Abstract #3323

An Atlas Pre-selection Method for Multi-atlas Based Brain Segmentation

Hengtong Li¹, Heather T Ma¹, Jingbo Ma¹, Chenfei Ye¹, and Shuai Mao¹

¹Department of Electronic and Information Engineering, Harbin Institute of Technology Shenzhen Graduate School, Shenzhen, China, People's Republic of

This study proposed 4L and LV-based methods to improve the brain segmentation accuracy and eddiciency of multi-atlas brain segmentation. The atlas database contains T1 images of brain from 77 subjects. These two methods were adopted to calculate the Dice between the target image and each atlas. We compared the proposed methods with MI-based method and randomly selected method for geriatric, adult and pediatric populations. The segmentation accuracy was evaluated by Dice and the results show that the accuracy of 4L and LV-based methods has great improvement. In addition, the proposed pre-selection is more efficient.

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

Join Here