Search

Supplemental Digital Content is available in the text., Background: Artificial intelligence (AI) for echocardiography requires training and validation to standards expected of humans. We developed an online platform and established the Unity Collaborative to build a dataset of expertise from 17 hospitals for training, validation, and standardization of such techniques. Methods: The training dataset consisted of 2056 individual frames drawn at random from 1265 parasternal long-axis video-loops of patients undergoing clinical echocardiography in 2015 to 2016. Nine experts labeled these images using our online platform. From this, we trained a convolutional neural network to identify keypoints. Subsequently, 13 experts labeled a validation dataset of the end-systolic and end-diastolic frame from 100 new video-loops, twice each. The 26-opinion consensus was used as the reference standard. The primary outcome was precision SD, the SD of the differences between AI measurement and expert consensus. Results: In the validation dataset, the AI’s precision SD for left ventricular internal dimension was 3.5 mm. For context, precision SD of individual expert measurements against the expert consensus was 4.4 mm. Intraclass correlation coefficient between AI and expert consensus was 0.926 (95% CI, 0.904–0.944), compared with 0.817 (0.778–0.954) between individual experts and expert consensus. For interventricular septum thickness, precision SD was 1.8 mm for AI (intraclass correlation coefficient, 0.809; 0.729–0.967), versus 2.0 mm for individuals (intraclass correlation coefficient, 0.641; 0.568–0.716). For posterior wall thickness, precision SD was 1.4 mm for AI (intraclass correlation coefficient, 0.535 [95% CI, 0.379–0.661]), versus 2.2 mm for individuals (0.366 [0.288–0.462]). We present all images and annotations. This highlights challenging cases, including poor image quality and tapered ventricles. Conclusions: Experts at multiple institutions successfully cooperated to build a collaborative AI. This performed as well as individual experts. Future echocardiographic AI research should use a consensus of experts as a reference. Our collaborative welcomes new partners who share our commitment to publish all methods, code, annotations, and results openly.

Background In stable coronary artery disease, medications are used for 2 purposes: cardiovascular risk reduction and symptom improvement. In clinical trials and clinical practice, medication use is often not optimal. The ORBITA (Objective Randomised Blinded Investigation With Optimal Medical Therapy of Angioplasty in Stable Angina) trial was the first placebo‐controlled trial of percutaneous coronary intervention. A key component of the ORBITA trial design was the inclusion of a medical optimization phase, aimed at ensuring that all patients were treated with guideline‐directed truly optimal medical therapy. In this study, we report the medical therapy that was achieved. Methods and Results After enrollment into the ORBITA trial, all 200 patients entered a 6‐week period of intensive medical therapy optimization, with initiation and uptitration of risk reduction and antianginal therapy. At the prerandomization stage, the median number of antianginals established was 3 (interquartile range, 2–4). A total of 195 patients (97.5%) reached the prespecified target of ≥2 antianginals; 136 (68.0%) did not stop any antianginals because of adverse effects, and the median number of antianginals stopped for adverse effects per patient was 0 (interquartile range, 0–1). Amlodipine and bisoprolol were well tolerated (stopped for adverse effects in 4/175 [2.3%] and 9/167 [5.4%], respectively). Ranolazine and ivabradine were also well tolerated (stopped for adverse effects in 1/20 [5.0%] and 1/18 [5.6%], respectively). Isosorbide mononitrate and nicorandil were stopped for adverse effects in 36 of 172 (20.9%) and 32 of 141 (22.7%) of patients, respectively. Statins were well tolerated and taken by 191 of 200 (95.5%) patients. Conclusions In the 12‐week ORBITA trial period, medical therapy was successfully optimized and well tolerated, with few drug adverse effects leading to therapy cessation. Truly optimal medical therapy can be achieved in clinical trials, and translating this into longer‐term clinical practice should be a focus of future study. Registration URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02062593.

Background: Dobutamine stress echocardiography is widely used to test for ischemia in patients with stable coronary artery disease. In this analysis, we studied the ability of the prerandomization stress echocardiography score to predict the placebo-controlled efficacy of percutaneous coronary intervention (PCI) within the ORBITA trial (Objective Randomised Blinded Investigation With Optimal Medical Therapy of Angioplasty in Stable Angina). Methods: One hundred eighty-three patients underwent dobutamine stress echocardiography before randomization. The stress echocardiography score is broadly the number of segments abnormal at peak stress, with akinetic segments counting double and dyskinetic segments counting triple. The ability of prerandomization stress echocardiography to predict the placebo-controlled effect of PCI on response variables was tested by using regression modeling. Results: At prerandomization, the stress echocardiography score was 1.56±1.77 in the PCI arm (n=98) and 1.61±1.73 in the placebo arm (n=85). There was a detectable interaction between prerandomization stress echocardiography score and the effect of PCI on angina frequency score with a larger placebo-controlled effect in patients with the highest stress echocardiography score ( P interaction =0.031). With our sample size, we were unable to detect an interaction between stress echocardiography score and any other patient-reported response variables: freedom from angina ( P interaction =0.116), physical limitation ( P interaction =0.461), quality of life ( P interaction =0.689), EuroQOL 5 quality-of-life score ( P interaction =0.789), or between stress echocardiography score and physician-assessed Canadian Cardiovascular Society angina class ( P interaction =0.693), and treadmill exercise time ( P interaction =0.426). Conclusions: The degree of ischemia assessed by dobutamine stress echocardiography predicts the placebo-controlled efficacy of PCI on patient-reported angina frequency. The greater the downstream stress echocardiography abnormality caused by a stenosis, the greater the reduction in symptoms from PCI. Clinical Trial Registration: URL: https://www.clinicaltrials.gov . Unique identifier: NCT02062593.

Background: There are no data on how fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR) are associated with the placebo-controlled efficacy of percutaneous coronary intervention (PCI) in stable single-vessel coronary artery disease. Methods: We report the association between prerandomization invasive physiology within ORBITA (Objective Randomised Blinded Investigation With Optimal Medical Therapy of Angioplasty in Stable Angina), a placebo-controlled trial of patients who have stable angina with angiographically severe single-vessel coronary disease clinically eligible for PCI. Patients underwent prerandomization research FFR and iFR assessment. The operator was blinded to these values. Assessment of response variables, treadmill exercise time, stress echocardiography score, symptom frequency, and angina severity were performed at prerandomization and blinded follow-up. Effects were calculated by analysis of covariance. The ability of FFR and iFR to predict placebo-controlled changes in response variables was tested by using regression modeling. Results: Invasive physiology data were available in 196 patients (103 PCI and 93 placebo). At prerandomization, the majority had Canadian Cardiovascular Society class II or III symptoms (150/196, 76.5%). Mean FFR and iFR were 0.69±0.16 and 0.76±0.22, respectively; 97% had ≥1 positive ischemia tests. The estimated effect of PCI on between-arm prerandomization-adjusted total exercise time was 20.7 s (95% confidence interval [CI], –4.0 to 45.5; P =0.100) with no interaction of FFR ( P interaction =0.318) or iFR ( P interaction =0.523). PCI improved stress echocardiography score more than placebo (1.07 segment units; 95% CI, 0.70–1.44; P P interaction P interaction P =0.072) with no detectable evidence of interaction with FFR ( P interaction =0.849) or iFR ( P interaction =0.783). However, PCI resulted in more patient-reported freedom from angina than placebo (49.5% versus 31.5%; odds ratio, 2.47; 95% CI, 1.30–4.72; P =0.006) but neither FFR ( P interaction =0.693) nor iFR ( P interaction =0.761) modified this effect. Conclusions: In patients with stable angina and severe single-vessel disease, the blinded effect of PCI was more clearly seen by stress echocardiography score and freedom from angina than change in treadmill exercise time. Moreover, the lower the FFR or iFR, the greater the magnitude of stress echocardiographic improvement caused by PCI. Clinical Trial Registration: URL: https://www.clinicaltrials.gov . Unique identifier: NCT02062593.