dots_segmentation¶
-
nighres.brain.dots_segmentation(tensor_image, mask, atlas_dir, wm_atlas=1, max_iter=25, convergence_threshold=0.005, s_I=0.023809523809523808, c_O=0.5, max_angle=67.5, save_data=False, overwrite=False, output_dir=None, file_name=None)[source]¶ DOTS segmentation
Segment major white matter tracts in diffusion tensor images using Diffusion Oriented Tract Segmentation (DOTS) algorithm.
Parameters: - tensor_image (niimg) – Input image containing the diffusion tensor coefficients in the following order: volumes 0-5: D11, D22, D33, D12, D13, D23
- mask (niimg) – Binary brain mask image which limits computation to the defined volume.
- atlas_dir (str) – Path to directory where the DOTS atlas information is stored. The atlas information should be stored in a subdirectory called ‘DOTS_atlas’ as generated by nighres.data.download_DOTS_atlas().
- wm_atlas (int, optional) – Define which white matter atlas to use. Option 1 for 23 tracts [2] and option 2 for 39 tracts [1]. (default is 1)
- max_iter (int, optional) – Maximum number of iterations in the conditional modes algorithm. (default is 20)
- convergence_threshold (float, optional) – Threshold for when the iterated conditonal modes algorithm is considered to have converged. Defined as the fraction of labels that change during one step of the algorithm. (default is 0.002)
- s_I (float, optional) – Parameter controlling how isotropic label energies propagate to their neighborhood. (default is 1/42)
- c_O (float, optional) – Weight parameter for unclassified white matter atlas prior. (default is 1/2)
- max_angle (float, optional) – Maximum angle (in degrees) between principal tensor directions before connectivity coefficient c becomes negative. Possible values between 0 and 90. (default is 67.5)
- save_data (bool, optional) – Save output data to file. (default is False)
- overwrite (bool, optional) – Overwrite existing results. (default is False)
- output_dir (str, optional) – Path to desired output directory, will be created if it doesn’t exist.
- file_name (str, optional) – Desired base name for output files without file extension, suffixes will be added.
Returns: Dictionary collecting outputs under the following keys (type of output files in brackets)
- segmentation (array_like): Hard segmentation of white matter.
- posterior (array_like): POsterior probabilities of tracts.
Return type: Notes
Algorithm details can be found in the references below.
References
[1] Bazin, Pierre-Louis, et al. “Direct segmentation of the major white matter tracts in diffusion tensor images.” Neuroimage (2011) doi: https://doi.org/10.1016/j.neuroimage.2011.06.020 [2] Bazin, Pierre-Louis, et al. “Efficient MRF segmentation of DTI white matter tracts using an overlapping fiber model.” Proceedings of the International Workshop on Diffusion Modelling and Fiber Cup (2009)
