Predicting post-operative right ventricular failure using video-based deep learning

Despite progressive improvements over the decades, the rich temporally resolved data in an echocardiogram remain underutilized. Human assessments reduce the complex patterns of cardiac wall motion, to a small list of measurements of heart function. All modern echocardiography artificial intelligence (AI) systems are similarly limited by design - automating measurements of the same reductionist metrics rather than utilizing the embedded wealth of data. This underutilization is most evident where clinical decision making is guided by subjective assessments of disease acuity. Predicting the likelihood of developing post-operative right ventricular failure (RV failure) in the setting of mechanical circulatory support is one such example. Here we describe a video AI system trained to predict post-operative RV failure using the full spatiotemporal density of information in pre-operative echocardiography. We achieve an AUC of 0.729, and show that this ML system significantly outperforms a team of human experts at the same task on independent evaluation.

Long-term survival in patients with post-LVAD right ventricular failure: multi-state modelling with competing outcomes of heart transplant

BACKGROUND

Multicenter data on long term survival following LVAD implantation that make use of contemporary definitions of RV failure are limited. Furthermore, traditional survival analyses censor patients who receive a bridge to heart transplant. Here we compare the outcomes of LVAD patients who develop post-operative RV failure accounting for the transitional probability of receiving an interim heart transplantation.

METHODS

We use a retrospective cohort of LVAD patients sourced from multiple high-volume centers based in the United States. Five- and ten-year survival accounting for transition probabilities of receiving a heart transplant were calculated using a multi-state Aalen Johansen survival model.

RESULTS

Of the 897 patients included in the study, 238 (26.5%) developed post-operative RV failure at index hospitalization. At 10 years the probability of death with post-op RV failure was 79.28% vs 61.70% in patients without (HR 2.10; 95% CI 1.72 - 2.57; p = < .001). Though not significant, patients with RV failure were less likely to be bridged to a heart transplant (HR 0.87, p = .4). Once transplanted the risk of death between both patient groups remained equivalent; the probability of death after a heart transplant was 3.97% in those with post-operative RV failure shortly after index LVAD implant, as compared to 14.71% in those without.

CONCLUSIONS AND RELEVANCE

Long-term durable mechanical circulatory support is associated with significantly higher mortality in patients who develop post-operative RV failure. Improving outcomes may necessitate expeditious bridge to heart transplant wherever appropriate, along with critical reassessment of organ allocation policies.

Patient-Specific Computational Fluid Dynamics Reveal Localized Flow Patterns Predictive of Post-Left Ventricular Assist Device Aortic Incompetence

BACKGROUND

Progressive aortic valve disease has remained a persistent cause of concern in patients with left ventricular assist devices. Aortic incompetence (AI) is a known predictor of both mortality and readmissions in this patient population and remains a challenging clinical problem.

METHODS

Ten left ventricular assist device patients with de novo aortic regurgitation and 19 control left ventricular assist device patients were identified. Three-dimensional models of patients' aortas were created from their computed tomography scans, following which large-scale patient-specific computational fluid dynamics simulations were performed with physiologically accurate boundary conditions using the SimVascular flow solver.

RESULTS

The spatial distributions of time-averaged wall shear stress and oscillatory shear index show no significant differences in the aortic root in patients with and without AI (mean difference, 0.67 dyne/cm2 [95% CI, -0.51 to 1.85]; P=0.23). Oscillatory shear index was also not significantly different between both groups of patients (mean difference, 0.03 [95% CI, -0.07 to 0.019]; P=0.22). The localized wall shear stress on the leaflet tips was significantly higher in the AI group than the non-AI group (1.62 versus 1.35 dyne/cm2; mean difference [95% CI, 0.15-0.39]; P<0.001), whereas oscillatory shear index was not significantly different between both groups (95% CI, -0.009 to 0.001; P=0.17).

CONCLUSIONS

Computational fluid dynamics serves a unique role in studying the hemodynamic features in left ventricular assist device patients where 4-dimensional magnetic resonance imaging remains unfeasible. Contrary to the widely accepted notions of highly disturbed flow, in this study, we demonstrate that the aortic root is a region of relatively stagnant flow. We further identified localized hemodynamic features in the aortic root that challenge our understanding of how AI develops in this patient population.

Computational Fluid Dynamics Simulations to Predict False Lumen Enlargement After Surgical Repair of Type-A Aortic Dissection

We aim to use computational fluid dynamics to investigate the hemodynamic conditions that may predispose to false lumen enlargement in this patient population. Nine patients who received surgical repairs of their type-A aortic dissections between 2017-2018 were retrospectively identified. Multiple contrast-enhanced post-operative CT scans were used to construct 3D models of aortic geometries. Computational fluid dynamics simulations of the models were run on a high-performance computing cluster using SimVascular - an open-source simulation package. Physiological pulsatile flow conditions (4.9 L/min) were used at the aortic true lumen inlet, and physiological vascular resistances were applied at the distal vascular ends. Exploratory analyses showed no correlation between rate of false lumen growth and blood pressure, immediate post-op aortic diameter, or the number of fenestrations (p = 0.2). 1-year post-operative CT scans showed a median false lumen growth rate of 4.31 (3.66, 14.67) mm/year Median (Interquartile range) peak systolic, mid-diastolic, and late diastolic velocity magnitudes were 0.90 (1.40); 0.10 (0.16); and 0.06 (0.06) cm/s respectively. Spearman's ranked correlations between fenestration velocity and 1-year false lumen growth rates were found to be statistically significant: Velocity magnitude at peak systolic (p = 0.025; rho = 0.75), mid diastolic (p = 0.025; rho = 0.75) and late diastolic phases of the cardiac cycle (p = 0.006; rho = 0.85). We have shown that false lumen growth is strongly correlated to fenestration flow velocity, which has potential implications for post-operative surveillance and risk stratification.

Comparison of pharmacodynamics between insulin glargine 100 U/mL and insulin glargine 300 U/mL in healthy cats

Insulin glargine (IGla) is a synthetic human-recombinant insulin analog that is used routinely in people as a q24h basal insulin. The 300 U/mL (U300) formulation of IGla is associated with longer duration of action and less within-day variability, making it a better basal insulin compared with the 100 U/mL (U100) formulation. We hypothesized that in healthy cats, IGlaU300 has a flatter time-action profile and longer duration of action compared with IGlaU100. Seven healthy neutered male, purpose-bred cats were studied in a randomized, crossover design. Pharmacodynamics of IGlaU100 and IGlaU300 (0.8 U/kg, subcutaneous) were determined by the isoglycemic clamp method. The time-action profile of IGlaU300 was flatter compared with IGlaU100 as demonstrated by lower peak (5.6 ± 1.1 mg/kg/min vs 8.3 ± 1.9 mg/kg/min, respectively; P = 0.04) with no difference in total metabolic effect (ME; P = 0.7) or duration of action (16.8 h ± 4.7 h vs 13.4 h ± 2.6 h; P = 0.2). The greater fraction of ME in the 12- to 24-h period postinjection (35 ± 23% vs 7 ± 8% respectively; P = 0.048) and lower intraday GIR% variability (7.8 ± 3.7% vs 17.4 ± 8.2% respectively; P = 0.03) supports a flatter time-action profile of IGlaU300. There were no differences in onset and end of the action. In summary, although both formulations have a similar duration of action that is well below 24 h, the ME of IGlaU300 is more evenly distributed over a 24 h period in healthy cats, making it a better candidate for once-daily injection in diabetics compared with IGlaU100.

The effect of the ghrelin-receptor agonist capromorelin on glucose metabolism in healthy cats

Somatostatin secretion from islet delta cells is important in maintaining low glycemic variability (GV) by providing negative feedback to beta cells and inhibiting insulin secretion. Capromorelin is a ghrelin-receptor agonist that activates the growth hormone secretagogue receptor on delta cells. We hypothesized that in cats, capromorelin administration will result in decreased GV at the expense of reduced insulin secretion and glucose tolerance. Seven healthy cats were treated with capromorelin from days 1-30. After the first day, fasting blood glucose increased (+13 ± 3 mg/dL, P < 0.0001), insulin decreased (+128 ± 122 ng/dL, P = 0.03), and glucagon was unchanged. Blood glucose was increased throughout an intravenous glucose tolerance test on day 1 with blunting of first-phase insulin response ([FPIR] 4,931 ± 2,597 ng/L/15 min) compared with day -3 (17,437 ± 8,302 ng/L/15 min, P = 0.004). On day 30, FPIR was still blunted (9,993 ± 4,285 ng/L/15 min, P = 0.045), but glucose tolerance returned to baseline. Mean interstitial glucose was increased (+19 ± 6 mg/dL, P = 0.03) on days 2-4 but returned to baseline by days 27-29 (P = 0.3). On days 2-4, GV was increased (SD = 9.7 ± 3.2) compared with baseline (SD = 5.0 ± 1.1, P = 0.02) and returned to baseline on days 27-29 (SD = 6.1 ± 1.1, P = 0.16). In summary, capromorelin caused a decline in insulin secretion and glycemic control and an increase in glucose variability early in the course of treatment, but these effects diminished toward the end of 30 d of treatment.

Isomerization of N-Allyl Amides To Form Geometrically Defined Di-, Tri-, and Tetrasubstituted Enamides

Enamides represent bioactive pharmacophores in various natural products, and have become increasingly common reagents for asymmetric incorporation of nitrogen functionality. Yet the synthesis of the requisite geometrically defined enamides remains problematic, especially for highly substituted and Z-enamides. Herein we wish to report a general atom economic method for the isomerization of a broad range of N-allyl amides to form Z-di-, tri-, and tetrasubstituted enamides with exceptional geometric selectivity. This report represents the first examples of a catalytic isomerization of N-allyl amides to form nonpropenyl disubstituted, tri- and tetrasubstituted enamides with excellent geometric control. Applications of these geometrically defined enamides toward the synthesis of cis vicinal amino alcohols and tetrasubstituted α-borylamido complexes are discussed.

Effect of polyethylene glycol on the supercoiling free energy of DNA

The supercoiling free energy of pUC19 DNA [2686 base pairs (bp)] was measured in various concentrations of PEG 8000 (polyethylene glycol; molecular weight 8000) by the topoisomer distribution method. The effective twist energy parameter (E(T)) that governs the supercoiling free energy declined linearly by 1.9-fold with increasing w/v % PEG from 0 to 7.5%, which lies below the threshold for intermolecular condensation. In principle, PEG could affect E(T) either via an osmotic exclusion mechanism or by altering the torsion elastic constant, bending rigidity, or self-repulsions of the DNA. Possible alterations of the DNA secondary structure and torsion elastic constant were assessed by CD spectroscopy and time-resolved fluorescence polarization anisotropy of intercalated ethidium. Up to 7.5% PEG, the secondary structure of the DNA remained largely unaltered, as evidenced by (1) the absence of any significant change in the CD spectrum, (2) an extremely small relative decrease (-0.0013) in intrinsic twist, and (3) a negligibly small change in the torsion elastic constant. The observed reduction in E(T) cannot be ascribed primarily to a decrease in torsion elastic constant, and most likely does not stem from a decrease in bending rigidity either. The decrease in medium dielectric constant due to PEG should increase the self-repulsions, and thereby increase E(T), which is opposite to the observed trend. Instead, the observed decline in E(T) is attributed to an osmotic exclusion mechanism. The change in molar volume excluded to the PEG (Delta V(ex)), when the linking difference converts from Delta l = 0 to Delta l = +/-1, was determined from the observed E(T) value and PEG osmotic pressure at each concentration. The experimental Delta V(ex) values agree well with theoretical estimates reckoned for a simple osmotic exclusion model, in which PEG is excluded by hard-core interactions from a concentric cylindrical volume around every duplex segment. The difference in volume excluded to PEG between the Delta l = 0 and the Delta l = +/-1 topoisomers is attributed entirely to the approximately 0.7 additional writhe "crossing" of two duplex strands at roughly 90 degrees, which is known to occur in the latter species. When the separation between the duplex centers at the "crossing" was adjusted so that the theoretical estimate of Delta V(ex) matched the experimental value at each PEG concentration, a value near 5.7 nm was obtained in each case. The invariance and plausible magnitude of this mean separation at the crossing provide strong support for this simple osmotic exclusion model. An alternative model, in which the PEG is excluded from the entire coil envelope of the DNA out to its radius of gyration, perhaps because it decreases the local dielectric constant, was also considered. The estimated difference in excluded volume in that case exceeds the experimental value by a factor of nearly 10(4), and could be ruled out on that basis.