Differentiating between Normal Aging, Mild Cognitive Impairment, and Alzheimer Ã¢Â€Â™ s disease with FDG-PET: Effects of Normalization Region and Partial Volume Correction Method
- Corresponding Author:
- Ronald J. Killiany
Director, Center for Biomedical Imaging
Boston University, 700 Albany Street W701
Boston, MA 02118, USA
Received date: March 28, 2013; Accepted date: May 04, 2013; Published date: May 10, 2013
Citation: Bauer CM, Cabral HJ, Greve DN, Killiany RJ (2013) Differentiating between Normal Aging, Mild Cognitive Impairment, and Alzheimerâ€™s disease with FDG-PET: Effects of Normalization Region and Partial Volume Correction Method. J Alzheimers Dis Parkinsonism 3:113. doi: 10.4172/2161-0460.1000113
Copyright: © 2013 Bauer CM, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Objective: In Alzheimer’s FDG PET research, the choice of reference region for normalization and use of partial volume correction are inconsistent and have not been studied in a large multi-center study. Herein, we identified which normalization region provided the highest degree of discrimination between subjects who were classified as normal aging, mild cognitive impairment, or Alzheimer’s disease. The effects of partial volume correction using either a gray matter mask or cortical thickness and subcortical volume residuals were also examined. Methods: Stepwise logistic regression models were used to identify the optimal normalization region and partial volume correction method to discriminate between disease stages in over 400 subjects from research sites across North America. Normalization region candidates were the brainstem, precentral gyrus, postcentral gyrus, cerebellum, and thalamus. Partial volume correction methods tested were anatomically or statistically based. Results: Pre- and post- central gyri, and the thalamus showed AD-related changes in FDG PET and did not qualify for further testing. Normalizing to the cerebellum while using the gray matter mask for partial volume correction provided the highest indicator of discrimination. Conclusions: Normalization region and partial volume correction are critical to FDG PET analysis and candidate normalization regions should be tested for disease effects in the study sample prior to use. Cerebellar normalization and gray matter mask partial volume correction are recommended for use with the ADNI dataset.