The heart rate monitors Rob ter Horst did not test – why not?

Last Updated on July 21, 2023 by pg@petergamma.org

Rob ter Horst is Postdoctoral researcher at the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences in VIENNA, AUSTRIA. He invested a lot of time in testing 100 smart watches:

The Quantitative Scientist Rob ter Horst tested 100 sports watches: so what now?

Why did Rob not test the the following devices? The problem is not the lack of Rob’s qualification:

0. Scott Harden’s Sound Card ECG (single-lead 3 electrodes)

• The most affordable ECG device we know.
• Sound cards usually have 16 bit A/D converters.
• The sound card could eventually be replaced by a circuit like this:

Are you a YouTube video maker looking for a new topic? We suggest to attach the circuit of How To Electronics to Home Assistant (which supports InfluxDB & Grafana)

1. Bitalino Biomedical Toolkit (12 bit) & biosignalsplux (16 bit) with the HRV add on ECG toolbox

2. The OpenBCI (24 bit) 3 lead ECG with for instance HeartPy

• modules from the OpenBCI store and and from Ebay and Aliexpress

3. The Adinstruments Equivital sensor belt

• which claims to have a noise and motion artifact-free ECG
• The Equivital has a low A/D conversion, 10 – 12 bit, for exact values see our journal
• It is only a 2 lead ECG device.
• ECG artefact removal requires the Adinstruments LabChart software, which is very expensive.
• Therefore accuracy tests should be done preferentially by someone who already has the LabChart software, for instance by the cardiovascular and exercise physiologist Dr. Evan Matthews from Montclair State University, USA.

4. The iWORX HSK-TA KIT IX-TA-220 IWIRE-B3G TM-220 with iWork ECG toolbox

• which are among the most most affordable ECG devices used for research applications

• used on ebay:

https://www.ebay.ch/itm/185245471492

• 3 lead ECG, which means that 3 ECG channels are amplified independently from each other, which eventually gives the option to get more accurate data.
• 16 bit A/D conversion. For research applications at least 16 bit is recommended
• This device was also used for respiratory rate estimation in papers written for clinical applications

5. The Holter ECG devices

• Rob cites a paper in which a Polar H10 chest strap was compared to a Holter ECG. But why does Rob not include this device in his tests also?

• He mentioned problems at higher speeds with ECG devices. But 3 lead ECG devices are used for clinical applications and are known to be more accurate than optical heart rate monitors at slowers speeds. Why does Rob not do any tests about this device, to find out, where exactly the speed limit is, where ECG devices get incaccurate?

• One of the problems ECG devices have are ECG motion artifacts. The Adinstruments Equivital claims to be ECG noise and motion artifact-free. It would be very interesting to compare this sensor belt to the most accurate smart watches Rob tested.

We asked Rob several times to do this, but failed. Mabe someone else takes over and tests these devices, makes a demo or eventually writes papers about this topic?