ABSTRACT
BACKGROUND: We investigate whether deep learning (DL) neural networks can reduce erroneous human "judgment calls" on bedside echocardiograms and help distinguish Takotsubo syndrome (TTS) from anterior wall ST segment elevation myocardial infarction (STEMI). METHODS: We developed a single-channel (DCNN[2D SCI]), a multi-channel (DCNN[2D MCI]), and a 3-dimensional (DCNN[2D+t]) deep convolution neural network, and a recurrent neural network (RNN) based on 17,280 still-frame images and 540 videos from 2-dimensional echocardiograms in 10 years (1 January 2008 to 1 January 2018) retrospective cohort in University of Iowa (UI) and eight other medical centers. Echocardiograms from 450 UI patients were randomly divided into training and testing sets for internal training, testing, and model construction. Echocardiograms of 90 patients from the other medical centers were used for external validation to evaluate the model generalizability. A total of 49 board-certified human readers performed human-side classification on the same echocardiography dataset to compare the diagnostic performance and help data visualization. FINDINGS: The DCNN (2D SCI), DCNN (2D MCI), DCNN(2D+t), and RNN models established based on UI dataset for TTS versus STEMI prediction showed mean diagnostic accuracy 73%, 75%, 80%, and 75% respectively, and mean diagnostic accuracy of 74%, 74%, 77%, and 73%, respectively, on the external validation. DCNN(2D+t) (area under the curve [AUC] 0·787â¯vs. 0·699, Pâ¯=â¯0·015) and RNN models (AUC 0·774â¯vs. 0·699, Pâ¯=â¯0·033) outperformed human readers in differentiating TTS and STEMI by reducing human erroneous judgement calls on TTS. INTERPRETATION: Spatio-temporal hybrid DL neural networks reduce erroneous human "judgement calls" in distinguishing TTS from anterior wall STEMI based on bedside echocardiographic videos. FUNDING: University of Iowa Obermann Center for Advanced Studies Interdisciplinary Research Grant, and Institute for Clinical and Translational Science Grant. National Institutes of Health Award (1R01EB025018-01).
ABSTRACT
BACKGROUND: Myocardial infarction can be a trigger of Takotsubo syndrome. We recently characterized imaging features of acute myocardial infarction-induced Takotsubo syndrome ("Takotsubo effect"). In this study, we investigate diagnostic and prognostic implications of Takotsubo effect in patients with anterior wall ST-segment elevation myocardial infarction. METHODS: We enrolled 111 consecutive patients who developed anterior wall ST-segment elevation myocardial infarction and received percutaneous coronary intervention, and studied systolic/diastolic function, hemodynamic consequences, adverse cardiac events, as well as 30-day and five-year outcomes in patients with and without Takotsubo effect. RESULTS: Patients with Takotsubo effect showed significantly worse average peak systolic longitudinal strain (-9.5 ± 2.6% vs -11.1 ± 3.6%, p = 0.038), left ventricular ejection fraction (38.5 ± 6.8% vs 47.7 ± 8.7%, p = 0.000) and myocardial performance index (0.54 ± 0.17 vs 0.37 ± 0.15, p = 0.000) within 48 h of myocardial infarction. There was no significant difference between the two groups in diastolic ventricular filling pressures, hemodynamic consequences, and 30-day rehospitalization and mortality (Gehan-Breslow-Wilcoxon test: p = 0.157). However, patients with Takotsubo effect developed more major adverse cardiac events (log-rank test: p = 0.019) when tested at the five-year follow-up. Cox regression analysis revealed that age, hypotension, tricuspid annular plane systolic excursion, and Takotsubo effect were independent prediction factors for five-year major adverse cardiac events. The Doppler/tissue Doppler parameter E/e' correlated with MACE only in patients without Takotsubo effect. CONCLUSION: Takotsubo effect secondary to anterior ST-segment elevation myocardial infarction predicts a worse long-term prognosis.
