As a signal processing engineer, I am relatively new to this kind of work.
I am not familiar with the idea of the U.S. FDA approving software.
However, generally in the field of electrical engineering, algorithms and
software are generally worked out and approved by means of standards
committees. When someone uses a G.728 speech coder, for example, everyone
knows what that means. I don't know what the standards body is for fMRI
signal processing. Would it be the IEEE? (In telecom the CCITT, but
that would not apply to fMRI. However, in telemedicine, you could see
the CCITT wanting to have standards). Usually the standards committes,
rather than government, issue the algorithms and software. Sometimes
I see equations in various papers disagreeing with each other, and
standardization of algorithms and software can be helpful. Of course,
standards efforts cost money in terms of employees' hours. There is a lot
standard. Corporations see the benefit of allocating employees' time for
this, but I don't know about research grants, since I am an engineer and
not a professor.
Linda Seltzer
> Dear Tae,
>
>
> I can recommend Analyze for what it does (I've mainly used the ROI tool
> and rendering, as well as some image algebra, morphological operations,
> played with SISCOM once, etc.) - it does that very well. However, as you
> probably know, it is so expensive that that is a prohibitive factor for
> many centres.
>
> Also, while I've successfully used it with Nifti, others have had
> problems.
>
> For image registration and tissue class segmentation, I use SPM5 or
> Exbrain and have never compared them to what Analyze does.
>
> You raise a very interesting point, namely approval. I wasn't aware of
> ANY neuroimaging software being FDA approved and have some difficulty
> with the concept. Presumably some kind of validation is necessary - but
> how would you do that, with innumerable actual or potential
> applications? Without standard operating procedures in addition, I am
> sure any user could come up with a protocol that does not do what is
> expected, and the documentation needs seem prohibitive to me. I can see
> that e.g. stethoscopes are FDA-approved - but if you haven't been
> trained, you could not realise yours is broken, or come to completely
> wrong conclusions, even with the best stethoscope in the world.
>
> So, the question is, _is_ there _any_ FDA-(or other agency)-approved
> neuroimaging software, and if so, where can one get hold of such a list
> and the criteria for approval? And if there is such a list, in what way
> would it be useful to you?
>
> (I have found one document relating to CURRY
> (http://www.fda.gov/cdrh/pdf/k001781.pdf) where the argumentation (if I
> understand correctly) is essentially that it intends to do similar
> things to something marketed before 1976 when some Act came into force
> and therefore it's ok to market - not what any scientist would regard as
> validation I'm sure!). If that's the standard, I think you risk paying
> for a marketing strategy which will not, per se, give you any additional
> advantage over open source tools.
>
> I'm very curious and looking forward to hearing more,
>
>
> Alexander
>
>
> -----Original Message-----
> From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
> On Behalf Of Woo-Suk, Tae
> Sent: 06 December 2007 07:41
> To: [log in to unmask]
> Subject: [SPM] FDA approved neuroimaging software
>
> Dear SPM community lists.
>
> Who can recommend FDA approved neuroimaging software such as Analyze
> (BIR, Mayo clinic) to me?
> The functions of image registration, segmentation, image algebra,
> multi-objects 3D visualization are needed.
>
> Woo Suk, Tae Ph.D.
> Seoul, Korea
>
|