INTRODUCTION This project investigated volumetric differences in certain subcortical structures as measured on high-resolution structural Magnetic Resonance Imaging (MRI) scans traced manually. The sample comprised 79 5-year old children, 52 with HIV and 27 uninfected controls. Infected children were all stable on antiretroviral therapy (ART) and were from the Children with HIV early antiretroviral (CHER) cohort who have been followed since birth. The study aimed to investigate the long-term effects of HIV and ART on the developing brain. While high-resolution structural data has been analysed using automated FreeSurfer to determine volume and cortical thickness, manual tracing remains the gold standard. Thus, manual tracing was used to validate automated measures and investigate subtle group differences in selected regions of interest. METHODS Extensive clinical data were available for all participants in the study. MR images were AC-PC transformed and converted to analyse format. Structures were traced using MultiTracer software. Structures selected included the caudate, nucleus accumbens (NA), putamen (Pu), globus pallidus (GP) and corpus callosum (CC). Four of these structures occur bilaterally. Tracing was performed in 79 subjects. Three subjects were excluded due to poor quality images or pathology; 5 HIV-1 infected children were excluded as they were not randomized between treatment groups. Certain subjects were retraced for inter and intrarater reliabilities. The effect and association of ethnicity, age, birthweight and sex as possible confounders were investigated. As the groups were not well matched for ethnicity, all Cape Coloured children were excluded from further analyses. Analysis of variance was used to test the effect on structure size between HIV-1 infected children and controls, as well as between 3 treatment arms (ART deferred until clinical criteria were met, early ART for 40 weeks, early ART for 96 weeks) and uninfected controls. Analysis of covariance was used to control for the possible confounding effects of sex and age. Each structure was tested for possible association with clinical variables (CD4, CD8, CD4/CD8 ratio and CD4%) both at enrolment and time of scanning. Linear regressions were modelled using clinical variables that showed significant correlation with structure size whilst controlling for covariates. Congruence between automated FreeSurfer and manual segmentation were evaluated via Bland-Altman, Pearson r and Cronbach's alpha.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/19894 |
| Date | January 2015 |
| Creators | Randall, Steven Ronald |
| Contributors | Meintjes, Ernesta M, Warton, Christopher |
| Publisher | University of Cape Town, Faculty of Health Sciences, Division of Anatomical Pathology |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Master Thesis, Masters, MSc (Med) |
| Format | application/pdf |
Page generated in 0.0024 seconds