Fine-grained functional parcellation maps of the infant cerebral cortex
- Key Laboratory of Biomedical Information Engineering of Ministry of Education, Department of Biomedical Engineering, School of Life Science and Technology, Xi'an Jiaotong University, China
- Department of Radiology and Biomedical Research Imaging Center, the University of North Carolina at Chapel Hill, United States
- Department of Psychiatry, the University of North Carolina at Chapel Hill, United States
Figures
Figure 1
Comparison of the mean local gradient maps of functional connectivity (FC) of the 3-month age group generated by different methods, where more detailed and clearer local gradient patterns are revealed by our proposed method.
(a) The local gradient map computed directly on the average FC matrix across individuals as in Gordon et al., 2016. (b) The local gradient map computed on each individual and then averaged across …
Figure 2
The local gradient maps and parcellation maps, where M stands for months, and 3M, 6M, 9M, 12M, 18M, and 24M represent different age groups in months defined in Table 1.
(a) The age-related local gradient maps generated at different ages, where brighter colors represent higher local gradient. (b) Age-related fine-grained functional parcellation maps were obtained …
Figure 3
The across-age variabilities of local gradient maps and the variabilities between age-related and age-independent local gradient maps, where M stands for months, and 3M, 6M, 9M, 12M, 18M, and 24M represent different age groups in months defined in Table 1.
(a) Across-age variabilities of local gradient maps between every two consecutive ages, where ‘3M to 6M’ measures the variability of the local gradient maps between 3 and 6 months age groups and so …
Figure 4
Local gradient map-derived parcels correspond to known cortical areas with low Hausdolff distances, and show lower variability and higher homogeneity than null models.
(a) The age-independent local gradient map obtained by averaging all age-related local gradient maps. (b) The age-independent parcellation map obtained by applying the watershed algorithm on the …
Figure 5
The spatiotemporal network developmental trend of infant brain functional networks.
Herein, M stands for months, and 3M, 6M, 9M, 12M, 18M, and 24M represent different age groups in months defined in Table 1. (a) Stabilities of different network numbers of different age groups …
Figure 6
Parcel-wise development maps of infant brain functional architecture.
(a) shows the parcel-wise homogeneity monotonically decreases with age, and (b) indicates that the efficiency of each parcel shows multipeak fluctuation.
Figure 7
Longitudinal distribution of scans.
Each point represents a scan with its scanned age (in months) shown on the x-axis, with males in blue and females in red, and each horizontal line represents one subject, with males in blue and …
Figure 8
An illustration of the procedure of infant cortical parcellation using local gradient map of functional connectivity.
Detailed steps are described in the following corresponding parts as: structural and functional MRI processing (sections ‘Structural MRI processing’, ‘Resting-state fMRI processing’, and ‘Cortical …
Author response image 1
The comparison of clustering stability of different methods.
The red line refers to the clustering stability when binarizing the correlation matrices first and then averaging the matrices across individuals, while the blue line refers to the clustering …
Author response image 2
Top panel: the network clustering results using all data in the original manuscript.
Bottom panel: the network clustering results using majority voting through 100 times of bootstrapping. Black circles and red arrows point to the parahippocampal gyrus, which was included in the …
Author response image 3
Parcellation boundaries from different sites.
The black color indicates the parcel boundaries agreed by both sites, and the red and yellow colors show the boundaries from UNC and UMN, respectively. The parcellation boundaries from the two sites …
Tables
Table 1
Demographic information of each age group from the longitudinal dataset under study.
| Age group | Age range (days) | fMRI scans | APscans | PAscans | Structural MRI scans (males/females) | Mean ± stdgestational age (days) | |
|---|---|---|---|---|---|---|---|
| 3M | 10–144 (98.2 ± 35.1) | 104 | 52 | 52 | 50 (25/25) | 280.2 ± 7.1 | |
| 6M | 145–223 (184.2 ± 22.8) | 154 | 77 | 77 | 48 (24/24) | 279.7 ± 9.6 | |
| 9M | 224–318 (279.1 ± 24) | 132 | 66 | 66 | 52 (26/26) | 280.3 ± 6.8 | |
| 12M | 319–410 (367.7 ± 21) | 132 | 66 | 66 | 48 (24/24) | 279.8 ± 7.8 | |
| 18M | 411–591 (494.5 ± 53.2) | 222 | 111 | 111 | 80 (40/40) | 277.6 ± 7.3 | |
| 24M | 592–874 (722.9 ± 66.3) | 242 | 121 | 121 | 82 (41/41) | 277.3 ± 6.2 | |
| Total | 10–874 (414.9 ± 220.1) | 986 | 493 | 493 | 360 (180/180) | 278.0 ± 7.8 | |
Author response table 1
Reproducibility of gradient density maps by rebalancing the sex in each age group (M: months).
| Threshold | 3 M | 6 M | 9 M | 12 M | 18 M | 24 M |
|---|---|---|---|---|---|---|
| Top 25% | 95.25% | 97.00% | 96.73% | 91.95% | 94.96% | 95.15% |
| Top 50% | 96.70% | 97.80% | 97.37% | 95.30% | 97.32% | 97.33% |
