[Insight-users] Re: Patented algorithms in ITK?

[Tobias Heimann]

Hi Luis,

Thanks for your clear reply to this question! Maybe the plans of getting 
rid of patented algorithms in ITK should be included in the Wiki as 
well, since the current text seems a bit contradictory and confusing (at 
least for me).
Anyway, in our case sharing an algorithm in ITK weights more than filing 
a patent for it, so I would be happy to contribute my code. Could you 
tell me which one of the developers I should contact for further 
technical/design discussions?

Thanks and best regards,
Tobias


Luis Ibanez wrote:

>
> Hi Tobias,
>
> The current status of patented algorithms in ITK is that
> those algorithms that are known to be patented have been
> isolated in a directories marked as "Patented".
>
> These directories only get build if you turn on the CMake
> option "ITK_USE_PATENTED".
>
> As soon as you do that, you are infringing the rights of
> the patent holder, unless you have negotiated a license
> with the holder.
>
> Patented algorithms are undesirable in an Open Source
> toolkit like ITK. We should do our best for getting rid
> of the current Patented algorithm and to make sure that
> the toolkit doesn't continued to get infected with such
> type of code.
>
> Algorithm developers whether they are in the Academic or
> Commercial setting will have to make their minds as to
> whether they want to Patent and charge for rights of use,
> or they want to share for free with the imaging community.
>
> Those who play the double-mind game of pretending to be nice
> by sharing, but at the same time patenting their algorithms
> in order to get profits from intellectual property should
> have to re-examine their motivations. It is my personal view
> that if an institution patent an algorithm and then it put
> it in an open source library, they should be charged for
> "advertising" since they are using the popularity of Open
> Source as a mechanism for popularizing their algorithms so
> they can charge for them later on.
>
> The sad reality of Algorithm Patents is that nobody gets
> rich with them. They only persuade people of not using
> the technology. The intellectual property system is simply
> getting to ridiculous extremes of patenting the obvious.
> When academic institutions pursue patents for algorithms
> they do it mostly with the intention of adding that to
> their annual activity report, that should sound something
> like:
>
>    "This year our group
>     a) published 45 papers in peer-review journals
>     b) published 67 papers in peer-review conferences
>     c) get awarded 12 patents
>     d) applied for 27 patents
>    "
>
>
> This is how delusional the current system for evaluating
> academic productivity has become. Of course, the reason
> is that this is the kind of annual report that a manager
> could be able to read without having any idea of what the
> group is actually doing. This report could equally well be
> produced by a group doing stem cell research that by a group
> doing compiler optimization.
>
> Whether any of those papers or patents provide any significant
> contribution to society or to scientific advancement is not even
> considered to be a relevant question. That is supposedly covered
> by the infallible ethereal properties of the "peer-review" system.
>
>
> ....
>
>
> VTK has recently succeeded to get rid of all its Patented classes.
> It will be great if we manage to bring ITK to the same state.
>
>
>
>   Regards,
>
>
>
>      Luis
>
>
>
> -------------------
> Tobias Heimann wrote:
>
>> Hi Luis,
>>
>> What's the current status regarding including patented algorithms in 
>> ITK?
>> The Wiki says it's not possible because patents and the ITK license 
>> don't fit together, yet there is a growing amount of algorithms in 
>> the Patented directory...
>> The matter is that I would like to integrate my 3D Active Shape Model 
>> construction into ITK, but in our department we are still thinking 
>> about patenting it for commercial applications. Is that a problem for 
>> the ITK integration?
>>
>> Best wishes,
>> Tobias
>>
>

-- 
Dipl.-Inform. Med. Tobias Heimann
Deutsches Krebsforschungszentrum         (German Cancer Research Center)
Div. Medical and Biological Informatics B010      Tel: (+49) 6221-423548
Im Neuenheimer Feld 280                           Fax: (+49) 6221-422345
D-69120 Heidelberg                              email: T.Heimann at dkfz.de
Germany                       http://www.dkfz.de/mbi/people/tobiash.html