Navigation

Recent Publications

More...

Home

Structural MRI measures in Alzheimer’s disease and effects of ethnicity

zhangl87 - Posted on 06 December 2012

Project Description: 

Background:
Worldwide, Alzheimer's disease (AD) is the most common form of dementia, posing a major burden for healthcare (~US$604 billion spent in 2010). In Hong Kong, 6.1% of people aged 70+ have dementia, 65% of which is AD. It is estimated that many others are misdiagnosed, so actual incidence rate may be even higher.
At present, there is no cure but early diagnosis and intervention can slow disease progression. The current gold standard for diagnosing AD is through cognitive tests, but this can only catch symptomatic patients. Presently, definitive diagnosis can only be made post-mortem. Neuroimaging is non-invasive, safe, and can measure many variables visually and quantitatively. Structural MRI measures such as hippocampal volumetry and cortical thickness have been shown to decrease with disease progression. Manual hippocampal volumetry is currently the gold standard for obtaining these measures, but semi-automatic data analysis programs, such as FreeSurfer (http://surfer.nmr.mgh.harvard.edu/), have been shown to provide similar levels of accuracy.
There is speculation that ethnicity may be a confounding variable when conducting large population studies. Morphometric differences between European and Japanese brains have been found, though the implications for subcortical structures are unknown. Genetically, the APOE allele frequency is known to be slightly different between ethnic groups, and it has been postulated that APOE genotypes can affect cortical thickness. With the increasing number of collaborative international neuroimaging studies, the effect of ethnicity on these measures should be investigated.

Purpose:
In our study, we will run the same structural MRI analyses using FreeSurfer on a local Chinese cohort and a matched Caucasian cohort obtained from the Alzheimer's Disease Neuroimaging Initiative’s database to address the following questions:
1. Will the use of cortical thickness and hippocampal volume combined increase sensitivity and specificity for AD than individually?
2a. Is it possible that ethnicity can have an effect on cortical thickness and brain morphometry?
2b. If so, would future cohorts need to be analysed separately using population-specific templates?

Researcher name: 
Linda Zhang
Researcher position: 
PhD student
Researcher department: 
Department of Diagnostic Radiology
Researcher email: 
lzhang87@gmail.com
Research Project Details
Project Duration: 
09/2012 to 05/2013
Project Significance: 
Possible outcomes: 1. Values of cortical thickness and hippocampal volume will be smaller in AD patient group, and have greater sensitivity and specificity for AD diagnosis than on their own. 2. Ethnicity will have some effect on these values, and using a population-specific template for normalisation can improve accuracy of the data. Relevance, significance, value: The goal of AD research, apart from finding a cure, is to find better methods of diagnosing patients, particularly when they are presymptomatic. If using structural MRI measures can increase the accuracy of diagnosis, it would be an important step towards a non-invasive and definitive diagnosis. As the number of large collaborative population projects increases, such as the Alzheimer's Disease Neuroimaging Initiative (ADNI) and the Australian Imaging, Biomarkers and Lifestyle study (AIBL), it is necessary to consider whether or not ethnicity could be a confounding variable, as these studies have subjects of multiple ethnicities. If there is an effect, future studies of this sort will need to have methods of avoiding bias, such as using population specific templates when conducting analysis.
Remarks: 
FreeSurfer is a processing-intensive program, requiring 48 hours to analyse the data of one participant on a regular workstation. HPC will hopefully reduce that time, so that analysis of larger groups will not take an impractical duration.