Spatio-temporal hybrid neural networks reduce erroneous human “judgement calls” in the diagnosis of Takotsubo syndrome

Fahim Zaman; Rakesh Ponnapureddy; Yi Grace Wang; Amanda Chang; Linda M Cadaret; Ahmed Abdelhamid; Shubha D Roy; Majesh Makan; Ruihai Zhou; Manju B Jayanna; Eric Gnall; Xuming Dai; Avneet Singh; Jingsheng Zheng; Venkata S Boppana; Feng Wang; Pahul Singh; Xiaodong Wu; Kan Liu

doi:10.1016/j.eclinm.2021.101115

Back

Spatio-temporal hybrid neural networks reduce erroneous human “judgement calls” in the diagnosis of Takotsubo syndrome

Journal article

Open access

Peer reviewed

Spatio-temporal hybrid neural networks reduce erroneous human “judgement calls” in the diagnosis of Takotsubo syndrome

Fahim Zaman, Rakesh Ponnapureddy, Yi Grace Wang, Amanda Chang, Linda M Cadaret, Ahmed Abdelhamid, Shubha D Roy, Majesh Makan, Ruihai Zhou, Manju B Jayanna, …

EClinicalMedicine, Vol.40, pp.101115-101115

10/2021

DOI: 10.1016/j.eclinm.2021.101115

PMCID: PMC8426197

PMID: 34522872

Files and links (1)

url

https://doi.org/10.1016/j.eclinm.2021.101115View

Published (Version of record) Open Access

Abstract

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). 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. 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] 0·787 vs. 0·699, P = 0·015) and RNN models (AUC 0·774 vs. 0·699, P = 0·033) outperformed human readers in differentiating TTS and STEMI by reducing human erroneous judgement calls on TTS. Spatio-temporal hybrid DL neural networks reduce erroneous human “judgement calls” in distinguishing TTS from anterior wall STEMI based on bedside echocardiographic videos. 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).

Anterior wall ST segment elevation myocardial infarction

Bedside echocardiographic videos

Deep convolution neural network

Judgment calls

Recurrent neural network

Spatiotemporal information

Takotsubo syndrome

Details

Title: Subtitle: Spatio-temporal hybrid neural networks reduce erroneous human “judgement calls” in the diagnosis of Takotsubo syndrome
Creators: Fahim Zaman - University of Iowa
Rakesh Ponnapureddy - University of Iowa
Yi Grace Wang - California State University, Dominguez Hills
Amanda Chang - University of Iowa
Linda M Cadaret - University of Iowa
Ahmed Abdelhamid - University of Iowa
Shubha D Roy - University of Iowa
Majesh Makan - Washington University in St. Louis
Ruihai Zhou - University of North Carolina at Chapel Hill
Manju B Jayanna - Lankenau Medical Center
Eric Gnall - Lankenau Medical Center
Xuming Dai - Cornell University
Avneet Singh - State University of New York System
Jingsheng Zheng - Pomona College
Venkata S Boppana - University of Kansas
Feng Wang - Washington State University
Pahul Singh - Department of Cardiology, Northwest Health Medical Center, Bentonville, AR, United States.
Xiaodong Wu - University of Iowa
Kan Liu - University of Iowa
Resource Type: Journal article
Publication Details: EClinicalMedicine, Vol.40, pp.101115-101115
DOI: 10.1016/j.eclinm.2021.101115
PMID: 34522872
PMCID: PMC8426197
NLM abbreviation: EClinicalMedicine
ISSN: 2589-5370
eISSN: 2589-5370
Publisher: Elsevier Ltd
Grant note: DOI: 10.13039/100000002, name: National Institutes of Health, award: 1R01EB025018-01; DOI: 10.13039/100008893, name: University of Iowa; DOI: 10.13039/100016596, name: Penn State Clinical and Translational Science Institute
Language: English
Date published: 10/2021
Academic Unit: Radiology; Electrical and Computer Engineering; Iowa Technology Institute; Cardiovascular Medicine; Radiation Oncology; Internal Medicine
Record Identifier: 9984198014802771

Metrics

21 Record Views

9 Times Cited - Web of Science