Diffusion Kurtosis Imaging as a Diagnostic Tool for Parkinson's Disease
MetadataShow full item record
Reliable diagnosis of Parkinson's disease requires long-term assessment of a patient's motor performance tests and response to medication. Though the development of magnetic resonance imaging (MRI) presents an additional tool in making a diagnosis, limited imaging biomarkers have been reported that support a clinical diagnosis of Parkinson's disease or its differentiation from similarly presenting diseases. Diffusion Kurtosis Imaging (DKI) is an MRI method that quantifies deviation of water diffusion from normal Gaussian distribution. DKI is a more sensitive technique than conventional diffusion tensor imaging (DTI) for assessing tissue microstructure. The parameters provided by DKI analysis, particularly the mean kurtosis, reflect structural changes within brain regions and demonstrate potential as a diagnostic tool for Parkinson's disease where the basal ganglia are known to markedly change. Here we examine the DKI maps of 86 patients from Hospital das Clinicas da FMUSP in Sao Paulo, Brazil. 49 patients presented with a previous diagnosis of Parkinson's disease based on the UK Parkinson's Brain Bank criteria (mean age, 65.3 + 8.7 [standard deviation]), 19 patients with a previous diagnosis of essential tremor based on the Movement Disorder Society standards (mean age, 64.7 + 6.7) and 27 patients were age-matched healthy controls (mean age, 64.5 + 10.9). All patients underwent the same 3T MRI procedure, consisting of a DTI scan with 32 different gradient directions and b values of 0, 1000, and 2000 s/mm2, necessary to construct a DKI map. Using a region of interest (ROI) analysis on the substantia nigra (rostral, middle, and caudal) and putamen for each patient and comparing mean kurtosis values, we find no significant differentiation of Parkinson's disease patients in the substantia nigra, but significantly higher mean kurtosis values in the putamen of Parkinson's patients (0.82 + 0.05 [standard deviation]) than healthy controls (0.60 + 0.04, p = 0.0158). Neither analysis demonstrated significant difference from essential tremor patients. Higher mean kurtosis estimates in the basal ganglia of Parkinson's disease patients may reflect changes in the microstructural environment of these structures related to disease progression. Further studies should investigate the histological correlates of these values and the reliability of DKI estimates as a diagnostic tool in various stages of the disease.