Table 2

Diagnostic performance of the machine learning algorithm before and after revisions to the automatic real-time user feedback that significantly increased the signal quality of the PPG recordings, with AFL recordings excluded

Diagnostic performanceBefore revisionAfter revisionP
Participants, n19684
Recordings, n12 8265179
Recording duration60 s60 s
PPG recordings with insufficient quality, n (%)526 (4.1)0 (0)<0.0001
PPG recordings with low certainty algorithmic classification, n (%)164 (1.4)29 (0.6)<0.0001
Sensitivity (%)99.6 [99.4–99.8]100.0 [100.0–100.0]0.0411
Specificity (%)99.6 [99.5–99.7]99.8 [99.6–99.9]0.1170
Accuracy (%)99.6 [99.5–99.7]99.8 [99.7–99.9]0.0195
PPV (%)99.2 [98.9–99.5]99.2 [98.7–99.8]0.9137
NPV (%)99.8 [99.7–99.9]100.0 [100.0–100.0]0.0065
Diagnostic performanceBefore revisionAfter revisionP
Participants, n19684
Recordings, n12 8265179
Recording duration60 s60 s
PPG recordings with insufficient quality, n (%)526 (4.1)0 (0)<0.0001
PPG recordings with low certainty algorithmic classification, n (%)164 (1.4)29 (0.6)<0.0001
Sensitivity (%)99.6 [99.4–99.8]100.0 [100.0–100.0]0.0411
Specificity (%)99.6 [99.5–99.7]99.8 [99.6–99.9]0.1170
Accuracy (%)99.6 [99.5–99.7]99.8 [99.7–99.9]0.0195
PPV (%)99.2 [98.9–99.5]99.2 [98.7–99.8]0.9137
NPV (%)99.8 [99.7–99.9]100.0 [100.0–100.0]0.0065

Corresponding 95% confidence intervals are shown within brackets.

AFL, atrial flutter; NPV, negative predictive value; PPG, photoplethysmography; PPV, positive predictive value.

Table 2

Diagnostic performance of the machine learning algorithm before and after revisions to the automatic real-time user feedback that significantly increased the signal quality of the PPG recordings, with AFL recordings excluded

Diagnostic performanceBefore revisionAfter revisionP
Participants, n19684
Recordings, n12 8265179
Recording duration60 s60 s
PPG recordings with insufficient quality, n (%)526 (4.1)0 (0)<0.0001
PPG recordings with low certainty algorithmic classification, n (%)164 (1.4)29 (0.6)<0.0001
Sensitivity (%)99.6 [99.4–99.8]100.0 [100.0–100.0]0.0411
Specificity (%)99.6 [99.5–99.7]99.8 [99.6–99.9]0.1170
Accuracy (%)99.6 [99.5–99.7]99.8 [99.7–99.9]0.0195
PPV (%)99.2 [98.9–99.5]99.2 [98.7–99.8]0.9137
NPV (%)99.8 [99.7–99.9]100.0 [100.0–100.0]0.0065
Diagnostic performanceBefore revisionAfter revisionP
Participants, n19684
Recordings, n12 8265179
Recording duration60 s60 s
PPG recordings with insufficient quality, n (%)526 (4.1)0 (0)<0.0001
PPG recordings with low certainty algorithmic classification, n (%)164 (1.4)29 (0.6)<0.0001
Sensitivity (%)99.6 [99.4–99.8]100.0 [100.0–100.0]0.0411
Specificity (%)99.6 [99.5–99.7]99.8 [99.6–99.9]0.1170
Accuracy (%)99.6 [99.5–99.7]99.8 [99.7–99.9]0.0195
PPV (%)99.2 [98.9–99.5]99.2 [98.7–99.8]0.9137
NPV (%)99.8 [99.7–99.9]100.0 [100.0–100.0]0.0065

Corresponding 95% confidence intervals are shown within brackets.

AFL, atrial flutter; NPV, negative predictive value; PPG, photoplethysmography; PPV, positive predictive value.

Close
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close

This PDF is available to Subscribers Only

View Article Abstract & Purchase Options

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Close