Abstract
1295
Objectives 18F-FP-CIT PET is an effective imaging study for dopamine transporter. In quantitative analysis of 18F-FP-CIT PET, operator-dependency is a concern when volume of interest (VOI) is manually drawn for the striatum and reference site. In this study, we adopted an automatic method for analysis of 18F-FP-CIT PET and clinically validated the method.
Methods A total of 77 18F-FP-CIT PET images were analyzed. The uptake of 18F-FP-CIT in the caudate and the putamen was visually graded by 2 experts using 4-grade system. Afterward, images were analyzed by automatic software, in which each image was normalized to averaged 18F-FP-CIT PET template using SPM algorithm, and uptakes of 18F-FP-CIT in the caudate, putamen, and cerebellum were automatically measured by statistical probabilistic anatomical map (SPAM) for template brain. BP of the caudate and the putamen was calculated from the uptake measurement and compared with visual grading. The effect of brain atrophy on the automatic method was also tested.
Results The automatic analysis was successful in all the cases and the reproducibility was perfect. When the calculated BP was compared with visual grade, BP of the caudate was well correlated with visual grade, and was significantly different with varying grade (4.00±2.74, 3.33±1.91, 2.60±0.91, 1.92±1.40 for grade 0, 1, 2, 3, respectively; p<0.001). Also in the putamen, BP was significantly different with varying grade (4.90±2.84, 3.21±1.42, 3.01±1.42, 2.57±1.55 for grade 0, 1, 2, 3, respectively; p<0.001). However, in case of severe brain atrophy, the correlation between calculated BP and visual grade was relatively low.
Conclusions In this study, we adopted SPM and SPAM method for automatic analysis of 18F-FP-CIT PET. The method showed perfect reproducibility and excellent correlation with conventional visual assessment although severe brain atrophy was a minor limitation. This method can be easily applied to clinical image analysis of 18F-FP-CIT PET