Accuracy of Artificial Intelligence-Based Technologies for the Diagnosis of Atrial Fibrillation: A Systematic Review and Meta-Analysis

Figures

• 

  Figure 1. Flow chart (n: number).

• 

  Figure 2. Assessment of risk of bias and applicability of the PPG studies. (Väliaho et al., 2019 (A): testing the AFEvidence algorithm; Väliaho et al., 2019 (B): testing the COSEn algorithm; Väliaho et al., 2021 (A): testing device performance when time interval between every measurement is 10 min; Väliaho et al., 2021 (B): testing device performance when time interval between every measurement is 20 min; Väliaho et al., 2021 (C): testing device performance when time interval between every measurement is 30 min; Väliaho et al., 2021 (D): testing device performance when time interval between every measurement is 60 min; Dörr et al., 2019 (A): testing performance of device when recording for 1 min; Dörr et al., 2019 (B): testing performance of device when recording for 3 min; Dörr et al., 2019 (C): testing performance of device when recording for 5 min).

• 

  Figure 3. PPG group: (a) Forest plot of sensitivity; (b) forest plot of specificity. (Väliaho et al., 2019; (A): testing the AFEvidence algorithm, Väliaho et al., 2019; (B): testing the COSEn algorithm, Väliaho et al., 2021; (A): testing device performance when time interval between every measurement is 10 min, Väliaho et al., 2021; (B): testing device performance when time interval between every measurement is 20 min, Väliaho et al., 2021; (C): testing device performance when time interval between every measurement is 30 min, Väliaho et al., 2021; (D): testing device performance when time interval between every measurement is 60 min, Dörr et al., 2019; (A): testing performance of device when recording for 1 min, Dörr et al., 2019; (B): testing performance of device when recording for 3 min, Dörr et al., 2019; (C): testing performance of device when recording for 5 min).

• 

  Figure 4. PPG group random effects meta-analysis.

• 

  Figure 5. Subgroup analysis of the PPG studies (inpatients vs. outpatients).

• 

  Figure 6. Assessment of risk of bias and applicability of the single-lead ECG studies. (Desteghe et al., 2017(A): testing the AliveCor in the cardiology ward population; Desteghe et al., 2017 (B): testing the MyDiagnostick in the cardiology ward population; Desteghe et al., 2017 (C): testing the AliveCor in the geriatric ward population; Desteghe et al., 2017 (D): testing the MyDiagnostick in the geriatric ward population; Ford et al., 2022 (A): testing the Apple Watch 4; Ford et al., 2022 (B): testing the KardiaBand; Fu et al., 2021 (A): testing the device in supine position, Fu et al., 2021 (B): testing the device in upright position, Fu et al., 2021 (C): testing the device after individuals climbed to the 3rd floor; Santala et al., 2021 (1): published in October 2021, Santala et al., 2021 (2) (A): published in May 2021 and testing the device between the palms, Santala et al., 2021, (2) (B): published in May 2021 and testing the device in the chest; Abu-Alrub et al., 2022(A): testing the Apple Watch 5, Abu-Alrub et al., 2022 (B): testing the Samsung Galaxy Watch Active 3, Abu-Alrub et al., 2022 (C): testing the Withings Move ECG; Wegner et al., 2020 (A): testing the lead I; Wegner et…

• 

  Figure 7. Single-lead ECG group: (a) forest plot of sensitivity, (b) forest plot of specificity. (Desteghe et al., 2017 (A): testing the AliveCor in the cardiology ward population; Desteghe et al., 2017 (B): testing the MyDiagnostick in the cardiology ward population; Desteghe et al., 2017 (C): testing the AliveCor in the geriatric ward population; Desteghe et al., 2017 (D): testing the MyDiagnostick in the geriatric ward population; Ford et al., 2022 (A): testing the Apple Watch 4, Ford et al., 2022 (B): testing the KardiaBand; Fu et al., 2021 (A): testing the device in supine position, Fu et al., 2021 (B): testing the device in upright position, Fu et al., 2021 (C): testing the device after individuals climbed to the 3rd floor; Santala et al., 2021 (1): published in October 2021, Santala et al., 2021 (2) (A): published in May 2021 and testing the device between the palms, Santala et al., 2021 (2) (B): published in May 2021 and testing the device in the chest; Abu-Alrub et al., 2022 (A): testing the Apple Watch 5, Abu-Alrub et al., 2022 (B): testing the Samsung Galaxy Watch Active 3, Abu-Alrub et al., 2022 (C): testing the Withings Move ECG, Wegner et al., 2020 (A): testing the lead I,…

• 

  Figure 8. Single-lead ECG group: random effects meta-analysis.

• 

  Figure 9. Single-lead ECG group: subgroup analysis (inpatients vs. outpatients).

• 

  Figure 10. Single-lead ECG group: subgroup analysis (duration of index test).

• 

  Figure 11. Miscellaneous technologies—assessment of risk of bias and applicability. (Lown et al., 2018 (A): testing the Watch BP, Lown et al., 2018 (B): testing the Polar H7 device, Lown et al., 2018 (C): testing the Bodyguard 2 device; D: domain; RoB: risk of bias; AC: applicability).