Non-compliant portable EEG devices

Lindsay Robertson

I need clarification around use of portable EEG devices, that do not have compliance with medical equipment standards.

I'm in a university departmental research and teaching environment. My head of department is very keen to do some outreach, using cheap readily available devices that do not conform to medical standards. They want to be able to go to outreach events, and put one of these devices on children attending and show them some very neat stuff.

My technical team have always insisted that all eeg activities in the dept use equipment that conforms to national and international safety standards for electrical medical devices. IEC 60601-1, AS/NZS 2500, AS/NZS 3551 etc.

Our understanding is that the medical standards only apply to therapeutic health care uses, however in the event of an incident, a court would expect us to have the same level of safety.

We have said we think that these devices should not be used in a research or education context, but the Head of department and other staff are often seeing other researchers using them, publishing papers, and doing very interesting outreach activities with them.

The world seems to be settling on the phrase “consumer eeg” devices, but I think that’s not very appropriate.
The devices are very simple to use, have very slick websites and are rapidly growing sales numbers.

Does anyone have any advice?

Senior Technician & Health and Safety Officer

University of Otago | Ōtākou Whakaihu Waka
Psychology | Te Tari Whakamātau Hinekaro
William James building, 275 Leith Walk, Dunedin | Ōtepoti
New Zealand


Peter Selvey

Super Moderator
If the EEG is powered from a PC or laptop's USB socket, then in reality the chance of something serious happening is pretty low, assuming the PC has it's own regulatory approvals (e.g. a well recognized brand, bought from a reputable store, not too old), and assuming the user is only connected for a short time and is in a normal environment. There are theoretical safety issues which result in medical standards having stricter rules, but the rules for a modern laptop/PC are also pretty strict and would be effective as well (otherwise, people would be getting shocked from PCs all the time!). However, keep in mind that in the rare chance that something did go wrong, it could be an uphill battle proving there was no negligence or getting insurance to cover the incident, even if the use of non-medical stuff had nothing to do with an incident.

If the EEG unit has it's own power supply then don't touch it. Taking an unknown, potentially dodgy power supply and then using that to power an electronic device that has electrodes making good electrical contact with a "human subject" is clear cut case of negligence. Go to jail, do not pass go kind of stuff.

Lindsay Robertson

Thanks Peter.
Generally these devices are internally powered, and we would only consider using that type. Ones with internal batteries, and use bluetooth or wifi to communicate to a laptop.

These devices use electrodes, that lower skin resistance to current flow, so the computer analogy is not very accurate in this case.

Peter Selvey

Super Moderator
Yes, I'm aware that an EEG has electrodes, the point is that while there is a little more risk from a low impedance electrode contact, it's only marginal in the environment described (at a show, as a demonstration). But anyway, if the device is internally powered/wireless it's even better.

The risks from wireless/internally powered device with electrode contact are:
- small dc currents can cause tissue necrosis, however this takes a long time (hours)
- currents in the order of a few mA can cause intended muscle action, and with a small contact area can cause burns, but it would have to be a really badly designed for this occur in practice, given the normal front end design of an EEG. These currents would not exist in normal use, but could appear in fault e.g. an input ESD diode shorts. But again, the typical EEG input circuit would usually require two or more broken components to get significant currents, and modern electronic components are very reliable.
- if the device uses impedance sensing to detect if the electrodes are attached, a fault in a poorly designed circuit again could lead to moderate currents. But again this is a stretch in practice.

Although these are not really plausible in practice, it would be good to have the devices tested to be sure that such currents are not possible in normal and fault condition. That said, most third party test labs are often not skilled in inspecting circuit diagrams to that level, so even a so called "IEC 60601-1 compliant" device might not have this level of analysis.

Lindsay Robertson

Yeah, we're not in a position do to destructive testing, but if circuit diagrams were available that would be hugely helpful, but I figure there's no chance of that. and theres no way of knowing how any particular device is wired.
The issue seems to be more around acceptance of the level of risk, and its difficult to find published work on what that risk actually is... so your info is valuable thank you.
Your numbers are a bit lower than I was thinking. Here's what I rationalised.
If a device had an 18v battery, and say a best case electrode impedance of 1k, assuming the fault somehow applied the battery directly to an electrode, 18/1000 = 18mA. And that on the scalp might have some adverse effects on the brain.
4ma DC is the accepted safe limit for Transcranial Direct Current Stimulation.
Do you think those numbers are too high?

In an outreach situation the unit would get removed by the supervisor, but I'm not in a position to judge the risk of 18mA for a few seconds in the brain. Seizure? Stroke?


Involved In Discussions
Hm, you are in a tough position.
To ease your equation a little you need two electrodes to pass such a current through a body, so you may divide your worst case scenario by 2 :)
But if covering your behind is what you are aiming at then the only way to go is to use certified medical devices.

On the other hand: to get an EXG device certified some aspects have to pass tests that are beyond your interest. Like you may decide that such a device is allowed to brick when receiving and ESD pulse while essential performance may dictate that it must get to a certain working state.
Another example: EEG devices must have a way of limiting the current in the single fault condition that the input amplifier fails. This under the assumption that the input amp fails from very high input impedance to short circuit to the worst case rail.
Of course the chance that an input amp fails this way is very very low, and if you have a system that visualizes the input values real-time you can usually see directly that this channel is faulty.
Finally: for medical devices one must assume that the patient is incapacitated, but this is very unlikely in a class room.

So, it come to this: how confident are you that the devices are reasonably serious built and that the person operating the stuff is knowing what he/she is doing?

(for what is worth: when my kids were in primary school I was one of the helping parents with the so called "Physics: electricity box" to teach them the basics of electricity. The way they worked with batteries, lamps, wires etc. resulted in various short circuits, blowing lamps, holding wires against other body parts than hands, etcetera. If this box was labeled "Medical devices box" it would not have passed any regulation.)

Lindsay Robertson

Thanks Loekje, thats a good summation, and good information.
It is very difficult to get a feel for the quality of the company and their designs. Most of them have very fancy and slick websites - and that puts me off!
There are a some factors you raise that increase the risk.
  • Many of these devices don't show the traces in realtime, all the time.
  • If you had one bad trace I think most academics would assume the impedance was high, not low, and not be very concerned. That could probably be mitigated with training.
  • Can you believe the very generalised statements made by companies trying to sell their product. As a retail purchaser you don't get to ask questions of the engineers, you only get to talk to the sales team.

The nice thing for me is that I'm a lowly paid, bottom of the hierarchy, tech. I'm not paid to make decisions like this, I can just gather info, make a recommendation and pass it up to the bosses. :)
Top Bottom