Table 3.
Minute signals used as the source signals for feature extraction in the CDI model.
| Signal name | Signal description |
|---|---|
| Heart rate | 10% Trim mean average of beat to beat heart rate values |
| Time domain heart rate variability | 10% Trim mean average of time domain heart rate variability |
| Respiration rate | 10% Trim mean average of respiration rate |
| Count of magnitude | Average of activity count using the 3-axis accelerometer vector magnitude |
| Gross activity | Average of root mean squared of 3-axis accelerometer signal |
| Magnitude of uni counts | Average of vector magnitude of 3-axis univariate accelerometer activity counts |
| A-Fib percent | Percent of time classified as exhibiting atrial fibrillation or atrial flutter |
| Sleep | Percent of time classified as sleeping |
| Step count | Number of steps |
| Tilt | 10% Trim mean of tilt angle of the torso (where 90 degrees is upright) |
| Trailing activity | Average of gross activity after application of a 3 minute moving average filter |
| Walk percent | Percent of time classified as ambulation (inclusive of walking and running) |
| Skin temperature | Median skin temperature, in degrees celcius |
| Activity residual | 10% Trim mean average of trailing activity residual |
| Heart rate residual | 10% Trim mean average of heart rate residual |
| Respiration rate residual | 10% Trim mean average of respiration rate residual |
| MCI | 10% TrIm Mean Average of MCI (multivariate change index)[?] |
The minute source signals are generated by the pinpointIQ™ platform.