Perinatal development of the corpus callosum in term and preterm infants
Poster No:
1243
Submission Type:
Abstract Submission
Authors:
Wei Wang1, Liyuan Yang1, Gaolang Gong1
Institutions:
1State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea, Beijing, Beijing
First Author:
Wei Wang
State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea
Beijing, Beijing
State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea
Beijing, Beijing
Co-Author(s):
Liyuan Yang
State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea
Beijing, Beijing
State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea
Beijing, Beijing
Gaolang Gong
State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea
Beijing, Beijing
State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Resea
Beijing, Beijing
Introduction:
The corpus callosum (CC) is the largest commissural tract, supporting interhemispheric communication and hemispheric specialization of various brain functions [1]. In fetuses and newborns, the development of this particular structure has been frequently adopted as a key indicator of the entire brain development, possibly due to the easy identification and quantification of the midsagittal CC on MRI or ultrasound images [2]. Previous studies have demonstrated CC abnormalities in adults with preterm birth [3, 4], but the impact of preterm birth on perinatal CC development remains unexplored. To address this, we included multimodal perinatal MRI dataset from term and preterm babies and compared their MRI-derived measures. Furthermore, we evaluated the association between perinatal CC microstructures and neurobehavioral outcomes at 18 months old.
Methods:
In total, 59 preterm and 381 term infants from the developing Human Connectome Project (dHCP) were included (Preterm group: 37 males, gestational age (GA) = 23.7-36.9 wks, postmenstrual age (PMA) = 37-44.9 wks; Term group: 205 males, GA = 37-42.3 wks, PMA = 37.4-44.7 wks).
The dHCP preprocessed structural and diffusion MRI dataset was used [5]. For each baby, the CC was manually delineated on the native-space midsagittal slice of aligned T2w images. The outlined midsagittal CC was then divided into five subregions according to the Hofer scheme [6]. Diffusion tensor and neurite orientation dispersion and density imaging (NODDI) were estimated. The mean fractional anisotropy (FA), diffusivity (MD), neurite density index (NDI), orientation dispersion index (ODI), and total area were calculated for each midsagittal CC subregion.
For each subregional measure above, we applied the normative modeling (implemented in the PCNtoolkit [7]) with data from all term infants. For this model, PMA and sex were input as predictors, and 10-fold cross-validation was applied. For each infant (both term and preterm), an individual z-score was derived from the normative model, representing his/her deviation from the normative trajectory during perinatal period.
For each infant, neurobehavioral outcomes at 18 months old were assessed by the Bayley Ⅲ Scales, yielding 5 scores for the cognition, receptive language, expressive language, gross motor, and fine motor ability, respectively.
For each measure, two-sample t-test was used to test group difference in the z-scores. In the two groups, partial least square correlation (PLSC) analysis was separately performed to examine the association between perinatal CC microstructural measures and Bayley Ⅲ scores at 18 months old. The Bonferroni method was applied to correct for multiple comparisons, and corrected p < 0.05 was considered as the significance level.
The dHCP preprocessed structural and diffusion MRI dataset was used [5]. For each baby, the CC was manually delineated on the native-space midsagittal slice of aligned T2w images. The outlined midsagittal CC was then divided into five subregions according to the Hofer scheme [6]. Diffusion tensor and neurite orientation dispersion and density imaging (NODDI) were estimated. The mean fractional anisotropy (FA), diffusivity (MD), neurite density index (NDI), orientation dispersion index (ODI), and total area were calculated for each midsagittal CC subregion.
For each subregional measure above, we applied the normative modeling (implemented in the PCNtoolkit [7]) with data from all term infants. For this model, PMA and sex were input as predictors, and 10-fold cross-validation was applied. For each infant (both term and preterm), an individual z-score was derived from the normative model, representing his/her deviation from the normative trajectory during perinatal period.
For each infant, neurobehavioral outcomes at 18 months old were assessed by the Bayley Ⅲ Scales, yielding 5 scores for the cognition, receptive language, expressive language, gross motor, and fine motor ability, respectively.
For each measure, two-sample t-test was used to test group difference in the z-scores. In the two groups, partial least square correlation (PLSC) analysis was separately performed to examine the association between perinatal CC microstructural measures and Bayley Ⅲ scores at 18 months old. The Bonferroni method was applied to correct for multiple comparisons, and corrected p < 0.05 was considered as the significance level.
Results:
As shown in Fig. 1, there were significant group differences in microstructural measures in at least one CC subregion. Particularly, preterm infants showed significantly lower NDI in all subregions. In contrast, no group difference of 2D midsagittal area was found in any of the subregions.
The PLSC analysis only showed a significant latent component (LC) in the preterm group (r = -0.52, p = 0.006; Fig. 2A). In term infant, applying the associated coefficients of the observed significant LC did not show a significant correlation between CC microstructural measures and neurobehavioral scores (r = 0.04, p = 0.442; Fig. 2B), suggesting the specificity of the observed LC to the preterm infants. For each CC measure and neurobehavioral score, its contribution to this significant LC (as referred to as salience) were illustrated in Fig. 2C.
The PLSC analysis only showed a significant latent component (LC) in the preterm group (r = -0.52, p = 0.006; Fig. 2A). In term infant, applying the associated coefficients of the observed significant LC did not show a significant correlation between CC microstructural measures and neurobehavioral scores (r = 0.04, p = 0.442; Fig. 2B), suggesting the specificity of the observed LC to the preterm infants. For each CC measure and neurobehavioral score, its contribution to this significant LC (as referred to as salience) were illustrated in Fig. 2C.
Conclusions:
The present study demonstrated microstructural underdevelopment of the CC around the perinatal period in preterm infants, especially around the genu and anterior body of the CC. Moreover, the perinatal CC microstructures are multivariably associated with neurobehavioral outcomes at 18 months old, highlighting a crucial role of CC early development in behavioral capability of later life.
Lifespan Development:
Early life, Adolescence, Aging 1
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 2
Multivariate Approaches
Univariate Modeling
Keywords:
MRI
Multivariate
Other - Corpus Callosum; Perinatal Development; Preterm Birth; Normative Modeling
1|2Indicates the priority used for review
Provide references using author date format
[2] De León Reyes, N. S. et al. (2020), 'Development and plasticity of the corpus callosum', Development, vol. 147, no. 18, pp. dev189738.
[3] Thompson, D. K. et al. (2011), 'Characterization of the corpus callosum in very preterm and full-term infants utilizing MRI', NeuroImage, vol. 55, no. 2, pp. 479–490.
[4] Siffredi, V. et al. (2023), 'Corpus callosum structural characteristics in very preterm children and adolescents: Developmental trajectory and relationship to cognitive functioning', Developmental Cognitive Neuroscience, vol. 60, pp. 101211.
[5] Edwards, A. D. et al. (2022), 'The Developing Human Connectome Project Neonatal Data Release', Frontiers in Neuroscience, vol. 16, pp. 886772.
[6] Hofer, S. et al. (2006), 'Topography of the human corpus callosum revisited--comprehensive fiber tractography using diffusion tensor magnetic resonance imaging', NeuroImage, vol. 32, no. 3, pp. 989–994.
[7] Rutherford, S. et al. (2022), 'The normative modeling framework for computational psychiatry', Nature Protocols, vol. 17, no. 7, pp. 1711–1734.