It's also true that we entered into this period after simple ICs plus some wire and analog components (or Heathkit) let you do interesting things when you started to need a lot of expensive gear to do interesting hardware. That's still true in some cases. But gear like Arduinos and Raspberry Pis (and BLE devices etc.) once again let you do practical but interesting things at home.

Agreed. I didn't get into electronics as heavy years ago because of cost and lack of solid information, and this site has changed that for me.

I also messed with electronics since I was a kid with 100-in-1 kits from radio shack, but several years ago adafruit, sparkfun, eevblog, hackaday, and seeedstudio (Bunnie) helped provide me with the background info and fun hacker on-my-own-time-projects I eventually required to help build serious real world products - the kind surgeons use today and required approval from both the FDA and FCC.

Thanks to all of them! And happy hacking!

As someone who worked with the fda, do you think there could be some pre certified framework that greatly reduces the work and the certification burden of medical product developers?

Just ignore the enormous risk or cost. Or how the EU does it (only slightly faster) for one moment.

Given that you, the creator, define the risk (dMFEA [1]) AND the "quality system" [2] (which is actually a process and has little to do with QA from a software definition or understanding), my guess is that the entire system for approved devices, drugs, and their requisite clinical trials has evolved for the big guys to eat the medium guys and prevent the small guys from even entering the competition. The final nail in the coffin is that there is ABSOLUTELY NO 3rd party risk or security analysis at any time or level - only what you say, document, and submit (claims [3]). They are only involved once there's a lawsuit - and every piece of documentation by the creator is "burden of proof" than the accuser must disprove. Now imagine a 10,000 patient study - how do you prove the drug was bad because it, e.g. blinded one person?

[1] https://en.wikipedia.org/wiki/Failure_mode_and_effects_analy... [2] http://www.fda.gov/medicaldevices/deviceregulationandguidanc...

Edit: add claims in parens [3] http://www.fda.gov/MedicalDevices/ResourcesforYou/Industry/u...

Edit 2: Short answer, no. But there may be hope for simple pure-software app developers: http://www.fda.gov/downloads/MedicalDevices/.../UCM263366.pd...