Abstract
1414
Objectives Hemispheric Cortical Surface Maps (HCSM) of FDDNP give good visualization of beta amyloid plaque (BA) and neurofibillary tangle (NFT) distribution in the brain. To assess how well the distribution pattern predicts cognitive decline, we examined the feasibility and reliability of using a statistical method to predict a subject’s MMSE based on their regional cortical FDDNP uptake pattern.
Methods We applied movement correction to dynamic FDDNP PET data from 23 subjects that have Mini Mental State Examination (MMSE) scores. Logan graphical plots were used, with the cerebellum as reference region, to generate the Distribution Volume Ratio (DVR) image of FDDNP. Early summed FDDNP images were computed. A HCSM for each subject was extracted from the MRI in ICBM space. The same transformations were applied to the DVR and summed FDDNP images to put them in the same cortical space. ROIs were drawn directly on the average HCSM. Linear regression was used to find the rate of change of FDDNP vs MMSE for each HCSM ROI. The models were then used jointly to predict MMSE from a set of ROI values for each subject. Cross validation was used to show the reliability of the estimate.
Results Use of all ROI values of DVR to predict MMSE gave a sample standard deviation (SD) of 2.3 that was comparable to using the global DVR alone. With a stepwise approach, we found that the SD can be reduced to 1.7 by using only DVR values from prefrontal, medial and lateral temporal, parietal regions plus the summed FDDNP values in posterior cingulate.
Conclusions HCSM of FDDNP can provide not only good visualization of brain cortical distribution of BA and NFT, but also reliable estimates of subjects' MMSE, when proper statistical estimation is employed.
Research Support DOE contract DE-FC03-87-ER60615
- © 2009 by Society of Nuclear Medicine