Physiological multi-sensor devices & InfluxDB – at fist sight still a rare combination

Soldering and coding according to this instruction:

is one of the solutions for a 32 channel EEG devices. 32 channel EEG devices are expensive, and teh question arrises, if we want to build it by ourselves from ADS1299 product offers:

we have here listed a comparison table:

Hardly any manufacturers of physiological multi-sensor devices such as BIOPAC, Adinstruments, iWorx, g.tec, or Schiller medical seems to use the ADS1299 chip from Texas Instruments in their devices. We sah only rarely competitive offers for ADS1299 products, and these where temporarily, as for instance HackEEG. HackEEG 32 was sold for 2 090 USD :

https://www.crowdsupply.com/starcat/hackeeg

A ADS1299 products wich is currently available is PiEEG. But can we trust PiEEG? We do not know of independent tests for PiEEG.

The PiEEG support promotes products such as FreeEEG128-alpha:

https://pieeg.com/forum-pieeg-low-cost-brain-computer-interface/topic/create-a-new-daily-use-bci-to-pacient/#postid-40

But which physiologist uses EEG devices based FreeEEG128-alpha? The ADS131M08-Q1 chip which is also from Texas Instruments as the ADS1299:

https://www.ti.com/product/ADS131M08-Q1?utm_source=google&utm_medium=cpc&utm_campaign=asc-null-null-GPN_EN-cpc-pf-google-eu_cons&utm_content=ADS131M08-Q1&ds_k=ADS131M08-Q1&DCM=yes&gad_source=1&gclid=Cj0KCQjwxsm3BhDrARIsAMtVz6NaTkMOUH0_qUeN2C0xDth9IjeOyUH__e1zyjoKdXXBBv6Mt_edMH0aAsnhEALw_wcB&gclsrc=aw.ds

But as far as we know the specifications of this chip are only little suitable for research applications. If we make a first look into the scientific literature, physiological multi-sensor devices & InfluxDB is still a rare combination, which is used for instance in this paper:

Developing a Cross-Platform Application for Integrating Real-Time Time-Series Data from Multiple Wearable Sensors

https://www.mdpi.com/2673-4591/58/1/4

In the paper it says about physiological signals:

We are equipped to collect a variety of physiological signals from:

  • the Apple Watch
  • Android Wear devices
  • the Empatica E4 (this devices is discontinued and was expensive, it was 2500 $, remark by MRIS)

each device is offering unique datasets essential for comprehensive health monitoring. Apple Watch and Android Wear provide raw acceleration data and processed heart rate data. The real-time collection of these data types is optimized through native development.

The E4, however, provides a richer dataset. It captures:

  • Inter-Beat Interval (IBI)
  • Blood Volume Pulse (BVP)
  • 3-axis acceleration
  • Electrodermal Activity (EDA)
  • and skin temperature

We personally have let go of smartwatches for our applications, since we do not have any scientific data which proof that they are accurate enough:

https://petergamma.org/category/optical-heart-rate-sensor

Rob ter Horst’s YouTube videos are not convincing for us personally, on the contrary. Different is the situation with 3 channel ECG devices which are gold standard for runs up to about 15 km/h on a treadmill.We think the closer we look at these instructions:

The easier the become. We can build and and manufacture such devices by ourselves. The OpenBCI architecture is 10 years old now and well established, as well as InfluxDB. But as we can see in the example of Home Assistant, and in YouTube videos of the Swiss YouTuber Adrian Spiess, different architectures are often used combined. And we think it is worth to have a look also into open access university libraries as to following at the level of our qualification:

https://infozentrum.ethz.ch/en