In this paper, we propose a novel patch-based mean-shift algorithm for constructing diffusion MRI atlases. Our method is less sensitive to outliers and is able to deal with inter-subject fiber dispersion. Preliminary experimental results indicate that our method yields improvements over the commonly used simple averaging method and generates diffusion atlases with cleaner fiber orientations and less artifacts caused by inter-subject orientation dispersion.
Acknowledgements This work was supported in part by NIH grants (NS093842, EB006733, EB009634, AG041721, MH100217, and AA012388) and the National Natural Science Foundation of China (No. 61540047).
2. Deshpande, R., Chang, L., Oishi, K.: Construction and application of human neonatal DTI atlases. Front. Neuroanat. 9, 138 (2015)
3. Johansen-Berg, H., Behrens, T.E.J.: Diffusion MRI: From Quantitative Measurement to In Vivo Neuroanatomy, 2nd edn. Academic, San Diego (2014). ISBN 9780123964601, http://dx.doi.org/10.1016/B978-0-12-396460-1.01001-5
4. Comaniciu, D., Meer, P.: Mean shift: a robust approach toward feature space analysis. IEEE Trans. Patt. Anal. Mach. Intell. 24(5), 603-619 (2002)
5. Buades, A., Coll, B., Morel, J.M.: A review of image denoising algorithms, with a new one. Multisc. Model. Simul. 4(2), 490-530 (2005)