Parametric response mapping of longitudinal PET scans and their use in detecting changes in Alzheimer’s diseases

Purpose The purpose of this study is to determine the effectiveness of an image analysis method called parametric response mapping (PRM) in predicting changes related to progression of Alzheimer’s disease (AD). Methods Forty patients were obtained from the ADNI database. Each patient underwent longitudinal 18 F Fludeoxyglucose (FDG) positron emission tomography (PET) imaging. The patients were divided into four groups, the NC group (n = 10) with stable normal control (NC) patients, the NC-to-MCI group (n = 10) with converting patients from NC to mild cognitive impairment (MCI), the MCI group (n = 10) with stable MCI patients, and the MCI-to-AD group (n = 10) with converting patients from MCI to AD. The PRM approach was applied to longitudinal PET scans using the following procedures; 1) Longitudinal scans were co-registered, 2) intensities were normalized using the cerebral mask, 3) an expert drew regions of interest (ROIs) for tumor volumes, and 4) voxels were classified into unchanged, increased, and decreased categories. Results PRM analysis results were analyzed with linear regression and the slopes of regression lines corresponding to different groups were compared. The PRM analysis was able to distinguish between the NC and NC-to-MCI group with p-value 0.05, while it was not able to distinguish between the MCI and the MCI-to-AD group with p-value 0.54. Conclusions An image analysis method, PRM approach, was able to distinguish between stable NC patients and converting patients from NC to MCI.