How accurate is the measurement?Updated a year ago

Evaluation of algorithm performance is performed by stepwise comparative analyses with current state-of-the-art medical devices (e.g., pulse oximeter) or questionnaires (e.g., PHQ-9, GAD-7) . Accuracy and precision are checked by calculating, for example, Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and Correlation Coefficient (R).

Currently, over 1 million data points (resulting from various collaborations with clinical organizations) are actively used for testing and validating the models.

Please refer to the validation and precision report (see Appendix) for an in-depth analysis of how the software is tested and validated. In summary, the accuracy and precision levels of the product can be found in the table below.