Date of Award

January 2021

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

John A. Persing


The focus of this thesis is on neuromorphometry of the cerebral cortex in syndromic craniosynostosis (sCS). The first aim is to evaluate cerebral cortical thickness in relation to intracranial hypertension risk factors for sCS patients. 171 syndromic magnetic resonance images (MRI) are processed via FreeSurfer pipeline to determine average cortical thickness and intracranial hypertension (ICH) risk factor data such as papilledema, occipito-frontal circumference, hydrocephalus, tonsillar position, and obstructive sleep apnea are gathered. Mixed regression modeling is used to assess for correlations between thickness and ICH data. We show that mean cortical thickness (2.78 ± 0.17 mm) is reduced in syndromic patients with a history of either papilledema (p = 0.036) or hydrocephalus (p = 0.007) which suggests structural maldevelopment due to ICH. A second aim is to determine if cortical thickness varies in Crouzon-Pfeiffer patients relative to synostosis pattern and primary vault procedure type. Cortical thickness of 34 Crouzon-Pfeiffer patients is estimated via FreeSurfer MRI analysis. Primary vault expansion and synostosis pattern data were obtained from medical records. Linear regression is used to assess correlation between procedure type and thickness and cohen’s d is calculated for effect of suture involvement on lobar thickness. We failed to detect thickness differences due to procedure type (p = 0.81). Regionally, we show a large negative effect on cingulate cortex thickness and contralateral lambdoid suture involvement (left d = -1.00 and right d = -1.23) suggesting overall cranial shape rather than local synostosis influences structural cortical development. A final aim is to evaluate cerebral cortical allometry in TWIST1 and FGFR-mediated sCS by assessing cortex surface area (CSA) to intracranial volume (ICV) scaling in comparison to typically developing controls. 203 MRIs (36 controls) are processed via FreeSurfer to generate ICV and CSA values. A scaling coefficient is derived from a log-transformed mixed regression model accounting for age, sex, and repeated measurements. Reduced scaling was observed in FGFR-mediated syndromes ( FGFR = 0.68) compared to controls ( control = 0.77) indicating cortical maldevelopment, particularly in the parietal ( parietal = 0.50) and occipital ( occipital = 0.67) lobes. These findings were also correlated with a lower scholastic achievement indicating a potential neural substrate for neurodevelopmental issues in sCS.


This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.