Talk:TeleHealth Hardware: Difference between revisions

From OLPC
Jump to navigation Jump to search
m (typo)
 
(9 intermediate revisions by 5 users not shown)
Line 8: Line 8:
* [http://www.create.ucsb.edu/~dano/CUI/ Create USB interface] - USB sensor input board. Cool but expensive. PIC.
* [http://www.create.ucsb.edu/~dano/CUI/ Create USB interface] - USB sensor input board. Cool but expensive. PIC.
* [http://www.hobbylab.us/default.aspx PC USB scope] - oscilloscope peripheral that attaches to normal computers via USB. Bloody expensive, Windows-only software... interesting to imitate (and a possible development tool for people with Windows desktops that don't have scopes, until we can make Measure a lovely o-scope interface.)
* [http://www.hobbylab.us/default.aspx PC USB scope] - oscilloscope peripheral that attaches to normal computers via USB. Bloody expensive, Windows-only software... interesting to imitate (and a possible development tool for people with Windows desktops that don't have scopes, until we can make Measure a lovely o-scope interface.)

== Hardware development ==

In regarding to the different projects that have been started would be nice to have a common heading in regarding to the implementation of a circuit that could handle all the peripherals or at least the majority of them..i thing is necessary to have an all in one circuit.

I'm not aware of any conditions under which the telehealth module or any part of it would need to be sterilized. Even in open-heart surgery these instruments are not in the body or even in the sterile field. Washable would be good. [[User:Niels|Niels]] 17:35, 18 May 2008 (EDT)

The ECG pads are a high failure rate part, so they will need to be disposable or at least cleanable and replaceable. Many pads in the US use a snap button (like those found in children's clothing), where the wire has the female side and the disposable pad has the male side of the snap bonded to a paper pad with adhesive on the reverse side. [[User:Niels|Niels]] 17:35, 18 May 2008 (EDT)

A twelve lead ECG can by reproduced in software from a five lead set-up. Basically you still have leads LA, RA, and LL, but you only use two leads for the precordials and interpolate the rest of the precordials. The precordials (V1 - V6) don't really tell you a heck of a lot besides R wave progression and sometimes ST-segment changes are most noticeable there for reasons of geometry that are dependent on the patient to begin with. [[User:Niels|Niels]] 17:35, 18 May 2008 (EDT)

There's a good book on the basic physics of ECG. I'm several hours from my library for the next month, but I think it's Rosen's Cardiac Electrophysiology, A Textbook. [[User:Niels|Niels]] 17:35, 18 May 2008 (EDT)

== Medical Devices Regulations ==

Little consideration appears to have been given to complying with medical devices regulations. These are indeed very onerous however this is for good reason. Any electrical device attached to a person has the potential to kill in ways you probably cannot imagine. You need more than just optical isolation between the sensor and any interface circuitry.

Nobody but a skilled and knowledgeable designer of medical devices should attempt this. I did a course that covered the basics some 12 years ago as part of my masters in Medical Physics. I know enough to know I don't know remotely enough to design any of this stuff. The concept of high school children even extremely talented ones doing this is a complete none starter.

Finally Any shipping product will need FDA or similar approval before it can be used and this is going to cost big $$$

Don't get me wrong this has great potential, I am sure that organizations like MSF etc. would bite OLPC's arms off for some of these devices (not that OLPC is interested in talking to them though). Without out appropriate approval the devices under consideration would be simply too risky to use from a liability standpoint.

--[[User:Jabuzzard|Jabuzzard]] 13:39, 30 March 2008 (EDT)

: Fear not! For one we're not deploying in the US, so the FDA isn't a problem. And yes, we have a team that includes biomedical engineers, electrical engineers and doctors that will vet this device out at great length before it is ever used in the field. The current discussions are ranging around a 6 month trial before a trial deployment. These pages should be updated soon. Stay tuned for details. [[User:Sethwoodworth|Seth]] 23:29, 22 April 2008 (EDT)

::Or maybe [http://www.phrases.org.uk/meanings/be%20afraid,%20be%20very%20afraid.html "Be afraid, be very afraid."] I would not recommend dismissing the potential regulatory hurdles so quickly. [http://www.fda.gov/cdrh/devadvice/39.html This FDA flowchart] diagramming the regulatory decision-tree on manufacturing and exporting an unapproved medical device looks to be about an order of magnitude more complicated than the entire circuit diagram for your ECG rig. I'm sure you are all very bright guys, but take some friendly advice from someone who works in an FDA-regulated industry and find yourself a regulatory affairs professional '''soon'''. Leave it to them to walk you through the regulatory minefield. Maybe your local IRB can steer you towards the right person. [[User:Cjl|Cjl]] 01:40, 23 April 2008 (EDT)

:::While people should always be careful when hooking other people up to electricity, I feel compelled to point out the flowchart Cjl linked above is approximately an order of magnitude *less* complicated than the ECG diagram, and is substantially easier to solve. Letters may need to be sent, of course, there are only 11 binary questions. It's not that hard.[[User:Niels|Niels]] 16:16, 18 May 2008 (EDT)

:: The ignorance of the subject this comment displays is breathtaking. Just about every country in the world has medical device regulations. This may be nothing more than having to have FDA or equivalent E.U. approval, but some approval will be required. A quick google search shows for Peru, one of the first OLPC countries the following [http://www.ita.doc.gov/td/health/peruregs.html]. You need FDA approval... The idea that because it is third world they allow their citizens to be exposed to any old untested rubbish is patronizing in the extreme.

:: There is also no mention whatsoever on these pages of ISO 13485. Hum, ISO standard on Medical Devices not mentioned anywhere. Tells me that the people doing this while well intentioned clearly have not the faintest clue what they are doing.

:: As I said previously medical devices can kill patients in all sorts of novel and unexpected ways. The reason medical devices are expensive is because designing and then building such devices costs money. Any electrical engineer undergraduate could design a circuit to do an ECG. However a suitable circuit that would pass medical devices approval is at least an order of magnitude more complicated has many more components and could ''perhaps'' be designed by a final year student as a project.

:: You need to stop '''NOW''' and get some proper professional advice on what you are doing before proceeding any further. You clearly don't have a clue at the moment. --[[User:Jabuzzard|Jabuzzard]] 18:57, 28 April 2008 (EDT)

::: Link to Wikipedia's page on [http://en.wikipedia.org/wiki/ISO_13485:2003 ISO 13485], and iso.org's [http://www.iso.org/iso/catalogue_detail?csnumber=36786 order form for ISO 13485 (~$150)]. Last I got into ISO standards (a laser project), what I found was that it's maddening amounts of bureaucratese (truly vast amounts of cut-and-paste text) making it very clear that no legislator, bureaucrat, military type, or diplomat has a clue about the science and they will only accept information that jives with what the basic scientists say, and the basic scientists, once you get to what they say, go back to same units of measure and normal ranges that you probably started off with. You just spent days of research and hundreds of dollars on ISO pubs to get the answer you started with. I wouldn't let that hold up prototyping, and I can't imagine a regulatory affairs type (aka a highly connected ex-bureaucrat) doing pro-bono work without some serious evidence of feasibility. Making prototyping even more important. You could probably get a high school senior with an interest in law to help with the ISO research in advance of the regulatory affairs type, because there's so much cut-and-paste text that it's really hard to get to the wrong answer.[[User:Niels|Niels]] 16:16, 18 May 2008 (EDT)

Niels, please forgive the hyperbole about complexity comparisons, let me also assure you that Seth and I are on excellent terms have developed a relationship of mutual respect (you'll find us chatting together on [http://wiki.laptop.org/go/Communication_channels#irc.freenode.net_channels IRC channel #olpc-health] many nights, where Seth goes by the handle isforinsects). Please do drop by and join in. [[User:Cjl|Cjl]] 17:23, 18 May 2008 (EDT)

Latest revision as of 22:25, 18 May 2008

Pics

Please post pics. Many people would love to see them. --Arjs 01:34, 31 January 2008 (EST)

Supplementary resources

Existing equipment and alternative designs - mostly digging these out from Ian's email/IRC messages.

  • Create USB interface - USB sensor input board. Cool but expensive. PIC.
  • PC USB scope - oscilloscope peripheral that attaches to normal computers via USB. Bloody expensive, Windows-only software... interesting to imitate (and a possible development tool for people with Windows desktops that don't have scopes, until we can make Measure a lovely o-scope interface.)

Hardware development

In regarding to the different projects that have been started would be nice to have a common heading in regarding to the implementation of a circuit that could handle all the peripherals or at least the majority of them..i thing is necessary to have an all in one circuit.

I'm not aware of any conditions under which the telehealth module or any part of it would need to be sterilized. Even in open-heart surgery these instruments are not in the body or even in the sterile field. Washable would be good. Niels 17:35, 18 May 2008 (EDT)

The ECG pads are a high failure rate part, so they will need to be disposable or at least cleanable and replaceable. Many pads in the US use a snap button (like those found in children's clothing), where the wire has the female side and the disposable pad has the male side of the snap bonded to a paper pad with adhesive on the reverse side. Niels 17:35, 18 May 2008 (EDT)

A twelve lead ECG can by reproduced in software from a five lead set-up. Basically you still have leads LA, RA, and LL, but you only use two leads for the precordials and interpolate the rest of the precordials. The precordials (V1 - V6) don't really tell you a heck of a lot besides R wave progression and sometimes ST-segment changes are most noticeable there for reasons of geometry that are dependent on the patient to begin with. Niels 17:35, 18 May 2008 (EDT)

There's a good book on the basic physics of ECG. I'm several hours from my library for the next month, but I think it's Rosen's Cardiac Electrophysiology, A Textbook. Niels 17:35, 18 May 2008 (EDT)

Medical Devices Regulations

Little consideration appears to have been given to complying with medical devices regulations. These are indeed very onerous however this is for good reason. Any electrical device attached to a person has the potential to kill in ways you probably cannot imagine. You need more than just optical isolation between the sensor and any interface circuitry.

Nobody but a skilled and knowledgeable designer of medical devices should attempt this. I did a course that covered the basics some 12 years ago as part of my masters in Medical Physics. I know enough to know I don't know remotely enough to design any of this stuff. The concept of high school children even extremely talented ones doing this is a complete none starter.

Finally Any shipping product will need FDA or similar approval before it can be used and this is going to cost big $$$

Don't get me wrong this has great potential, I am sure that organizations like MSF etc. would bite OLPC's arms off for some of these devices (not that OLPC is interested in talking to them though). Without out appropriate approval the devices under consideration would be simply too risky to use from a liability standpoint.

--Jabuzzard 13:39, 30 March 2008 (EDT)

Fear not! For one we're not deploying in the US, so the FDA isn't a problem. And yes, we have a team that includes biomedical engineers, electrical engineers and doctors that will vet this device out at great length before it is ever used in the field. The current discussions are ranging around a 6 month trial before a trial deployment. These pages should be updated soon. Stay tuned for details. Seth 23:29, 22 April 2008 (EDT)
Or maybe "Be afraid, be very afraid." I would not recommend dismissing the potential regulatory hurdles so quickly. This FDA flowchart diagramming the regulatory decision-tree on manufacturing and exporting an unapproved medical device looks to be about an order of magnitude more complicated than the entire circuit diagram for your ECG rig. I'm sure you are all very bright guys, but take some friendly advice from someone who works in an FDA-regulated industry and find yourself a regulatory affairs professional soon. Leave it to them to walk you through the regulatory minefield. Maybe your local IRB can steer you towards the right person. Cjl 01:40, 23 April 2008 (EDT)
While people should always be careful when hooking other people up to electricity, I feel compelled to point out the flowchart Cjl linked above is approximately an order of magnitude *less* complicated than the ECG diagram, and is substantially easier to solve. Letters may need to be sent, of course, there are only 11 binary questions. It's not that hard.Niels 16:16, 18 May 2008 (EDT)
The ignorance of the subject this comment displays is breathtaking. Just about every country in the world has medical device regulations. This may be nothing more than having to have FDA or equivalent E.U. approval, but some approval will be required. A quick google search shows for Peru, one of the first OLPC countries the following [1]. You need FDA approval... The idea that because it is third world they allow their citizens to be exposed to any old untested rubbish is patronizing in the extreme.
There is also no mention whatsoever on these pages of ISO 13485. Hum, ISO standard on Medical Devices not mentioned anywhere. Tells me that the people doing this while well intentioned clearly have not the faintest clue what they are doing.
As I said previously medical devices can kill patients in all sorts of novel and unexpected ways. The reason medical devices are expensive is because designing and then building such devices costs money. Any electrical engineer undergraduate could design a circuit to do an ECG. However a suitable circuit that would pass medical devices approval is at least an order of magnitude more complicated has many more components and could perhaps be designed by a final year student as a project.
You need to stop NOW and get some proper professional advice on what you are doing before proceeding any further. You clearly don't have a clue at the moment. --Jabuzzard 18:57, 28 April 2008 (EDT)
Link to Wikipedia's page on ISO 13485, and iso.org's order form for ISO 13485 (~$150). Last I got into ISO standards (a laser project), what I found was that it's maddening amounts of bureaucratese (truly vast amounts of cut-and-paste text) making it very clear that no legislator, bureaucrat, military type, or diplomat has a clue about the science and they will only accept information that jives with what the basic scientists say, and the basic scientists, once you get to what they say, go back to same units of measure and normal ranges that you probably started off with. You just spent days of research and hundreds of dollars on ISO pubs to get the answer you started with. I wouldn't let that hold up prototyping, and I can't imagine a regulatory affairs type (aka a highly connected ex-bureaucrat) doing pro-bono work without some serious evidence of feasibility. Making prototyping even more important. You could probably get a high school senior with an interest in law to help with the ISO research in advance of the regulatory affairs type, because there's so much cut-and-paste text that it's really hard to get to the wrong answer.Niels 16:16, 18 May 2008 (EDT)

Niels, please forgive the hyperbole about complexity comparisons, let me also assure you that Seth and I are on excellent terms have developed a relationship of mutual respect (you'll find us chatting together on IRC channel #olpc-health many nights, where Seth goes by the handle isforinsects). Please do drop by and join in. Cjl 17:23, 18 May 2008 (EDT)