Hera Leto Two earphones for HRV analysis

Blog post by Marco Altini

I was recently contacted by Kow Ping, co-founder of Well Being Digital Limited (WBD101), a company which makes heart rate sensing technology for hearables, as we have been in touch regarding PPG-based technologies for HRV analysis.

WBD101 makes technology which can be embedded in different products, with the goal of providing high quality heart rate data. The technology has been embedded in the Hera Leto Two earphones, sold by Actywell.

Kow has been kind enough to send me a device for testing, and therefore I report here an initial validation showing really good RR intervals when compared to both our camera-based solution and our trusted reference chest strap, the Polar H7.

Measurement setup

Data was acquired using the following devices at the same time:

• Hera Leto Two earphones, connected to a device running the HRV Logger app
• Polar H7 (previously validated with respect to ECG here), also connected to a device running the HRV Logger app
• Phone camera and the Camera Heart Rate Variability app 

During data acquisition, I collected data a few minutes while breathing freely, and a few minutes while deep breathing, to elicitate higher HRV due to RSA. You will see in the plots below visually the effect of deep breathing as we get greater swings in RR intervals.

A final note on data synchronization: data cannot be perfectly synchronized because it is not timestamped by the sensors. What we can do is either to log real time and then to split data in windows based on when data was collected, then compute HRV features on these windows or to sum up RR intervals over time. For this analysis we went with the second option and also tried to visually align the data streams.

As usual, if you use the camera for your measurements, please double check our camera based measurement best practices, to make sure you'll collect high quality data on which meaningful analytics can be derived. ​

RR intervals

We will start by looking at RR intervals, the basic unit we need to compute HRV features. RR intervals (peak to peak differences in consecutive heart beats) are provided by the two sensors directly, so we don't really need to do much to collect them, apart from linking the sensor to the HRV Logger app and export the csv files.

As a third comparison, we will add also an RR intervals time series collected using the phone camera, which is the method we have introduced and normally recommend using.

What can we derive from these data? You can see clearly  almost perfect correlation between Polar H7 and Hera Leto Two and Phone Camera for all conditions (relaxed vs paced breathing as highlighted by bigger oscillations in RR intervals or instantaneous heart rate), meaning that the sensor works really well.

Heart Rate Variability: rMSSD

As features, we will look only at rMSSD. rMSSD is a clear marker of parasympathetic activity and the main feature we use for our analysis in HRV4Training, from which we derive Recovery Points (check out this post for an overview of HRV features).

What we expect given the data above is to see extremely close values between the Polar H7 chest strap and Hera Leto Two, as well as the camera based measurements.

For the plot below, I computed rMSSD for each 60 seconds time window:

Results are very good considering normal variation in physiology and limitations in data synchronization.

We can see how in the first 3 minutes results also include little difference between consecutive minutes, as the protocol there involved simply self-paced breathing, similarly to what we recommend doing in the morning. rMSSD then increases as we enter the deep breathing phase, and finally reduces post deep breathing session.