Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context
Summary: Background: Echocardiography (echo) based machine learning (ML) models may be useful in identifying patients at high-risk of all-cause mortality. Methods: We developed ML models (ResNet deep learning using echo videos and CatBoost gradient boosting using echo measurements) to predict 1-yea...
| Main Authors: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-04-01
|
| Series: | EBioMedicine |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352396423000440 |
| _version_ | 1811162778713456640 |
|---|---|
| author | Akshay Valsaraj Sunil Vasu Kalmady Vaibhav Sharma Matthew Frost Weijie Sun Nariman Sepehrvand Marcus Ong Cyril Equibec Jason R.B. Dyck Todd Anderson Harald Becher Sarah Weeks Jasper Tromp Chung-Lieh Hung Justin A. Ezekowitz Padma Kaul |
| author_facet | Akshay Valsaraj Sunil Vasu Kalmady Vaibhav Sharma Matthew Frost Weijie Sun Nariman Sepehrvand Marcus Ong Cyril Equibec Jason R.B. Dyck Todd Anderson Harald Becher Sarah Weeks Jasper Tromp Chung-Lieh Hung Justin A. Ezekowitz Padma Kaul |
| author_sort | Akshay Valsaraj |
| collection | DOAJ |
| description | Summary: Background: Echocardiography (echo) based machine learning (ML) models may be useful in identifying patients at high-risk of all-cause mortality. Methods: We developed ML models (ResNet deep learning using echo videos and CatBoost gradient boosting using echo measurements) to predict 1-year, 3-year, and 5-year mortality. Models were trained on the Mackay dataset, Taiwan (6083 echos, 3626 patients) and validated in the Alberta HEART dataset, Canada (997 echos, 595 patients). We examined the performance of the models overall, and in subgroups (healthy controls, at risk of heart failure (HF), HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF)). We compared the models' performance to the MAGGIC risk score, and examined the correlation between the models’ predicted probability of death and baseline quality of life as measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ). Findings: Mortality rates at 1-, 3- and 5-years were 14.9%, 28.6%, and 42.5% in the Mackay cohort, and 3.0%, 10.3%, and 18.7%, in the Alberta HEART cohort. The ResNet and CatBoost models achieved area under the receiver-operating curve (AUROC) between 85% and 92% in internal validation. In external validation, the AUROCs for the ResNet (82%, 82%, and 78%) were significantly better than CatBoost (78%, 73%, and 75%), for 1-, 3- and 5-year mortality prediction respectively, with better or comparable performance to the MAGGIC score. ResNet models predicted higher probability of death in the HFpEF and HFrEF (30%–50%) subgroups than in controls and at risk patients (5%–20%). The predicted probabilities of death correlated with KCCQ scores (all p < 0.05). Interpretation: Echo-based ML models to predict mortality had good internal and external validity, were generalizable, correlated with patients’ quality of life, and are comparable to an established HF risk score. These models can be leveraged for automated risk stratification at point-of-care. Funding: Funding for Alberta HEART was provided by an Alberta Innovates - Health Solutions Interdisciplinary Team Grant no. AHFMR ITG 200801018. P.K. holds a Canadian Institutes of Health Research (CIHR) Sex and Gender Science Chair and a Heart & Stroke Foundation Chair in Cardiovascular Research. A.V. and V.S. received funding from the Mitacs Globalink Research Internship. |
| first_indexed | 2024-04-10T06:34:38Z |
| format | Article |
| id | doaj.art-f1bcb4de0ad0442394b6df9a89b3bc3f |
| institution | Directory Open Access Journal |
| issn | 2352-3964 |
| language | English |
| last_indexed | 2024-04-10T06:34:38Z |
| publishDate | 2023-04-01 |
| publisher | Elsevier |
| record_format | Article |
| series | EBioMedicine |
| spelling | doaj.art-f1bcb4de0ad0442394b6df9a89b3bc3f2023-03-01T04:32:03ZengElsevierEBioMedicine2352-39642023-04-0190104479Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in contextAkshay Valsaraj0Sunil Vasu Kalmady1Vaibhav Sharma2Matthew Frost3Weijie Sun4Nariman Sepehrvand5Marcus Ong6Cyril Equibec7Jason R.B. Dyck8Todd Anderson9Harald Becher10Sarah Weeks11Jasper Tromp12Chung-Lieh Hung13Justin A. Ezekowitz14Padma Kaul15Bits Pilani KK Birla Goa Campus, Goa, IndiaCanadian VIGOUR Centre, University of Alberta, Alberta, Canada; Department of Computing Science, University of Alberta, Edmonton, Alberta, Canada; Faculty of Medicine & Dentistry, University of Alberta, Alberta, CanadaAligarh Muslim University, Uttar Pradesh, IndiaUS2.ai, SingaporeCanadian VIGOUR Centre, University of Alberta, Alberta, Canada; Department of Computing Science, University of Alberta, Edmonton, Alberta, CanadaCanadian VIGOUR Centre, University of Alberta, Alberta, Canada; Faculty of Medicine & Dentistry, University of Alberta, Alberta, CanadaUS2.ai, SingaporeUS2.ai, SingaporeFaculty of Medicine & Dentistry, University of Alberta, Alberta, CanadaLibin Cardiovascular Institute, Cumming School of Medicine, University of Calgary, Alberta, CanadaFaculty of Medicine & Dentistry, University of Alberta, Alberta, CanadaLibin Cardiovascular Institute, Cumming School of Medicine, University of Calgary, Alberta, CanadaSaw Swee Hock School of Public Health, National University of Singapore & National University Health System, Singapore; Duke-NUS Medical School, SingaporeMacKay Memorial Hospital, Taipei City, TaiwanCanadian VIGOUR Centre, University of Alberta, Alberta, Canada; Faculty of Medicine & Dentistry, University of Alberta, Alberta, CanadaCanadian VIGOUR Centre, University of Alberta, Alberta, Canada; Faculty of Medicine & Dentistry, University of Alberta, Alberta, Canada; Corresponding author. University of Alberta, Centre for Pharmacy and Health Research, 4-120 Katz Group, Edmonton, T6G2E1, Alberta, Canada.Summary: Background: Echocardiography (echo) based machine learning (ML) models may be useful in identifying patients at high-risk of all-cause mortality. Methods: We developed ML models (ResNet deep learning using echo videos and CatBoost gradient boosting using echo measurements) to predict 1-year, 3-year, and 5-year mortality. Models were trained on the Mackay dataset, Taiwan (6083 echos, 3626 patients) and validated in the Alberta HEART dataset, Canada (997 echos, 595 patients). We examined the performance of the models overall, and in subgroups (healthy controls, at risk of heart failure (HF), HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF)). We compared the models' performance to the MAGGIC risk score, and examined the correlation between the models’ predicted probability of death and baseline quality of life as measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ). Findings: Mortality rates at 1-, 3- and 5-years were 14.9%, 28.6%, and 42.5% in the Mackay cohort, and 3.0%, 10.3%, and 18.7%, in the Alberta HEART cohort. The ResNet and CatBoost models achieved area under the receiver-operating curve (AUROC) between 85% and 92% in internal validation. In external validation, the AUROCs for the ResNet (82%, 82%, and 78%) were significantly better than CatBoost (78%, 73%, and 75%), for 1-, 3- and 5-year mortality prediction respectively, with better or comparable performance to the MAGGIC score. ResNet models predicted higher probability of death in the HFpEF and HFrEF (30%–50%) subgroups than in controls and at risk patients (5%–20%). The predicted probabilities of death correlated with KCCQ scores (all p < 0.05). Interpretation: Echo-based ML models to predict mortality had good internal and external validity, were generalizable, correlated with patients’ quality of life, and are comparable to an established HF risk score. These models can be leveraged for automated risk stratification at point-of-care. Funding: Funding for Alberta HEART was provided by an Alberta Innovates - Health Solutions Interdisciplinary Team Grant no. AHFMR ITG 200801018. P.K. holds a Canadian Institutes of Health Research (CIHR) Sex and Gender Science Chair and a Heart & Stroke Foundation Chair in Cardiovascular Research. A.V. and V.S. received funding from the Mitacs Globalink Research Internship.http://www.sciencedirect.com/science/article/pii/S2352396423000440EchocardiographyMachine learningDeep learningMortalityHeart failurePrognostic models |
| spellingShingle | Akshay Valsaraj Sunil Vasu Kalmady Vaibhav Sharma Matthew Frost Weijie Sun Nariman Sepehrvand Marcus Ong Cyril Equibec Jason R.B. Dyck Todd Anderson Harald Becher Sarah Weeks Jasper Tromp Chung-Lieh Hung Justin A. Ezekowitz Padma Kaul Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context EBioMedicine Echocardiography Machine learning Deep learning Mortality Heart failure Prognostic models |
| title | Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context |
| title_full | Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context |
| title_fullStr | Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context |
| title_full_unstemmed | Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context |
| title_short | Development and validation of echocardiography-based machine-learning models to predict mortalityResearch in context |
| title_sort | development and validation of echocardiography based machine learning models to predict mortalityresearch in context |
| topic | Echocardiography Machine learning Deep learning Mortality Heart failure Prognostic models |
| url | http://www.sciencedirect.com/science/article/pii/S2352396423000440 |
| work_keys_str_mv | AT akshayvalsaraj developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT sunilvasukalmady developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT vaibhavsharma developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT matthewfrost developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT weijiesun developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT narimansepehrvand developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT marcusong developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT cyrilequibec developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT jasonrbdyck developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT toddanderson developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT haraldbecher developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT sarahweeks developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT jaspertromp developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT chungliehhung developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT justinaezekowitz developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext AT padmakaul developmentandvalidationofechocardiographybasedmachinelearningmodelstopredictmortalityresearchincontext |