Cardiac Phase Space Tomography: A novel method of assessing coronary artery disease utilizing machine learning.

BACKGROUND: Artificial intelligence (AI) techniques are increasingly applied to cardiovascular (CV) medicine in arenas ranging from genomics to cardiac imaging analysis. Cardiac Phase Space Tomography Analysis (cPSTA), employing machine-learned linear models from an elastic net method optimized by a genetic algorithm, analyzes thoracic phase signals to identify unique mathematical and tomographic features associated with the presence of flow-limiting coronary artery disease (CAD). This novel approach does not require radiation, contrast media, exercise, or pharmacological stress. The objective of this trial was to determine the diagnostic performance of cPSTA in assessing CAD in patients presenting with chest pain who had been referred by their physician for coronary angiography. METHODS: This prospective, multicenter, non-significant risk study was designed to: 1) develop machine-learned algorithms to assess the presence of CAD (defined as one or more ≥ 70% stenosis, or fractional flow reserve ≤ 0.80) and 2) test the accuracy of these algorithms prospectively in a naïve verification cohort. This report is an analysis of phase signals acquired from 606 subjects at rest just prior to angiography. From the collective phase signal data, features were extracted and paired with the known angiographic results. A development set, consisting of signals from 512 subjects, was used for machine learning to determine an algorithm that correlated with significant CAD. Verification testing of the algorithm was performed utilizing previously untested phase signals from 94 subjects. RESULTS: The machine-learned algorithm had a sensitivity of 92% (95% CI: 74%-100%) and specificity of 62% (95% CI: 51%-74%) on blind testing in the verification cohort. The negative predictive value (NPV) was 96% (95% CI: 85%-100%). CONCLUSIONS: These initial multicenter results suggest that resting cPSTA may have comparable diagnostic utility to functional tests currently used to assess CAD without requiring cardiac stress (exercise or pharmacological) or exposure of the patient to radioactivity.

Intracardiac acoustic radiation force impulse (ARFI) and shear wave imaging in pigs with focal infarctions.

Four pigs, three with focal infarctions in the apical intraventricular septum (IVS) and/or left ventricular free wall (LVFW), were imaged with an intracardiac echocardiography (ICE) transducer. Custom beam sequences were used to excite the myocardium with focused acoustic radiation force (ARF) impulses and image the subsequent tissue response. Tissue displacement in response to the ARF excitation was calculated with a phase-based estimator, and transverse wave magnitude and velocity were each estimated at every depth. The excitation sequence was repeated rapidly, either in the same location to generate 40 Hz M-modes at a single steering angle, or with a modulated steering angle to synthesize 2-D displacement magnitude and shear wave velocity images at 17 points in the cardiac cycle. Both types of images were acquired from various views in the right and left ventricles, in and out of infarcted regions. In all animals, acoustic radiation force impulse (ARFI) and shear wave elasticity imaging (SWEI) estimates indicated diastolic relaxation and systolic contraction in noninfarcted tissues. The M-mode sequences showed high beat-to-beat spatio-temporal repeatability of the measurements for each imaging plane. In views of noninfarcted tissue in the diseased animals, no significant elastic remodeling was indicated when compared with the control. Where available, views of infarcted tissue were compared with similar views from the control animal. In views of the LVFW, the infarcted tissue presented as stiff and non-contractile compared with the control. In a view of the IVS, no significant difference was seen between infarcted and healthy tissue, whereas in another view, a heterogeneous infarction was seen to be presenting itself as non-contractile in systole.

Short-lag spatial coherence imaging of cardiac ultrasound data: initial clinical results.

Short-lag spatial coherence (SLSC) imaging is a novel beamforming technique that reduces acoustic clutter in ultrasound images. A clinical study was conducted to investigate clutter reduction and endocardial border detection in cardiac SLSC images. Individual channel echo data were acquired from the left ventricle of 14 volunteers, after informed consent and institutional review board approval. Paired B-mode and SLSC images were created from these data. Contrast, contrast-to-noise, and signal-to-noise ratios were measured in paired images, and these metrics were improved with SLSC imaging in most cases. Three cardiology fellows rated the visibility of endocardial segments in randomly ordered B-mode and SLSC cine loops. SLSC imaging offered 22%-33% improvement (p < 0.05) in endocardial border visibility when B-mode image quality was poor (i.e., 80% or more of the endocardial segments could not be visualized by the three reviewers). The percentage of volunteers with poor-quality images was decreased from 21% to 7% with the SLSC beamformer. Results suggest that SLSC imaging has the potential to improve clinical cardiac assessments that are challenged by clutter.

Intracardiac echocardiography measurement of dynamic myocardial stiffness with shear wave velocimetry.

Acoustic radiation force (ARF)-based methods have been demonstrated to be a viable tool for noninvasively estimating tissue elastic properties, and shear wave velocimetry has been used to measure quantitatively the stiffening and relaxation of myocardial tissue in open-chest experiments. Dynamic stiffness metrics may prove to be indicators for certain cardiac diseases, but a clinically viable means of remotely generating and tracking transverse wave propagation in myocardium is needed. Intracardiac echocardiography (ICE) catheter-tip transducers are demonstrated here as a viable tool for making this measurement. ICE probes achieve favorable proximity to the myocardium, enabling the use of shear wave velocimetry from within the right ventricle throughout the cardiac cycle. This article describes the techniques used to overcome the challenges of using a small probe to perform ARF-driven shear-wave velocimetry and presents in vivo porcine data showing the effectiveness of this method in the interventricular septum.

Ranolazine-related dyspnea on exertion.

BACKGROUND: Ranolazine is increasingly being prescribed for the treatment of chronic stable angina. This report describes an adverse effect that may be related to ranolazine. CASE SUMMARY: A 77-year-old white man with chronic renal insufficiency was evaluated for moderate dyspnea on exertion (DOE). Cardiac and pulmonary workup revealed nonobstructive coronary artery disease and mild obstructive lung disease. The patient had been taking ranolazine 500 mg daily for possible angina for the past 2 months. Given the temporal association of his symptoms with drug initiation, ranolazine was discontinued during the hospitalization. One month after discontinuing ranolazine, the patient's DOE had completely resolved; the only intervention had been discontinuation of ranolazine. The patient's Naranjo algorithm score was 3, indicating a possible adverse drug reaction. CONCLUSIONS: No previous cases of ranolazine-related DOE requiring drug cessation have been published. Ranolazine may be associated with DOE in this elderly man.

Regulation of the platelet-derived growth factor receptor-beta by G protein-coupled receptor kinase-5 in vascular smooth muscle cells involves the phosphatase Shp2.

Smooth muscle cell (SMC) proliferation and migration are substantially controlled by the platelet-derived growth factor receptor-beta (PDGFRbeta), which can be regulated by the Ser/Thr kinase G protein-coupled receptor kinase-2 (GRK2). In mouse aortic SMCs, however, we found that prolonged PDGFRbeta activation engendered down-regulation of GRK5, but not GRK2; moreover, GRK5 and PDGFRbeta were coordinately up-regulated in SMCs from atherosclerotic arteries. With SMCs from GRK5 knock-out and cognate wild type mice (five of each), we found that physiologic expression of GRK5 increased PDGF-promoted PDGFRbeta seryl phosphorylation by 3-fold and reduced PDGFRbeta-promoted phosphoinositide hydrolysis, thymidine incorporation, and overall PDGFRbeta tyrosyl phosphorylation by approximately 35%. Physiologic SMC GRK5 activity also increased PDGFRbeta association with the phosphatase Shp2 (8-fold), enhanced phosphorylation of PDGFRbeta Tyr(1009) (the docking site for Shp2), and reduced phosphorylation of PDGFRbeta Tyr(1021). Consistent with having increased PDGFRbeta-associated Shp2 activity, GRK5-expressing SMCs demonstrated greater PDGF-induced Src activation than GRK5-null cells. GRK5-mediated desensitization of PDGFRbeta inositol phosphate signaling was diminished by Shp2 knock-down or impairment of PDGFRbeta/Shp2 association. In contrast to GRK5, physiologic GRK2 activity did not alter PDGFRbeta/Shp2 association. Finally, purified GRK5 effected agonist-dependent seryl phosphorylation of partially purified PDGFRbetas. We conclude that GRK5 mediates the preponderance of PDGF-promoted seryl phosphorylation of the PDGFRbeta in SMCs, and, through mechanisms involving Shp2, desensitizes PDGFRbeta inositol phosphate signaling and enhances PDGFRbeta-triggered Src activation.

The platelet-derived growth factor receptor-beta phosphorylates and activates G protein-coupled receptor kinase-2. A mechanism for feedback inhibition.

G protein-coupled receptor kinase-2 (GRK2) serine-phosphorylates the platelet-derived growth factor receptor-beta (PDGFRbeta), and thereby diminishes signaling by the receptor. Because activation of GRK2 may involve phosphorylation of its N-terminal tyrosines by c-Src, we tested whether the PDGFRbeta itself could tyrosine-phosphorylate and activate GRK2. To do so, we used wild type (WT) and Y857F mutant PDGFRbetas in HEK cells, which lack endogenous PDGFRs. The Y857F PDGFRbeta autophosphorylates normally but does not phosphorylate exogenous substrates. Although PDGF-stimulated Y857F and WT PDGFRbetas activated c-Src equivalently, the WT PDGFRbeta tyrosine-phosphorylated GKR2 60-fold more than the Y857F PDGFRbeta in intact cells. With purified GRK2 and either WT or Y857F PDGFRbetas immunoprecipitated from HEK cells, GRK2 tyrosyl phosphorylation was PDGF-dependent and required the WT PDGFRbeta, even though the WT and Y857F PDGFRbetas autophosphorylated equivalently. This PDGFRbeta-mediated GRK2 tyrosyl phosphorylation enhanced GRK2 activity: GRK2-mediated seryl phosphorylation of the PDGFRbeta was 9-fold greater for the WT than for the Y857F in response to PDGF, but equivalent when GRK2 was activated by sequential stimulation of beta2-adrenergic and PDGF-beta receptors. Furthermore, both PDGFRbeta-mediated GRK2 tyrosyl phosphorylation and GRK2-mediated PDGFRbeta seryl phosphorylation were reduced approximately 50% in intact cells by mutation to phenylalanine of three tyrosines in the N-terminal domain of GRK2. We conclude that the activated PDGFRbeta itself phosphorylates GRK2 tyrosyl residues and thereby activates GRK2, which then serine-phosphorylates and desensitizes the PDGFRbeta.