Linking Contemporary High Resolution Magnetic Resonance Imaging to the Von Economo

Transcription

1 Supplementary Materials of Title Linking Contemporary High Resolution Magnetic Resonance Imaging to the Von Economo legacy: A study on the comparison of MRI cortical thickness and histological measurements of cortical structure As published in Human Brain Mapping Epub ahead of print. Link to full paper: Authors Lianne H. Scholtens, Marcel A. de Reus, Martijn P. van den Heuvel Affiliations Brain Center Rudolf Magnus, Department of Psychiatry, University Medical Center Utrecht Address for Correspondence Martijn van den Heuvel, Brain Center Rudolf Magnus, Department of Psychiatry, University Medical Center Utrecht, Heidelberglaan 100, 3508 GA Utrecht, PO Box 85500, Room: A01.126, The Netherlands; Phone: ; Fax: ; M.P.vandenheuvel@umcutrecht.nl Supplementary Materials Von Economo MRI analysis for non-overlapping regions!1

2 To match regions of the Von Economo and Desikan-Killiany atlas, cortical values were averaged in the case of multiple smaller Von Economo areas linked to one single Desikan-Killiany region (see methods). To rule out any potential effect of this averaging, a second post-hoc analysis was performed in which only regions were included that involved a 1 to 1 mapping between the two atlases (23 regions, Supplemental Table 1, see also methods), revealing highly consistent findings with the main analysis (p=7.65x10-04, r=0.65). Consistency of Von Economo MRI cortical thickness measurements in a year subset. Several studies have reported a modest, but relevant association between age and reduction in cortical thickness in the age range between 20 and 40 years (for example (Lemaitre, et al., 2012) (Schnack, et al., 2014)). To verify whether our results are influenced by the difference in age range between our 215 included Human Connectome Project (HCP) subjects (aged years, in age classes of 5 years) and the Von Economo-Koskinas dataset (30-40 years) we performed two additional analyses. First, we compared the Von Economo-Koskinas cortical width measurements to only the subset of MRI thickness values derived from the oldest included HCP subjects (in age class 31-35). Similar to our main findings (see main text), this resulted in a strong positive correlation between MRI cortical thickness and Von Economo-Koskinas thickness of the cortical mantle (HCP years, n=96: r=0.64, p=1.15x10-07, data shown in Supplemental Figure S3-A). Second, we examined the cross-technique correlation for a dataset of n=40 healthy participants within the exact age range as examined by Von Economo and Koskinas (all subjects ranging between years, mean/std: 34.1/3.4 years) taken from recent studies of our group (all datasets acquired on the same 3 Tesla Philips Achieva scanner, all with the same T1-weighted imaging protocol, 3D FFE using parallel imaging, TR/TE 10ms/4.6ms, 200 slices, 0.75 mm voxel size). Analyzing this new dataset in the same way as the HCP data, again revealed a significant, positive correlation between MRI estimates of cortical thickness and cortical width values of the Von Economo atlas (Utrecht set!2

3 30-40 years: r=0.56, p=7.97x10-06, data shown in Supplemental Figure S3-B). Supplemental Figure S1. In the main part of our study, a mapping of the Von Economo data to the Desikan-Killiany subdivided atlas of 114 (57 single hemisphere) regions was used. Using other (also commonly used) more fine-grained subdivisions of the Desikan-Killiany atlas (as made by Hagmann and colleagues (Cammoun, et al., 2011)(Hagmann, et al., 2008)) revealed highly consistent results. Panel A, B and C show correlations between the Von Economo and Koskinas estimates of width of the cortical mantle and MRI derived cortical thickness when the 219 (111 single left hemisphere regions), 448 (225 single left hemisphere regions) and 998 (499 single left!3

4 hemisphere regions) subdivisions of the Desikan-Killiany atlas are used. All three additional subdivisions revealed high correlations (and highly significant effects), being DK-228 (r=0.62, p=8.04x10-13), DK-500 (r=0.66, p=3.844x10-29) and DK-1000 (r=0.56, p=1.95x10-42). Supplemental Figure S2. In addition to a cross-technique analysis using MRI cortical thickness values based on a common subdivision of the Desikan-Killiany atlas (see main text), a second analysis was performed using a direct mapping of the cortical Von Economo regions to a newly segmented atlas in FreeSurfer (see main text, and Supplemental Figure S1 for this mapping). Panel A shows the original Von Economo Koskinas atlas. Panel B shows a lateral (left panel) and!4

5 medial (right panel) view of the new FreeSurfer-VonEconomo atlas on the 3D cortical surface rendering of five random exemplary subjects of the Human Connectome Project. Supplemental Figure S3. Panel A shows the correlation between Von Economo and Koskinas measurements of cortical width and MRI derived cortical thickness values for the subset of HCP subjects in the age class of (96 subjects, Desikan-Killiany atlas 57 single hemisphere subregions). Panel B shows the Von Economo - MRI correlation for the n=40 Utrecht subset (age years, Desikan-Killiany atlas 57 single hemisphere subregions). Supplemental Table 1. Mapping Von Economo atlas to Desikan-Killiany atlas. Region numbers in the Desikan-Killiany atlas correspond to the region numbers shown in Figure 1. Von Economo - Koskinas atlas regioncode regionname Desikan-Killiany atlas regionnumber(s) and regionname(s)* Frontal lobe FA precentral gyrus precentral_1, precentral_2, precentral_3, superiorfrontal_4!5

6 FB granular frontal area precentral_4, superiorfrontal_4, caudalmiddlefrontal_1 FC intermediate frontal area superiorfrontal_3, caudalmiddlefrontal_1 FCBm magnocellular agranular intermediated frontal parsopercularis_1 FD granular frontal area superiorfrontal_2, rostralmiddlefrontal_1 FDdelta middle granular frontal area rostralmiddlefrontal_2 FDT triangular granular frontal area parstriangularis_1 FE frontalpolar area parsorbitalis_1, rostralmiddlefrontal_3, superiorfrontal_1, frontalpole_1 FF orbital area (agranula orbital area Ffalpha) parsorbitalis_1, lateralorbitofrontal_1, lateralorbitofrontal _2 FG straight area lateralorbitofrontal_2 FH prefrontal area medialorbitofrontal_1 FI frontoinsular area lateralorbitofrontal_1 FK frontal piriform area lateralorbitofrontal_1 FL parolfactory area medialorbitofrontal_1 FM geniculate area medialorbitofrontal_1 FN precommissural medialorbitofrontal_1 Parietal lobe PA postcentral giant pyramidal area PA1 and postparacentral giant pyramidal area PA2 postcentral_1, postcentral_2, postcentral_3, paracentral_1 PB1 oral postcentral area granulosa postcentral_1, postcentral_2, postcentral_3 PB2 oral postcentral area simplex postcentral_1, postcentral_2, postcentral_3 PC intermediate postcentral area postcentral_1, postcentral_2, postcentral_3 PD caudal postcentral area postcentral_1, postcentral_2, postcentral_3 PE superior parietal area superiorparietal_1, superiorparietal_2, superiorparietal_3, precuneus_2, precuneus_1 PF supramarginal area supramarginal_2, supramarginal_1!6

7 PG angular area inferiorparietal_1, inferiorparietal_2 PH Basal temporo-occipital parietal area inferiortemporal_2, middletemporal_1, bankssts_1, fusiform_1, lingual_1 Insular cortex IA precentral insular area insula_2 IB postcentral insular area insula_1 Occipital lobe OA peristriate area lateraloccipital_2, inferiorparietal_1, superiorparietal_3, cuneus_1, lingual_1, lingual_2 OB parastriate area cuneus_1, lingual_1, lingual_2, 1 lateraloccipital_1 OC striate area 1 lateraloccipital_1, pericalcarine_1 Temporal Lobe TA superior temporal area transversetemporal_1, superiortemporal_2 TB supratemporal area simplex transversetemporal_1 TC supratempral area granulosa transversetemporal_1 TD intercalated supratemporal area transversetemporal_1 TE temporal area proper inferiortemporal_2, middletemporal_1, superiortemporal_2, middletemporal_2, inferiortemporal_1, superiortemporal_1 TF fusiform area fusiform_2 TG temporopolar area temporalpole_1, inferiortemporal_1 Cingulate cortex LA1 LA2 precingulate agranular anterior limbic area precingulate agranular anterior limbic area rostralanteriorcingulate_1, caudalanteriorcingulate_1 rostralanteriorcingulate_1, caudalanteriorcingulate_1 LC1 dorsal posterior cingulate area precuneus_2 LC2 ventral posterior cingulate area posteriorcingulate_1, isthmuscingulate_1 LC3 precingulate agranular anterior limbic area posteriorcingulate_1!7 LD angranular retrosplenial area isthmuscingulate_1

8 LE1 superior retrosplenial area granulosa isthmuscingulate_1 LE2 inferior retrosplenial area granulosa isthmuscingulate_1 Hippocampal lobe HA uncinate area entorhinal_1 HB parauncinate area entorhinal_1 HC rhinal area limitans parahippocampal_1 * region subnumbers are study specific!8