EUVIMEDEUVIMED entdecken
ValidierungsstudieOpen Access · CC BY-NC-ND 4.0

Prediction of cardiovascular risk factors from retinal fundus photographs: Validation of a deep learning algorithm in a prospective non‐interventional study in Kenya

Tom White, Viknesh Selvarajah iD, Fredrik Wolfhagen-Sand iD, Nils Svangård, Gayathri Mohankumar, Peter Fenici, Kathryn Rough, Nelson Onyango, Kendall Lyons, Christina Mack, Videlis Nduba, Mansoor Noorali Saleh und 8 weitere

Diabetes, Obesity and Metabolism · 2024-07 · 26 (7) · 2722-2731

DOI: 10.1111/dom.15587 ↗

Kurz gefasst

Validierungsstudie in der Zeitschrift Diabetes, Obesity and Metabolism, online erschienen am 15.04.2024. Verfasst von Tom White und 19 weiteren Autor:innen. Beteiligte Einrichtung: Data Science and Advanced Analytics, Data Science & Artificial Intelligence, R&D, AstraZeneca, Cambridge, UK. Hauptthemen laut MEDLINE: Deep Learning, Cardiovascular Diseases, Heart Disease Risk Factors. Die Arbeit wurde bisher 19-mal zitiert (Crossref) und nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: AstraZeneca.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Aim

Hypertension and diabetes mellitus (DM) are major causes of morbidity and mortality, with growing burdens in low‐income countries where they are underdiagnosed and undertreated. Advances in machine learning may provide opportunities to enhance diagnostics in settings with limited medical infrastructure. Materials and Methods A non‐interventional study was conducted to develop and validate a machine learning algorithm to estimate cardiovascular clinical and laboratory parameters. At two sites in Kenya, digital retinal fundus photographs were collected alongside blood pressure (BP), laboratory measures and medical history. The performance of machine learning models, originally trained using data from the UK Biobank, were evaluated for their ability to estimate BP, glycated haemoglobin, estimated glomerular filtration rate and diagnoses from fundus images.

Results

In total, 301 participants were enrolled. Compared with the UK Biobank population used for algorithm development, participants from Kenya were younger and would probably report Black/African ethnicity, with a higher body mass index and prevalence of DM and hypertension. The mean absolute error was comparable or slightly greater for systolic BP, diastolic BP, glycated haemoglobin and estimated glomerular filtration rate. The model trained to identify DM had an area under the receiver operating curve of 0.762 (0.818 in the UK Biobank) and the hypertension model had an area under the receiver operating curve of 0.765 (0.738 in the UK Biobank).

Conclusions

In a Kenyan population, machine learning models estimated cardiovascular parameters with comparable or slightly lower accuracy than in the population where they were trained, suggesting model recalibration may be appropriate. This study represents an incremental step toward leveraging machine learning to make early cardiovascular screening more accessible, particularly in resource‐limited settings.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningCardiovascular DiseasesHeart Disease Risk FactorsHumansKenyaMaleFemaleMiddle AgedProspective StudiesAdultHypertensionAlgorithmsPhotographyFundus OculiAgedDiabetes MellitusRisk FactorsDiabetic Retinopathy

Stichwörter der Autor:innen

UK Biobankcardiovascular screeningdiabetes mellitushypertensionmachine learningpredictive modellingretinal fundus photographs

Autor:innen und Institutionen

  • Data Science and Advanced Analytics, Data Science & Artificial Intelligence, R&D, AstraZeneca, Cambridge, UK.
  • Research and Early Development, Cardiovascular, Renal and Metabolism (CVRM), BioPharmaceuticals R&D, AstraZeneca, Cambridge, UK.
  • Data Science and Advanced Analytics, Data Science & Artificial Intelligence, R&D, AstraZeneca, Gothenburg, Sweden.
  • Centre for Artificial Intelligence, Data Science & Artificial Intelligence, R&D, AstraZeneca, Gaithersburg, Maryland, USA.
  • School of Medicine and Surgery, Catholic University, Rome, Italy.
  • Biomagnetism and Clinical Physiology International Center (BACPIC), Rome, Italy.

ORCID-Profile: Viknesh Selvarajah, Fredrik Wolfhagen-Sand, Victoria E R Parker

Förderung

  • AstraZeneca

Zitieren

APA

White, T., Selvarajah, V., Wolfhagen-Sand, F., Svangård, N., Mohankumar, G., Fenici, P., Rough, K., Onyango, N., Lyons, K., Mack, C., Nduba, V., Noorali Saleh, M., Abayo, I., Siddiqui, A., Majdanska-Strzalka, M., Kaszubska, K., Hegelund-Myrback, T., Esterline, R., Manzur, A., & Parker, V. E. R. (2024). Prediction of cardiovascular risk factors from retinal fundus photographs: Validation of a deep learning algorithm in a prospective non‐interventional study in Kenya. Diabetes, Obesity and Metabolism, 26(7), 2722-2731. https://doi.org/10.1111/dom.15587

Vancouver

White T, Selvarajah V, Wolfhagen-Sand F, Svangård N, Mohankumar G, Fenici P, et al. Prediction of cardiovascular risk factors from retinal fundus photographs: Validation of a deep learning algorithm in a prospective non‐interventional study in Kenya. Diabetes, Obesity and Metabolism. 2024;26(7):2722-2731. doi:10.1111/dom.15587

BibTeX
@article{White2024,
  author = {White, Tom and Selvarajah, Viknesh and Wolfhagen-Sand, Fredrik and Svangård, Nils and Mohankumar, Gayathri and Fenici, Peter and Rough, Kathryn and Onyango, Nelson and Lyons, Kendall and Mack, Christina and Nduba, Videlis and Noorali Saleh, Mansoor and Abayo, Innocent and Siddiqui, Afrah and Majdanska-Strzalka, Malgorzata and Kaszubska, Katarzyna and Hegelund-Myrback, Tove and Esterline, Russell and Manzur, Antonio and Parker, Victoria E R},
  title = {Prediction of cardiovascular risk factors from retinal fundus photographs: Validation of a deep learning algorithm in a prospective non‐interventional study in Kenya},
  journal = {Diabetes, Obesity and Metabolism},
  year = {2024},
  volume = {26},
  number = {7},
  pages = {2722--2731},
  doi = {10.1111/dom.15587}
}
Kostenlos starten →