Understanding the Differences in HRV calculation via Dataquest A.R.T. Analysis vs. Ponemah
Dataquest A.R.T. reached its End of Life (EOL) in July 2015 and is no longer available for purchase. For information on continued support, please see DSI Software Support Lifecycle policy for software product support and discontinuation guidelines.Due to technology upgrades and software compatibility, DSI cannot guarantee the functionality of Dataquest ART on computers manufactured after July 2015.
Background:
HRV calculation is a method to determine the variation of the heart beats over time. This can be done by interpolating the IBI (inter-beat-interval) to produce a continuous signal, which an FFT can then be applied to extract its frequencies components. The frequencies components are then broken down into bins, which can then be compared to determine the variation of the heart beats over time.
Differences in calculations:
- IBI derivation
Dataquest A.R.T. Analysis and Ponemah each have its own ways of marking the ECG or BP waveform to extract the IBI. There will be instances that either of the software platforms will miss a beat or two at the beginning/ending of the signal. This will contribute the amount of data available when applying an FFT to the signal; hence, slight variation in the result will be noticeable.
- Default FFT Windowing Methods
Dataquest ART Analysis Ponemah Windowing Method Rectangular Hanning Overlapping Sub-series 1 5
- Scaling Factor
By default, Dataquest A.R.T. Analysis scales the HRV result by a factor of 2.7. This is due the differences in the algorithms when applying the FFT to the continuous IBI signal.
Example
When both software platforms use the same calculation method and a scaling of 2.7 in Ponemah:
Dataquest ART Analysis | Ponemah | |
Windowing Method | Hanning | Hanning |
Overlapping Sub-series | 5 | 5 |
Scaling | 1 | 2.7 |
Note: The data set in the example above was edited so that the beats were marking the same on both platforms. Trying to replicate maybe yield different results.
Comments
0 comments
Please sign in to leave a comment.