[IGSTK-Users] Same TRE for varying target points

Andinet Enquobahrie andinet.enqu at kitware.com
Mon Mar 24 09:27:43 EDT 2008


Hi Nathaniel,

The Landmark3DRegistrationErrorEstimator in IGSTK is implementation of the
TRE derivation by West et al.

J.B West et al, "Fiducial Point Placement and the Accuracy of Point-based,
Rigid Body Registration", Neurosurgery, pp 810-816, Vol.48, No.4, April
2001.

You mentioned in your email a Matlab script implementation. Is this script a
full implementation of the TRE equation in Matlab or is it a script that
calls igstk::landmark3DRegistrationErrorEstimator to compute TRE?

If your Matlab implementation is independent of IGSTK and if both
implementations produce the same result, then we will need to look into the
derivation to understand your observation.

Let us know.

-Andinet






> I'm conducting a study into fiducial configurations and their related TRE.
> I used the algorithm presented in
> igstkLandmark3DRegistrationErrorEstimator to develop a MATLAB script to
> calculate the TRE for given image landmarks, FRE, and a target point.
> However, when I vary the target point by its x-coodinate I receive the same
> TRE for a given fiducial configuration. This is not experienced when I vary
> the y- or z-coodinates of the target point. I validated the MATLAB script
> with Landmark3DRegistrationErrorEstimatorExample::ErrorEstimator1 in
> IGSTK-Sandbox v2.0 and received the same results.
>
> Has anyone else experienced this situation or can provide enlightenment
> for a solution?
>
> Thanks for your time and help,
> Nathaniel Hamming
>
> FYI: I included the image landmarks, track landmarks, and target points
> that were used to validate the script using
> Landmark3DRegistrationErrorEstimatorExample::ErrorEstimator1 along with the
> resulting TRE.
>
> *Image Landmarks:*
> (0.114845139504747         -84.6604824506397         -6.26710676551675)
> (72.9584125146068          6.53918641170636          32.1024344060193)
> (-80.5111003508657          4.75960126764604          40.0533742041848)
> (58.1672017212154         -47.5649370736371         -12.6361990926914)
> (-54.487081424932         -55.2545031775054         -26.0381247399557)
> (75.5947613145267          29.6918711527396          11.7797978481002)
> (-84.5277722958269          26.1312335752415          12.7930502907778)
> (60.4385919682889         -13.0993934111475         -31.2299212457651)
> (-65.2827107691361         -21.1801171028438         -34.2503757501768)
> (2.04690415405402         -41.7124883445493         -64.6657459246422)
> (-73.3214245559623         -19.8518398326938           -3.984626698198)
> (69.9212556595938         -1.46702964398868        -0.472367941194386)
> (-25.6974808869996          -75.065947078444         -33.4563926906414)
> (33.8170959437249         -63.0341849792339         -37.9607564776401)
>
> *Tracker Landmarks:*
> (46.600717682397         -30.8121035263332         -885.990194627345)
> (128.792976006704          44.5188525815836          -935.56621842578)
> (127.021257370086         -107.728230217206         -949.701561980191)
> (82.9686506055305          27.1399645537516         -882.972750919913)
> (79.6707385764108         -85.7007660779085         -874.429003896307)
> (154.517592060753          46.6545433797318         -919.449505921427)
> (152.644949852231         -112.921598475625         -926.667845321769)
> (120.254718830609          28.8688629954591         -870.901038232264)
> (114.048783584512         -96.1666948609119         -872.597788374716)
> (98.5072568480499         -30.8629325840406         -835.649463957748)
> (110.292486691374         -102.893696403642         -902.229885054371)
> (126.40396743689          39.8625678957445         -902.309420522653)
> (60.4020948254781         -57.3617285910359         -862.234734609912)
> (72.5897769707638          1.66170144246608         -856.570128840823)
>
> *Target Points:*
> 1 - (-83.5623069243003         -4.91518218770797          32.0031623919239
> )
> 2 - (-6.28957965157304         -4.91518218770797          32.0031623919239
> )
> 3 - (70.9831476211542         -4.91518218770797          32.0031623919239)
> 4 - (-6.28957965157304         -104.915182187708          32.0031623919239
> )
> 5 - (-6.28957965157304           15.084817812292          32.0031623919239
> )
> 6 - (-6.28957965157304           95.084817812292          32.0031623919239
> )
> 7 - (-6.28957965157304         -4.91518218770797         -76.1786557898943
> )
> 8 - (-6.28957965157304         -4.91518218770797          1.09407148283303
> )
> 9 - (-6.28957965157304         -4.91518218770797          78.3667987555603
> )
>
> *TRE^2 (results from
> Landmark3DRegistrationErrorEstimatorExample::ErrorEstimator1) :*
>  1 -   0.054657382367051761
> 2 -   0.054657382367051761
> 3 -   0.054657382367051761
> 4 -   0.12959851653502544
> 5 -   0.070150472712894457
> 6 -   0.23372722469439022
> 7 -   0.085080085180609902
> 8 -   0.033015214490287000
> 9 -   0.13262218720315677
>
>
>
> _______________________________________________
> IGSTK-Users mailing list
> IGSTK-Users at public.kitware.com
> http://public.kitware.com/cgi-bin/mailman/listinfo/igstk-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/igstk-users/attachments/20080324/7198b93b/attachment-0003.html>


More information about the IGSTK-Users mailing list