Raymond Alley
Developing technology for modifying the index of refraction In vivo inside an Intraocular Lens (IOL), modifying vision
February 2017
< 1 min reading time
Marked as spam
|
Meet your next client here. Join our medical devices group community.
Private answer
Ee Bin Liew
Define "medical grade computer".. and for that matter define a "medical grade mobile phone"..
Marked as spam
|
|
Private answer
Aaron Liang
By medical grade do you mean being able to withstand cleaning and disinfection without damaging the internals, include antimicrobial housing or meeting the electrical safety requirements for medical electric equipment, IEC 60601? It all depends on the product, how it is intended to be used and the risks posed by its intended use. That should drive your design and hardware selection.
Marked as spam
|
|
Private answer
Raymond Alley
I wanted to thank all for your input. I had a suspension some manufactures might be throwing phrases around to freely.
Marked as spam
|
|
Private answer
Caitlin Morse, PMP
Electrical safety (60601), water ingress(IP rating), etc are possibilities. They may also be referring to HIPPA protections of data or being able to prove system performance with a standard operating system. Phone apps seem to work fine for biofeedback devices, but where the software is actually controlling the device, our clients have often opted for a dedicated tablet and included it as part of the medical system to simplify the validation.
Marked as spam
|
|
Private answer
I think you are more likely referring to what we call an "industrial PC". For a medical device, if you are using a Windows based environment or similar, the medical standards will require you to be able to lock down that device to prevent tampering or use for anything other than the operation of said medical device. The benefit of an industrial PC is you can guarantee a continuity of supply for several years rather than just a short period like a commercial PC. This stops you have to re-design and re-certify your system every year or so because the PC you've chosen has gone out of manufacturing.
Marked as spam
|
|
Private answer
Are you talking about the computer you connect it to in order to control the device and manage its data or the processor inside the device? If you're talking about the external computer I'd say technically no but there will always be a cost vs future proofing/technical support/software quality trade off. If you are talking about the processor inside the device, I'd say the cost of a medical grade or high end processor pales to insignificance compared to the cost of a medical grade embedded OS and software development/maintenance life cycle. You just wouldn't take the risk on a lower end device in case it disappears from the market (too) soon after you release your product.
Marked as spam
|
|
Private answer
Raymond Alley
Again, thank you for your comments. The computer is inside the instrument and used for controlling precision components. A reliable computer is important for a long-lasting device. Some of the medical grade computers I have seen add a good premium for that claim. That said a few have value added features that look promising. Some are completely enclosed, no vents to clean which is a plus long-term.
Marked as spam
|
|
Private answer
Phill Brown
Hmm, interesting. If it's for precision control, why not use an industrial PLC or a microcontroller? Texas instruments is actively promoting MCU chips for medical applications. If you're doing control and data collection it's not a bad idea to have slave MCUs to do the precision work and then output the data to the PC for user interface stuff.
Marked as spam
|
|
Private answer
if it is a PC inside of an instrument then I would suggest an industrial PC is the way to go unless you are considering developing some custom electronics with embedded software/firmware that gets you away from needing to use a Windows based environment or similar. You'll pay a higher price but will reap the rewards in the longevity of supply, quality, and not having to re-certify the device as often if your pc changes.
Marked as spam
|
|
Private answer
If the computer is in a patient care area and can come in contact with the patient per the following guidance: http://www.mdsr.ecri.org/summary/detail.aspx?doc_id=8286
Marked as spam
|
|
Private answer
To be used in the OR the device needs to go through a certification process to be deemed medical grade. Some devices are not much different than their non-medical grade equivalent. It is an added cost for developing a medical product. For liability reasons it is highly recommended.
Marked as spam
|
|
Private answer
Raymond Alley
Charles, this is an out-patient medical device that will be employed in an office.
Marked as spam
|
|
Private answer
Raymond Alley
Thank you for your continued comments on this subject matter. It’s always an honor to be part of a team Developing medical devices that can improve people’s lives and our system will help millions with improved vision. Safety is always a Paramount concern to ensure all will have a positive outcome. Yes, following the QMS (ISO 13485) is an arduous process, however as an Engineer if this was easy it wouldn’t be challenging and exciting.
Marked as spam
|
|
Private answer
Raymond Alley
Hi Phill, That’s simple the evolution of development. Using a computer is much easier in development than implementing an embedded system. I see embedded control in future generations, which reduces the footprint, costs, and improve reliability. You have to walk before you run.
Marked as spam
|
|
Private answer
Hi Ray, This is such an expansive question. There are three key points I make when talking to medical device manufacturers. 1) Separate the medical device from the interface device to the fullest extent possible. This allows for independent management of the disparate life cycles. This enables the medical device owner to manage the network interface better and the manufacturers to respond faster to emerging computer vulnerabilities. 2) Design for worst case environment. Even outpatient care locations have inherent risk of bodily fluid contamination. 3) Think Cyber security first, then clinical function. Enable the customer to have maximum control like: turning off all ports not needed to provide clinical function, signed software that prevents unauthorized code changes, store data off the device by default or on a drive separate from the computing platform . Remember, your medical device will be kept in use much longer than the computing platform support life it was built on.
Marked as spam
|
|
Private answer
In my opinion there is a lot more that could be done.
From the basics of better house keeping to the choices(and continual choices) of kernel-level software. I don't want to get into the semantics about how easy it is to reset/recover windows passwords. It's equally easy to navigate in native windows os or live linux distro. There's very little encryption being done on a software or hardware level. Additionally, too much data is being stored at the end user's computer. Picking window's isn't an issue, but at least maximize the full potential of everything microsoft has to offer. Eliminate the need for multiple PC's that have RAID 1,01,10 setups,Instead use shares with a fiber connection,include on/offsite backups into a SLA. Make recoveries/backups easier with the use of modern VM technology. Same principle applies to installations. You could even take it as far as to facilitate AD into the mix. Manage credentials at a server level. Marked as spam
|
|
Private answer
It would help to know what regulatory compliance standards this medical system needs to meet. From an EMC perspective, OTS ITE computers often times do not fare well under immunity tests like ESD and EFT Burst since they are verified to ITE standards which are less stringent than medical. Also, once you attach the computer to other medical devices, the system performance to EMI including emissions may degrade.
Marked as spam
|
|
Private answer
Yes for sure If it has an incidence on the measurement or put into action a medical device, you need to make a risk assessment and to assess the computer. Also make sure data are registered and think about the upgrade of the version and the way to keep all data. All the best
Marked as spam
|
|
Private answer
Leonard Eisner
There are several standards and concerns to look at. 1) IEC 60601-1 requires medical devices in clinics, out patient centers, hospitals, homes, etc. to be isolated enough to keep leakage currents down low enough so this includes the computer in your product so review Annex I of IEC 60601-1, edition 3.0 or 3.1 (3.1 is 3rd ed. + Amendment 1) to determine the best configuration for your design. Also, as mentioned you need to isolate all your I/O to meet the requirements of the standard. 2) EMC per IEC 60601-1-2 (3rd or 4th ed - 4th ed required by FDA, Health Canada & EU MDD for your type of device Dec 31, 2018 end of transition date) is getting much tougher for the 4th ed. and even with the 3rd ed. requirements non medical computers may not meet all the requirements for EMC. See my post on 4th ed. of IEC 60601-1-2 (transition dates have changed since that post - note above dates are correct) at . See my next reply to continue this discussion.
Marked as spam
|
|
Private answer
Raymond Alley
Were in the process of determining which stand is appropriate for this device and of course this is not straight forward. The other difficult part is the interruption of FDA guidelines. We have engaged Regulatory consultants and obtain varying answers to what standard this falls under. However, the consensus is to apply for the easiest one.
Windows has its issues but it also has it benefits. Imaging and Image Processing is quickly addressed in Windows along with process flow and control. Of course, encryption and cyber security is always a concern and is on the radar. We have engaged cyber security experts looking forward. The main concept to keep in mind is continuous improvement. The goal is developing a safe outpatient medical device to address a need in the Intra Ocular Lens domain. Than continually improve on the system as it matures. Marked as spam
|
|
Private answer
Julie Omohundro
Yuhan Zhu, why would medical device software be inherently simpler than "pure software"? I'm not a software expert, but I would have thought that some medical device software is more complex than some pure software, and vice versa.
Marked as spam
|