Abstract #3828

Comparison of multi-atlas based and template based method in the quantitative analysis of [18F]-FP-DTBZ PET

Yiwei Pan¹, Shuying Liu², Yao Zeng¹, Chenfei Ye³, Hongwen Qiao², Tianbing Song², Haiyan Lv⁴, Piu Chan², Jie Lu², and Ting Ma^1,2,3

¹Department of Electronic and Information Engineering, Harbin Institute of Technology at Shenzhen, Shenzhen, China, ²Xuanwu Hospital Capital Medical University, Beijing, China, ³Peng Cheng Laboratory, Shenzhen, China, ⁴Mindsgo Life Science Shenzhen Co. Ltd, Shenzhen, China

Synopsis

[18F]-FP-DTBZ PET provides reliable information for the diagnosis of Parkinson’s disease. Quantitative analysis of PET images requires a precise segmentation of the region of interest. Two [18F]-FP-DTBZ PET image quantification methods, multi-atlas based and template based methods were compared in this study. A total of 68 subjects were included, each of them underwent 3D T1-weighted MR imaging and [18F]-FP-DTBZ PET imaging. Twenty subjects were used to build atlases and 48 were used for validation. Using dice coefficient and ICC coefficient for evaluation, we observed that the multi-atlas based method showed better performance than the template based method.

This abstract and the presentation materials are available to members only; a login is required.

Join Here