The Effectiveness of a Deep Learning Model to Detect Left Ventricular Systolic Dysfunction from Electrocardiograms
Deep learning models can be applied to electrocardiograms (ECGs) to detect left ventricular (LV) dysfunction. We hypothesized that applying a deep learning model may improve the diagnostic accuracy of cardiologists in predicting LV dysfunction from ECGs. We acquired 37,103 paired ECG and echocardiography data records of patients who underwent echocardiography between January 2015 and December 2019. We trained a convolutional neural network to identify the data records of patients with LV dysfunction (ejection fraction < 40%) using a dataset of 23,801 ECGs. When tested on an independent set of 7,196 ECGs, we found the area under the receiver operating characteristic curve was 0.945 (95% confidence interval: 0.936-0.954). When 7 cardiologists interpreted 50 randomly selected ECGs from the test dataset of 7,196 ECGs, their accuracy for predicting LV dysfunction was 78.0% ± 6.0%. By referring to the model's output, the cardiologist accuracy improved to 88.0% ± 3.7%, which indicates that model support significantly improved the cardiologist diagnostic accuracy (P = 0.02). A sensitivity map demonstrated that the model focused on the QRS complex when detecting LV dysfunction on ECGs. We developed a deep learning model that can detect LV dysfunction on ECGs with high accuracy. Furthermore, we demonstrated that support from a deep learning model can help cardiologists to identify LV dysfunction on ECGs.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2021 |
|---|---|
| Erschienen: |
2021 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:62 |
|---|---|
| Enthalten in: |
International heart journal - 62(2021), 6 vom: 30., Seite 1332-1341 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Katsushika, Susumu [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
|---|
| Anmerkungen: |
Date Completed 13.12.2021 Date Revised 14.12.2021 published: Print Citation Status MEDLINE |
|---|
| doi: |
10.1536/ihj.21-407 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM33390091X |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM33390091X | ||
| 003 | DE-627 | ||
| 005 | 20231225222609.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231225s2021 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1536/ihj.21-407 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1112.xml |
| 035 | |a (DE-627)NLM33390091X | ||
| 035 | |a (NLM)34853226 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Katsushika, Susumu |e verfasserin |4 aut | |
| 245 | 1 | 4 | |a The Effectiveness of a Deep Learning Model to Detect Left Ventricular Systolic Dysfunction from Electrocardiograms |
| 264 | 1 | |c 2021 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 13.12.2021 | ||
| 500 | |a Date Revised 14.12.2021 | ||
| 500 | |a published: Print | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a Deep learning models can be applied to electrocardiograms (ECGs) to detect left ventricular (LV) dysfunction. We hypothesized that applying a deep learning model may improve the diagnostic accuracy of cardiologists in predicting LV dysfunction from ECGs. We acquired 37,103 paired ECG and echocardiography data records of patients who underwent echocardiography between January 2015 and December 2019. We trained a convolutional neural network to identify the data records of patients with LV dysfunction (ejection fraction < 40%) using a dataset of 23,801 ECGs. When tested on an independent set of 7,196 ECGs, we found the area under the receiver operating characteristic curve was 0.945 (95% confidence interval: 0.936-0.954). When 7 cardiologists interpreted 50 randomly selected ECGs from the test dataset of 7,196 ECGs, their accuracy for predicting LV dysfunction was 78.0% ± 6.0%. By referring to the model's output, the cardiologist accuracy improved to 88.0% ± 3.7%, which indicates that model support significantly improved the cardiologist diagnostic accuracy (P = 0.02). A sensitivity map demonstrated that the model focused on the QRS complex when detecting LV dysfunction on ECGs. We developed a deep learning model that can detect LV dysfunction on ECGs with high accuracy. Furthermore, we demonstrated that support from a deep learning model can help cardiologists to identify LV dysfunction on ECGs | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a Artificial intelligence | |
| 650 | 4 | |a Echocardiography | |
| 700 | 1 | |a Kodera, Satoshi |e verfasserin |4 aut | |
| 700 | 1 | |a Nakamoto, Mitsuhiko |e verfasserin |4 aut | |
| 700 | 1 | |a Ninomiya, Kota |e verfasserin |4 aut | |
| 700 | 1 | |a Inoue, Shunsuke |e verfasserin |4 aut | |
| 700 | 1 | |a Sawano, Shinnosuke |e verfasserin |4 aut | |
| 700 | 1 | |a Kakuda, Nobutaka |e verfasserin |4 aut | |
| 700 | 1 | |a Takiguchi, Hiroshi |e verfasserin |4 aut | |
| 700 | 1 | |a Shinohara, Hiroki |e verfasserin |4 aut | |
| 700 | 1 | |a Matsuoka, Ryo |e verfasserin |4 aut | |
| 700 | 1 | |a Ieki, Hirotaka |e verfasserin |4 aut | |
| 700 | 1 | |a Higashikuni, Yasutomi |e verfasserin |4 aut | |
| 700 | 1 | |a Nakanishi, Koki |e verfasserin |4 aut | |
| 700 | 1 | |a Nakao, Tomoko |e verfasserin |4 aut | |
| 700 | 1 | |a Seki, Tomohisa |e verfasserin |4 aut | |
| 700 | 1 | |a Takeda, Norifumi |e verfasserin |4 aut | |
| 700 | 1 | |a Fujiu, Katsuhito |e verfasserin |4 aut | |
| 700 | 1 | |a Daimon, Masao |e verfasserin |4 aut | |
| 700 | 1 | |a Akazawa, Hiroshi |e verfasserin |4 aut | |
| 700 | 1 | |a Morita, Hiroyuki |e verfasserin |4 aut | |
| 700 | 1 | |a Komuro, Issei |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t International heart journal |d 2005 |g 62(2021), 6 vom: 30., Seite 1332-1341 |w (DE-627)NLM155069675 |x 1349-3299 |7 nnns |
| 773 | 1 | 8 | |g volume:62 |g year:2021 |g number:6 |g day:30 |g pages:1332-1341 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1536/ihj.21-407 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 62 |j 2021 |e 6 |b 30 |h 1332-1341 | ||