4
\$\begingroup\$

Please note: I would have posted this question on HealthIT.SE but they are apparently now closed for business, and I believe this site is the next most appropriate place to ask this. I believe this question is on topic because it involves circuit verification and certification.


I'm very interested in the open source OpenBCI device and the bevy of apps and devices I could see it allowing integration with. Their open source license allows derivation (provided attribution), and I think this is going to open up flood gates with respect to commercial devices that build off of its hardware design.

In the commercial domain, I was curious as to what regulations/standards may come into focus for any companies selling OpenBCI and/or hardware componentry built up off of the OpenBCI mainboard. I ask because the biodata being streamed from the OpenBCI is EEG (brain wave) data, EMG data, EKG data, etc... this may very well constitute biomedical data that may be protected by things like HIPAA, etc.

I've heard of IEC 6061 but that is the only biomedical/technical safety standard I'm aware of.

So I ask: if I was building a commercial device that used OpenBCI as its core, and that was streaming EEG/EMG/EKG data from it, what sort of standards (besides IEC 6061) might I have to have the electronics verified/certified to?

\$\endgroup\$
  • \$\begingroup\$ I think this is a good question for Engineering Stack Exchange \$\endgroup\$ – Mahendra Gunawardena Apr 5 '16 at 10:33
  • 1
    \$\begingroup\$ OpenBCI is not a medical device, it cannot be used for diagnostic and/or therapeutic purposes. Whatever certifications a hypothetical device based on it may require fully depends on what device that is. \$\endgroup\$ – Dmitry Grigoryev Apr 5 '16 at 12:44
  • \$\begingroup\$ Thanks @DmitryGrigoryev (+1) - but when you say "OpenBCI...*cannot* be used for diagnostic and/or therepeutic purposes", what do you mean by "cannot"?! Do you mean cannot, as in, legally-cannot (if so, can you cite sources)? Or do you mean cannot from a technological perspective? Thanks for any clarification here! \$\endgroup\$ – smeeb Apr 5 '16 at 13:30
  • 1
    \$\begingroup\$ Yes, I meant legally-cannot. A simple look on the website and the device make it clear it is not certified for medical use. My point is, whatever certification is needed depends on the actual medical device, not on the capabilities of the components used to built it. That's why your question strikes me as overly broad and theoretical. \$\endgroup\$ – Dmitry Grigoryev Apr 5 '16 at 13:40
  • 1
    \$\begingroup\$ Thanks again @DmitryGrigoryev (+1) - then perhaps this is really the root of my question: if I am planning on using an OpenBCI-derived device for non-medical purposes, then am I required to satisfy any medical-related standard (FDA/IEC, etc.)? This may sound like an obvious "No!", but my concern is that, because the origin of the data is in fact a part of person's body, that perhaps it would automatically be classified as "biomedical" in nature, and therefore, subject to certain biomedical criteria. Thoughts? \$\endgroup\$ – smeeb Apr 5 '16 at 13:59
4
\$\begingroup\$

There is nothing wrong in using open source software or hardware to design medical devices. But at end of the day provider of these equipment has to comply with regulation in jurisdiction the devices are sold. I have experiences ubuntu (Linux) computers used in US doctors offices.

In general in USA and Europe the devices have to be compliant to government regulations such as FDA - 21 CFR 820 (I have attach references below). IEC 6061 which you have referenced is commonly used in both and USA and Europe. This document contains product performance in with respect to EMC/EMI and others.

One of the major issue with open source software and hardware is compliance to quality requirements. There are many more ...

All in all there are companies starting medical device development work using open source tools but migrate into traditionally accept development process as the product is getting closer to regulatory approval.


References:

\$\endgroup\$
  • 1
    \$\begingroup\$ And then if any of these devices pass data, that equipment is subject to another set of standards that it must comply with. \$\endgroup\$ – user65586 Apr 5 '16 at 12:56
  • \$\begingroup\$ Thanks @MahendraGunawardena (+1) - I appreciate the solid feedback! \$\endgroup\$ – smeeb Apr 5 '16 at 13:31
  • \$\begingroup\$ Thanks @jdv (+1) - what sort of other standards are you referring to? \$\endgroup\$ – smeeb Apr 5 '16 at 13:31
2
\$\begingroup\$

This question is more legal than engineering one, and I'm not a lawyer. However, my understanding is that a medical device is a function, not a technological capability. If the device is used to diagnose or threat actual patients, it is a medical device and has to be certified as such. If you built a toy brain wave monitor, no certification is required. You may however be subject to other regulations due to the wearable nature of your device (I bet they will forbid you to power it from mains and use lead in head-attached probes), as well as general regulations for consumer electronics.

As such lots of devices deal with biometric data (e.g. eye-tracking cameras, fingerprint sensors and the like), yet they are not medical devices. Even when people are lead to assume it is one, a prompt warning seems to be enough:

enter image description here

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.