Hello Gordon, thank you very much for your answers. Due to our resolution limited to 2 mm isotropic we have a substantial lower TR (3270 vs. HCP 5520 ms) and a TE of 110.4 ms (HCP 89.50 ms). I wasn't sure if it is the one gradient direction that needs to be changed (since it always stopped at the same point) or something else (in this case using a longer TR/TE). Thanks to your answer, I will now gradually increase TR in a trial-and-error-fashion.
Any chance, the mentioned 1.5mm-DTI-TRIO-sequence can be posted here? (I guess, there are a couple more TRIO users in the mailing list) Best regards, Martin ________________________________________ Von: Xu, Junqian [[email protected]] Gesendet: Samstag, 6. Juli 2013 16:57 An: Reiß-Zimmermann, Martin Cc: [email protected] Betreff: Re: [HCP-Users] HCP-dMRI gradient table on a TRIO - any experience? Martin, Sorry that I misunderstood your problem. However It is still gradient duty cycle related. The Siemens VB software version doesn't estimate gradient heating as accurately/conservatively as the Siemens VD software version. Hence the minimum TE/TR allowed by the sequence can still trip the gradient duty cycle limit in real time, causing scan crash. The solution is to relax the TE/TR little by little with try-and-error until you find a stable combination that passes your test. Gordon On Jul 5, 2013, at 9:12 PM, "Xu, Junqian" <[email protected]> wrote: > Martin, > > The culprit is the gradient power amplifier (GPA) duty cycle too high on the > Y axis (i.e. the consecutive high value of 0.992 and -0.975). We have > encountered similar problem in HCP and have implemented a solution to solve > the gradient duty cycle issue. > > You can find the solution to this problem in Sotiropoulos et al. NeuroImage > special issue > > http://www.ncbi.nlm.nih.gov/pubmed/23702418 > > Section: Q-space sampling > > "However, even when realistic sequences of diffusion gradient orientations > were used (even with interleaved shells), the gradient duty cycle, due to > accumulated heating, could still limit the achievable minimal TR. The order > of the diffusion gradient orientations in the final HCP dMRI protocol was > therefore optimised by swapping adjacent orientations to reduce the > auto-correlation (lag = 1) of absolute gradient amplitude (~ < 0.5) for each > physical gradient axis." > > Please let me know if this is clear enough for you to solve your specific > case. > > Regards > Gordon > > On Jul 5, 2013, at 11:06 AM, "Reiß-Zimmermann, Martin" > <[email protected]> wrote: > >> Dear all, >> >> I tried implementing and converting the HCP dMRI-sequences for our TRIO. I >> followed the suggestions in the Appendix I and with minor modifications (2 >> mm isotropic) it works well with MB3 and the MDDW-gradient table (60 >> directions). >> >> However, when I use the HCP gradient table the measurement stops at >> reproducable stops within the sequence (see following) and always with the >> same error (hope I can translate that somehow from German into English): >> Error in gradient amplifier, GPA not ready. duty cycle is to high. Please >> check sequence parameters. >> >> The sequences always stop at these points: >> >> 95dir-dMRI: direction number 47 (-1.000 , -0.001 , -0.011) >> 96dir-dMRI: direction number 75 (-0.094 , 0.992 , -0.082) >> 97dir-dMRI: direction number 55 (-0.134 , -0.975 , 0.180) >> >> The images before look great! >> >> I took a look at the page of Emmanuel Caruyer - but if I create a gradient >> table of 90 dir (the HCP 96dir-table seems to have 6 b0 and the rest >> b1000-3000, the E.C.-table seems to encode b>0 values, right?) it is totally >> different from the HCP one. >> >> I also read the note in the HCP gradient table that these gradients are >> customized for the Connectome-Scanner. >> >> So my question is: Shall I just use one created on Emmanuel's page or can I >> (or better: someone else) change the one from HCP? >> >> The Appendix I also mentions that the dMRI-sequence has been successfully >> tested on a TRIO and Verio with 1.5 mm isotropic resolution (and b2000). Is >> it possible that you could post the protocol (including the gradient table). >> >> Thank you very much. Best regards, >> Martin >> ____________________________________________________ >> Dr. med. Martin Reiss-Zimmermann >> Abteilung für Neuroradiologie >> Universitätsklinikum Leipzig AöR >> Liebigstr. 20 >> 04103 Leipzig >> >> Mail: [email protected] >> Fon: +49 - 341 - 97 17 585 >> Fax: +49 - 341 - 97 17 490 >> Net: http://www.neuroradiologie-leipzig.de >> ........................................................................................................................................................................................................ >> Universitätsklinikum Leipzig - Anstalt öffentlichen Rechts | >> Aufsichtsratsvorsitzender: Prof. Dr. Wolfram Knapp >> Medizinischer Vorstand und Sprecher: Prof. Dr. med. Wolfgang E. Fleig | >> Kaufmännischer Vorstand: Dipl.-Kfm. Ekkehard Zimmer >> Sitz: Liebigstraße 18, 04103 Leipzig | Telefon +49-341-971709 | St.Nr.: >> 23114904074 | IK: 261401052 >> Bankverbindung: Deutsche Bank | BLZ: 86070000 | Kto.: 1247998 | Swift-Code: >> DEUTDE8L | IBAN: DE27860700000124799800 >> Internet: www.uniklinikum-leipzig.de >> >> Universität Leipzig | Ritterstr. 26, 04109 Leipzig | Körperschaft des >> Öffentlichen Rechts | Vertreten durch die Rektorin Prof. Dr. Beate Schücking >> Zuständige Aufsichtsbehörde: SMWK, Wigardstr. 17, 01097 Dresden, www.smwk.de >> Umsatzsteuer-Identifikationsnummer gemäß §27a Umsatzsteuergesetz DE 141510383 >> >> Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte >> Informationen. Sollten Sie nicht der richtige Empfänger sein oder Sie haben >> diese E-Mail irrtümlich erhalten, informieren Sie uns bitte sofort und >> vernichten diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte >> Weitergabe dieser E-Mail und der darin enthaltenen Informationen ist nicht >> gestattet. >> >> _______________________________________________ >> HCP-Users mailing list >> [email protected] >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > _______________________________________________ > HCP-Users mailing list > [email protected] > http://lists.humanconnectome.org/mailman/listinfo/hcp-users _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
