Synthesis and Evaluation of Butyrylcholinesterase Ligands for Neuroimaging Alzheimer's Disease

Abstract

Background: White matter lesions (WML) are common in MRI images of the elderly. We have previously developed an analysis program AAL, the automated anatomical labeling of brain structures in SPM using a macroscopic anatomical parcellation. In this study we added an additional function to label the white matter lesions and test the program in 64 MRI studies collected from patients with Alzheimer’s disease (AD), mild cognitive impairment (MCI), and normal controls (NC). Methods: Sixty-four subjects, including 39 patients with AD (age 1⁄4 72.5 6 8.5, M/F 1⁄4 14/25), 13 individuals with MCI (age 1⁄4 72.8 6 4.3, M/F 1⁄4 3/10) and 12 normal controls (age 1⁄4 71.1 6 8.1, M/F 1⁄4 5/7) were analyzed in the study. The high-resolution 3D T1-weighted images, T2-weighted images, and FLAIR images acquired with a Siemens 3T scanner were input into the program. The gross compartment volumes in GM, WM, CSF and intracranial volume (ICV) were obtained in all subjects. The AAL program also yielded volumes from 106 regions-of-interest (ROI). For analysis of WML, all three aligned datasets (T1, T2, FLAIR) were segmented into GM, WM and CSF in the intracranial volume, and a lesion probability map was generated. The WML was defined as the pixels with the probability above a threshold, and the volume was integrated over the entire white matter space. Results: The AD group had a smaller GM volume than normal controls (P 1⁄4 0.011), and had higher CSF volume than MCI (P 1⁄4 0.003) and control groups (P 1⁄4 0.000). The volumes of medial temporal lobe structures in the AD group were significantly smaller than in the MCI or NC groups. The cerebellum was not significantly different among all 3 groups. Figure 1 shows an example of the WML segmented using the automated labeling procedure. Conclusions: The AAL program provides an efficient method for segmentation of each individual brain, and shows a significant gray matter loss in AD patients. The initial results of WML segmentation were promising, and currently being verified against other established methods, including visual rating and manual segmentation.