As Matt says, figure 2 is where we show that lower, legacy resolutions would also benefit from surface-based analysis. This was a late addition to the paper, so there isn't as much coverage of lower resolution.
Figure 7 only looks at all of cortex as a single entity, so it doesn't capture problems with crossing sulcal banks, or areas not being lined up. The main point of it is to show that high-resolution data does much better with surface-based methods than it does with volume-based methods, since tissue alignment across subjects is only a problem for volume-based methods. Additionally, the distribution of cortical signal capture across the brain is more uniform with surface-based methods, while volume-based methods do much worse in some regions than in others. Figure 8 does not represent different resolutions, it uses 2mm resolution (HCP fMRI), the "+ 4mm" refers to adding smoothing to the analysis (surface-based smoothing for MSMAll, volume-based for FNIRT). Smoothing is a traditional part of many processing pipelines, and we consider such a practice to be harmful, especially when the smoothing is volume-based. Tim On Fri, Jun 1, 2018 at 8:06 PM, Joseph Orr <[email protected]> wrote: > This is a great paper for encouraging the adoption of HCP-style > acquisitions for future studies, but I'm curious about the benefit of > applying the surface-based approach to existing, traditionally acquired > datasets. In Figure 7 of your paper you simulate traditional acquisition > resolutions which is very helpful. But these simulations are still assuming > HCP-style sub-millimeter T1s and T2s for the surface registration. In a > recent dataset I collected .8mm T1s but we didn't collect T2s (not my > decision). > > What are your recommendations for more traditional study acquisitions like > this? I think this question is what you're getting at in Figure 8, but I > need to spend a little more time with the paper. Without a T2 is FreeSurfer > my best bet? > > Thanks, > Joe > > -- > Joseph M. Orr, Ph.D. > Assistant Professor > Department of Psychological and Brain Sciences > Texas A&M Institute for Neuroscience > Texas A&M University > College Station, TX > > On Fri, Jun 1, 2018 at 7:12 PM, Glasser, Matthew <[email protected]> > wrote: > >> I would note that we don’t recommend mapping group average volume data to >> group average surfaces. It is a much better idea to map the data onto the >> individuals' surfaces and register it on the surface (and also avoid >> smoothing in the volume or even better avoid smoothing entirely): >> >> See this article currently in press at PNAS: https://www.biorxiv.org/ >> content/early/2018/04/23/255620 >> <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.biorxiv.org_content_early_2018_04_23_255620&d=DwMF-g&c=ODFT-G5SujMiGrKuoJJjVg&r=ZKy1VO33u0kvO-PqY1gpb9Ld-AGhtT8c9PAcpsEyp70&m=NDSQ3kMMhk_BcjZ91RPExirFBMmqlMSPFqIRXr0e6Mk&s=Op6bJ6Q2t5EvGlDljAWbc9tKe1pHdybU1cXQUKBjEsA&e=> >> >> >> Peace, >> >> Matt. >> >> From: <[email protected]> on behalf of Joseph Orr < >> [email protected]> >> Date: Friday, June 1, 2018 at 6:37 PM >> To: Kristian M Eschenburg <[email protected]> >> Cc: HCP Users <[email protected]> >> Subject: Re: [HCP-Users] Conte69 164k and 32k mesh >> >> I'd recommend using the S900 group maps available at the BALSA site which >> is referred to on the caret page: https://balsa.wustl.edu/ >> <https://urldefense.proofpoint.com/v2/url?u=https-3A__balsa.wustl.edu_&d=DwMF-g&c=ODFT-G5SujMiGrKuoJJjVg&r=ZKy1VO33u0kvO-PqY1gpb9Ld-AGhtT8c9PAcpsEyp70&m=NDSQ3kMMhk_BcjZ91RPExirFBMmqlMSPFqIRXr0e6Mk&s=nertDHfdcmNZRWyvrQ2xkwan957VrE9FyosIzhKP5Mw&e=> >> >> You can register volumetric data to the surfaces following the >> instructions in the Workbench tutorial or this blog post: >> http://mvpa.blogspot.com/2014/03/connectome-workbench-adding-volumes.html >> <https://urldefense.proofpoint.com/v2/url?u=http-3A__mvpa.blogspot.com_2014_03_connectome-2Dworkbench-2Dadding-2Dvolumes.html&d=DwMF-g&c=ODFT-G5SujMiGrKuoJJjVg&r=ZKy1VO33u0kvO-PqY1gpb9Ld-AGhtT8c9PAcpsEyp70&m=NDSQ3kMMhk_BcjZ91RPExirFBMmqlMSPFqIRXr0e6Mk&s=-3VEDq5QbXj8H94XYv4qckbANmT62lOByFhcIoukpWo&e=> >> >> I've found workbench to be more user friendly than caret, assuming that's >> what you were intending to use. >> >> Joe >> >> -- >> Joseph M. Orr, Ph.D. >> Assistant Professor >> Department of Psychological and Brain Sciences >> Texas A&M Institute for Neuroscience >> Texas A&M University >> College Station, TX >> >> On Fri, Jun 1, 2018 at 5:48 PM, Kristian M Eschenburg <[email protected]> >> wrote: >> >>> Hi HCP Admins >>> >>> Where can I download the 164k and 32k resolution Conte69 template >>> meshes? I'm trying to register data I've acquired to the same MNI >>> space-surfaces as the HCP data. I searched the SumsDB site ( >>> http://brainvis.wustl.edu/wiki/index.php//Caret:Atlases/Conte69_Atlas >>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__brainvis.wustl.edu_wiki_index.php__Caret-3AAtlases_Conte69-5FAtlas&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=ZKy1VO33u0kvO-PqY1gpb9Ld-AGhtT8c9PAcpsEyp70&m=VgTjzj6GtNY5OkkbA0JzFKw5Q0JILeOmYmRx6TQulFY&s=F3ndSFg5e0RefqpYKJ5c2JOgCVlaCJP6wVu0gQQHqXc&e=>) >>> and it seems there was a "catastrophic failure" w.r.t to making this data >>> freely available. >>> >>> Any information is appreciated. >>> >>> >>> >>> Kristian >>> >>> _______________________________________________ >>> HCP-Users mailing list >>> [email protected] >>> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.humanconnectome.org_mailman_listinfo_hcp-2Dusers&d=DwMF-g&c=ODFT-G5SujMiGrKuoJJjVg&r=ZKy1VO33u0kvO-PqY1gpb9Ld-AGhtT8c9PAcpsEyp70&m=NDSQ3kMMhk_BcjZ91RPExirFBMmqlMSPFqIRXr0e6Mk&s=wCVk1igEI_a0pOwWL3FwdqRffyD5_iX1hkVHWMC_owE&e=> >>> >> >> _______________________________________________ >> HCP-Users mailing list >> [email protected] >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >> <https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.humanconnectome.org_mailman_listinfo_hcp-2Dusers&d=DwMF-g&c=ODFT-G5SujMiGrKuoJJjVg&r=ZKy1VO33u0kvO-PqY1gpb9Ld-AGhtT8c9PAcpsEyp70&m=NDSQ3kMMhk_BcjZ91RPExirFBMmqlMSPFqIRXr0e6Mk&s=wCVk1igEI_a0pOwWL3FwdqRffyD5_iX1hkVHWMC_owE&e=> >> > > _______________________________________________ > 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
