 10%, cardiac index 1.82 ± 0.46 L·min-1·m-2, pulmonary capillary wedge pressure 25 ± 7 mm Hg, and lactate 27 ± 31 mg/dL. Before Impella placement, 32% (n = 11) of patients required intra-aortic balloon pump. Mean duration of Impella support was 91 ± 74 hours; 21 of 34 patients (62%) survived the index hospitalization and were discharged with an improved mean LVEF of 37.32% ± 20.31% (P = .001); 15 patients recovered with successful support, 5 patients were transferred to another hospital on initial Impella support, 1 patient underwent orthotopic heart transplantation. Ten patients required transition to another mechanical circulatory support device. CONCLUSIONS: This is the largest analysis of Impella-supported myocarditis cases to date. The use of Impella appears to be safe and effective in the settings of myocarditis complicated by cardiogenic shock.

Right Ventricular Dysfunction in Acute Myocardial Infarction Complicated by Cardiogenic Shock: A Hemodynamic Analysis of the Should We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock (SHOCK) Trial and Registry.

BACKGROUND: The prevalence and significance of right ventricular dysfunction (RVD) in patients with cardiogenic shock due to acute myocardial infarction (AMI-CS) have not been well characterized. We hypothesized that RVD is common in AMI-CS and associated with worse clinical outcomes. METHODS AND RESULTS: We retrospectively analyzed patients with available hemodynamics enrolled in the Should We Emergently Revascularize Occluded Coronaries for Cardiogenic Shock (SHOCK) trial (n = 139) and registry (n = 258) to identify RVD in AMI-CS. RVD was defined by an elevated central venous pressure (CVP), elevated CVP-pulmonary capillary wedge pressure (PCWP) ratio, decreased pulmonary artery pulsatility index, and decreased right ventricular stroke work index. A P value of <.01 was used to infer significance. In the SHOCK trial and registry, respectively, 38% and 37% of patients had RVD, but RVD was not associated with 30-day or 6-month survival (hazard ratio [HR] 1.51, (99% CI 0.92-2.49; P = .10). RV failure with the use of inclusion criteria from the Recover Right Trial for RV Failure (RR-RVF) requiring percutaneous mechanical circulatory support included elevated CVP and CVP/PCWP and a low cardiac index despite ≥1 inotrope or vasopressor. In the SHOCK trial and registry, respectively, 45% (n = 63/139) and 38% (n = 98/258) of patients met RR-RVF criteria. The RR-RVF criteria were not significantly associated with 30-day mortality in the registry cohort (HR 1.44, 99% CI 1.01-2.04; P = .04), or in the trial cohort (HR 1.51, 99% CI 0.92-2.49; P = .10). CONCLUSIONS: Hemodynamically defined RVD is common in AMI-CS. Routine assessment with pulmonary artery catherization allows detection of RVD; however, further work is needed to identify interventions that will result in improved outcomes for these patients.

Acute Hemodynamic Effects of Intra-aortic Balloon Counterpulsation Pumps in Advanced Heart Failure.

BACKGROUND: The utility of intra-aortic balloon counterpulsation pumps (IABPs) in low cardiac output states is unknown and no studies have explored the impact of IABP therapy on ventricular workload in patients with advanced heart failure (HF). For these reasons, we explored the acute hemodynamic effects of IABP therapy in patients with advanced HF. METHODS: We prospectively studied 10 consecutive patients with stage D HF referred for IABP placement before left ventricular assist device (LVAD) surgery and compared with 5 control patients with preserved left ventricular (LV) ejection fraction (EF) who did not receive IABP therapy. Hemodynamics were recorded using LV conductance and pulmonary artery catheters. Cardiac index (CI)-responder and CI-nonresponder status was assigned a priori as being "equal to or above" or below the median of the IABP effect on CI, respectively, within 24 hours after IABP activation. RESULTS: Compared with controls, patients with advanced HF had lower LVEF, lower LV end-systolic pressure, lower LV stroke work, and higher LV end-diastolic pressures and volumes before IABP activation. IABP activation reduced LV stroke work primarily by reducing end-systolic pressure. IABP therapy increased CI by a median of 20% as well as increased diastolic pressure time index and the myocardial oxygen supply:demand ratio. Compared with CI-nonresponders, CI-responders had higher systemic vascular resistance, lower right heart filling pressures, and a trend toward lower left heart filling pressures with improved indices of right heart function. Compared with CI-nonresponders, the diastolic pressure time index was increased among CI-responders. CONCLUSIONS: IABP therapy may be effective at reducing LV stroke work, increasing CI, and favorably altering the myocardial oxygen supply:demand ratio in patients with advanced HF, especially among patients with low right heart filling pressures and high systemic vascular resistance.

Quantity Over Quality: How the Rise in Quality Measures is Not Producing Quality Results.

Over the past decade, quality measures (QMs) have been implemented nationally in order to establish standards aimed at improving the quality of care. With the expansion of their role in the Affordable Care Act and pay-for-performance, QMs have had an increasingly significant impact on clinical practice. However, adverse patient outcomes have resulted from adherence to some previously promulgated performance measures. Several of these QMs with unintended consequences, including the initiation of perioperative beta-blockers in noncardiac surgery and intensive insulin therapy for critically ill patients, were instituted as QMs years before large randomized trials ultimately refuted their use. The future of quality care should emphasize the importance of evidence-based, peer-reviewed measures.

Hemodynamic support with percutaneous devices in patients with heart failure.

The use of surgically implanted durable mechanical circulatory support (MCS) in high-risk patients with heart failure is declining and short-term, nondurable MCS device use is growing. Percutaneously delivered MCS options for advanced heart failure include the intra-aortic balloon pump, Impella axial flow catheter, TandemHeart centrifugal pump, and venoarterial extracorporeal membrane oxygenation. Nondurable MCS devices have unique implantation characteristics and hemodynamic effects. Algorithms and guidelines for optimal nondurable MCS device selection do not exist. Emerging technologies and applications will address the need for improved left ventricular unloading using lower-profile devices, longer-term ambulatory support, and the potential for myocardial recovery.

HAGMN-UQ: Hyper association graph matching network with uncertainty quantification for coronary artery semantic labeling.

Coronary artery disease (CAD) is one of the leading causes of death worldwide. Accurate extraction of individual arterial branches from invasive coronary angiograms (ICA) is critical for CAD diagnosis and detection of stenosis. Generating semantic segmentation for coronary arteries through deep learning-based models presents challenges due to the morphological similarity among different types of coronary arteries, making it difficult to maintain high accuracy while keeping low computational complexity. To address this challenge, we propose an innovative approach using the hyper association graph-matching neural network with uncertainty quantification (HAGMN-UQ) for coronary artery semantic labeling on ICAs. The graph-matching procedure maps the arterial branches between two individual graphs, so that the unlabeled arterial segments are classified by the labeled segments, and the coronary artery semantic labeling is achieved. Leveraging hypergraphs not only extends representation capabilities beyond pairwise relationships, but also improves the robustness and accuracy of the graph matching by enabling the modeling of higher-order associations. In addition, employing the uncertainty quantification to determine the trustworthiness of graph matching reduces the required number of comparisons, so as to accelerate the inference speed. Consequently, our model achieved an accuracy of 0.9211 for coronary artery semantic labeling with a fast inference speed, leading to an effective and efficient prediction in real-time clinical decision-making scenarios.

Bridging auditory perception and natural language processing with semantically informed deep neural networks.

Sound recognition is effortless for humans but poses a significant challenge for artificial hearing systems. Deep neural networks (DNNs), especially convolutional neural networks (CNNs), have recently surpassed traditional machine learning in sound classification. However, current DNNs map sounds to labels using binary categorical variables, neglecting the semantic relations between labels. Cognitive neuroscience research suggests that human listeners exploit such semantic information besides acoustic cues. Hence, our hypothesis is that incorporating semantic information improves DNN's sound recognition performance, emulating human behaviour. In our approach, sound recognition is framed as a regression problem, with CNNs trained to map spectrograms to continuous semantic representations from NLP models (Word2Vec, BERT, and CLAP text encoder). Two DNN types were trained: semDNN with continuous embeddings and catDNN with categorical labels, both with a dataset extracted from a collection of 388,211 sounds enriched with semantic descriptions. Evaluations across four external datasets, confirmed the superiority of semantic labeling from semDNN compared to catDNN, preserving higher-level relations. Importantly, an analysis of human similarity ratings for natural sounds, showed that semDNN approximated human listener behaviour better than catDNN, other DNNs, and NLP models. Our work contributes to understanding the role of semantics in sound recognition, bridging the gap between artificial systems and human auditory perception.

Mechanical Preload Reduction: Harnessing a Cornerstone of Heart Failure Management to Improve Clinical Outcomes.

Decongestion is a cornerstone therapeutic goal for those presenting with decompensated heart failure. Current approaches to clinical decongestion include reducing cardiac preload, which is typically limited to diuretics and hemofiltration. Several new technologies designed to mechanically reduce cardiac preload are in development. In this review, we discuss the pathophysiology of decompensated heart failure; the central role of targeting cardiac preload; emerging mechanical preload reduction technologies; and potential application of these devices.

Circulating Proteome Analysis Identifies Reduced Inflammation After Initiation of Hemodynamic Support with Either Veno-Arterial Extracorporeal Membrane Oxygenation or Impella in Patients with Cardiogenic Shock.

In-hospital mortality associated with cardiogenic shock (CS) remains high despite the use of percutaneous assist devices. We sought to determine whether support with VA-ECMO or Impella in patients with CS alters specific components of the plasma proteome. Plasma samples were collected before device implantation and 72 h after initiation of support in 11 CS patients receiving ECMO or Impella. SOMAscan was used to detect 1305 circulating proteins. Sixty-seven proteins were changed after ECMO (18 upregulated and 49 downregulated, p < 0.05), 38 after Impella (10 upregulated and 28 downregulated, p < 0.05), and only eight proteins were commonly affected. Despite minimal protein overlap, both devices were associated with markers of reduced inflammation and increased apoptosis of inflammatory cells. In summary, ECMO and Impella are associated with reduced expression of inflammatory markers and increased markers of inflammatory cell death. These circulating proteins may serve as novel targets of therapy or biomarkers to tailor AMCS use.

Factors associated with acute limb ischemia in cardiogenic shock and downstream clinical outcomes: Insights from the Cardiogenic Shock Working Group.

BACKGROUND: There are limited data depicting the prevalence and ramifications of acute limb ischemia (ALI) among cardiogenic shock (CS) patients. METHODS: We employed data from the Cardiogenic Shock Working Group (CSWG), a consortium including 33 sites. We constructed a multi-variable logistic regression to examine the association between clinical factors and ALI, we generated another logistic regression model to ascertain the association of ALI with mortality. RESULTS: There were 7,070 patients with CS and 399 (5.6%) developed ALI. Patients with ALI were more likely to be female (40.4% vs 29.4%) and have peripheral arterial disease (13.8% vs 8.3%). Stratified by maximum society for cardiovascular angiography & intervention (SCAI) shock stage, the rates of ALI were stage B 0.0%, stage C 1.8%, stage D 4.1%, and stage E 10.3%. Factors associated with higher risk for ALI included: peripheral vascular disease OR 2.24 (95% CI: 1.53-3.23; p < 0.01) and ≥2 mechanical circulatory support (MCS) devices OR 1.66 (95% CI: 1.24-2.21, p < 0.01). ALI was highest for venous-arterial extracorporeal membrane oxygenation (VA-ECMO) patients (11.6%) or VA-ECMO+ intra-aortic balloon pump (IABP)/Impella CP (16.6%) yet use of distal perfusion catheters was less than 50%. Mortality was 38.0% for CS patients without ALI but 57.4% for CS patients with ALI. ALI was significantly associated with mortality, adjusted OR 1.40 (95% CI 1.01-1.95, p < 0.01). CONCLUSIONS: The rate of ALI was 6% among CS patients. Factors most associated with ALI include peripheral vascular disease and multiple MCS devices. The downstream ramifications of ALI were dire with a considerably higher risk of mortality.

Intermediate acoustic-to-semantic representations link behavioral and neural responses to natural sounds.

Recognizing sounds implicates the cerebral transformation of input waveforms into semantic representations. Although past research identified the superior temporal gyrus (STG) as a crucial cortical region, the computational fingerprint of these cerebral transformations remains poorly characterized. Here, we exploit a model comparison framework and contrasted the ability of acoustic, semantic (continuous and categorical) and sound-to-event deep neural network representation models to predict perceived sound dissimilarity and 7 T human auditory cortex functional magnetic resonance imaging responses. We confirm that spectrotemporal modulations predict early auditory cortex (Heschl's gyrus) responses, and that auditory dimensions (for example, loudness, periodicity) predict STG responses and perceived dissimilarity. Sound-to-event deep neural networks predict Heschl's gyrus responses similar to acoustic models but, notably, they outperform all competing models at predicting both STG responses and perceived dissimilarity. Our findings indicate that STG entails intermediate acoustic-to-semantic sound representations that neither acoustic nor semantic models can account for. These representations are compositional in nature and relevant to behavior.

Loss of Bone Morphogenetic Protein-9 Reduces Survival and Increases MMP Activity After Myocardial Infarction.

No studies have explored a functional role for bone morphogenetic protein (BMP)-9, a transforming growth factor-ß superfamily ligand, in cardiac remodeling after myocardial infarction (MI). Using BMP-9 null mice, we observed that loss of BMP-9 decreases survival and increases cardiac rupture after MI. We further observed that loss of BMP-9 not only increases collagen abundance, but also promotes matrix metalloproteinase-9 activity and collagen degradation after MI. These findings identify BMP-9 as a necessary component of cardiac remodeling after MI and a potentially important target of therapy to improve outcomes after MI.

Elevated soluble fms-like tyrosine kinase-1 levels in acute coronary occlusion.

OBJECTIVE: Early recognition of an acute coronary occlusion (ACO) improves clinical outcomes. Soluble fms-like tyrosine kinase-1 (sFLT1) is an endothelium-derived protein induced by hypoxia. We tested whether sFLT1 levels are elevated in ACO. METHODS AND RESULTS: Serum sFLT1 levels were measured by enzyme-linked immunosorbent assay in patients with ST-segment elevations and angiographically confirmed ACO, unstable angina/non ST-segment elevation myocardial infarction, and 2 control groups. To further explore sFLT1 release, a mouse model of ACO and in vitro human coronary artery endothelial cell injury were used. sFLT1 levels were increased in ACO compared with unstable angina/non-ST-elevation myocardial infarction, catheterized controls, or healthy volunteers (200.7±15.5 versus 70.7±44.0 versus 10.2±4.0 versus 11.7±1.7 pg/mL respectively, P<0.001 versus ACO). At presentation, all ACO patients had elevated sFLT1 levels (>15 pg/mL, 99th percentile in controls), whereas 57% had levels of the MB isoform of creatine kinase levels >10 ng/mL (P<0.01) and 85% had ultrasensitive troponin I levels >0.05 ng/mL (P<0.05). Within 60 minutes after symptom onset, sFLT1 was more sensitive than the MB isoform of creatine kinase or ultrasensitive troponin I for ACO (100% versus 20% versus 20% respectively; P≤0.01 for each). Within 60 minutes of ACO in mice, sFLT1 levels were elevated. Hypoxia and thrombin increased sFLT1 levels within 15 minutes in human coronary artery endothelial cells. CONCLUSIONS: sFLT1 levels may be an early indicator of endothelial hypoxia in ACO.

Myocardial Injury Promotes Matrix Metalloproteinase-9 Activity in the Renal Cortex in Preclinical Models of Acute Myocardial Infarction.

New mechanistic insight into how the kidney responds to cardiac injury during acute myocardial infarction (AMI) is required. We hypothesized that AMI promotes inflammation and matrix metalloproteinase-9 (MMP9) activity in the kidney and studied the effect of initiating an Impella CP or veno-arterial extracorporeal membrane oxygenation (VA-ECMO) before coronary reperfusion during AMI. Adult male swine were subjected to coronary occlusion and either reperfusion (ischemia-reperfusion; IR) or support with either Impella or VA-ECMO before reperfusion. IR and ECMO increased while Impella reduced levels of MMP-9 in the myocardial infarct zone, circulation, and renal cortex. Compared to IR, Impella reduced myocardial infarct size and urinary KIM-1 levels, but VA-ECMO did not. IR and VA-ECMO increased pro-fibrogenic signaling via transforming growth factor-beta and endoglin in the renal cortex, but Impella did not. These findings identify that AMI increases inflammatory activity in the kidney, which may be attenuated by Impella support.