Last Updated on August 19, 2023 by pg@petergamma.org
Rob ter Horst uses his Polar H10 chest strap as a reference device and says it is his gold standard ECG device for research. According to this information is possible to build a highly accurate ECG device for a very prize:
The Apple watch Ultra currently costs about 800 USD at digitec Switzerland:
A Cyton board with issue-free WIFI shield currently costs around 370 USD on eBay:
https://www.ebay.com/itm/284876280716
Rob ter Horst says his Polar H10 chest strap is the most accurate device with a signal quality of 99.6. Among the smartwatches Rob ter Horst tested the Apple Watch Ultra is the most accurate one. Rob furthermore says that a Schiller Medilog AR12 Plus Holter ECG recorder is not as accurate as a Polar H10 if intense activities with strong body movements are investigated. But the Schiller Medilog AR12 Plus has no options to handle with noisy ECG signals. Even the open source Python toolbox HeartPy can handle with very noisy ECG signals.
We previously have discussed in this journal, how many features for instance the iWorx ECG toolbox offers to analyse ECG signals. The iWorx toolbox costs 1000 USD, but other ECG toolboxes are free. There are for instance Matlab ECG toolboxes which are used for research applications which are free. And we suppose that the Python open source toolbox HeartPy offers similar or even better specifications than the Matlab ECG toolboxes.
If we start building a 3 lead ECG device based on Cyton and HeartPy and start testing the HeartPy toolbox and other algorithms which are available to handle with noisy ECG signals, are we not able to challenge Rob ter Horst? Is it possible to build an ECG device which can compete with the heart rate monitors which are in the champions league in Rob ter Horsts accuracy plots? To summarize some specifications of the devices we discussed here we have written a comparison table of several devices:
| Device | Hardware | Software | Sampling Rate | Channels | Leads | Electrodes | can handle with noisy ECG signals | Costs | remarks | |
| 1 | iWorx | TA 220 | iWorx ECG Toolbox $1000 | 16 | 3 | 3 | 5 | ECG artifact removal caused by acitivity is available | used on ebay $1400 | We previously discussed in our journal the different features which this device offers. |
| 2 | Schiller Medilog AR12 Plus Holter Recorder | GE Healthcare USA | Fire of Life (included with the device ?) | 12 | 3 | 7 | no, only motion detection available | $1700 | Rob ter Horst cites a paper (*2) which demonstrates that this device cannot compete with the Polar H10 chest strap as far as accuracy is concerned if intense activities with strong body movements are investigated. | |
| 3 | OpenBCI | Cyton | we suggest to use the Python open source toolbox HeartPy | 24 | 8 | 3 | 5 | can handle with very noisy ECG signals | new for $ 370 on eBay with WIFI shield | A paper (*1) says that a 3 lead ECG device based on Cyton can be used instead of an Adinstruments device which costs $ 15’000 for certain applications. A second paper (*3) says that Cyton can be used for neurophysiological research. |
| 4 | Apple watch | Ultra | we do not know of software from Apple who analyses PPG signals from Apple watches | unknown | 4 optical heart rate sensors | none | none | yes | $ 800 | Rob ter Horst says the Apple watch Ultra is among the most accurate optical HRMs he tested, but no scientific paper has confirmed this to this date. |
| 5 | Polar | H10 | We do not know of software from Polar who analyse ECG signals from chest straps. | unknown | unknown | unknown | 2 | yes | $ 90 | Rob ter Horst says that the Polar H10 ECG chest strap has the best heart rate and is the gold standard in research with a signal quality of 99.6 %. Rob cites a paper (*2) which was partly funded by Polar as source on which his statements are based on.. |
(*1)
(*2) Rob ter Horst: “Polar H10 Scientific Review: Best for Heart Rate (99.6% Accurate)“
paper which Rob ter Horst reviews in this video.
https://pubmed.ncbi.nlm.nih.gov/31004219/
(*3)