Multidimensional Neuroimaging Biomarkers in MCI

Project Outlines and Aims

The Multidimensional Neuroimaging Biomarkers in MCI (Mild cognitive impairment) is a study of computational neuroanatomy.

Dementia in the elderly is a major and increasing public health concern. Mild cognitive impairment is more common in older adults than dementia, and this population is at increased risk of developing dementia. Advances in neuroscience and neuroimaging have led to an increasing recognition that there is a decline in brain size with increasing age, the gradient of which increases in neurodegenerative disease. Certain brain anatomical structures may be preferentially affected in certain diseases. The distribution of structural changes reflects the underlying pathology and may determine the clinical phenomenology. The increasing sophistication of magnetic resonance imaging (MRI) allows brain structures to be visualized in vivo in unprecedented detail, e.g. clinical MRI is able to give good soft-tissue contrast and high spatial (<1 mm) resolution. Due to rapid advances in computing power and algorithm development, we are now well placed to study the extraordinary MRI-visible morphological variability of the human brain through mathematical models sensitive to subtle changes in neuroanatomical shape, complexity and tissue characteristics.

One of the key requirements of neuroimaging research is the multi-disciplinary collaborations between neuroscience, medicine and engineering and a host of sub-disciplines. While the research subjects and the clinical needs are provided by medicine, the theoretical underpinnings come from mathematics and statistics, and the methodological approaches from computer science and biomedical engineering, all forming critical components of the enterprise.

The main aim of the study is to use and further develop computational and mathematical algorithms which have been developed within our group for the past few years for measuring brain structural features in a reliable and automated manner. These methods/algorithms will then be further developed to establish the proper multidimensional biomarkers for the prediction of conversion of normal cognition (NC) to MCI, and from MCI to dementia. We plan to develop them as clinical tools to optimise the use of structural MRI for early diagnosis of MCI and dementia. While MRI is now widely available and frequently used in the diagnostic assessment of dementia, its visual interpretation by an “expert” gleans only a fraction of the information available and the interpretation is subject to bias and poor reliability. The tools we plan to develop will help unleash the full power of brain MRI and automate the process to improve reliable interpretation.

We (UNSW) are collaborating with three universities overseas, i.e. Johns Hopkins University, USA; Washington University in St. Louis, USA and Simon Fraser University, Canada in this study.

The MR images we study include 3D T1-weighted scans, T2-weighted (such as FLAIR sequence) scans, DTI (diffusion tensor imaging), 1H MRS (magnetic resonance spectroscopy) and will extend to ASL (arterial-spin labeling) etc.