deepNeuroSeg 0.1

Deep-learning Tool for White Matter (WM) lesions and Claustrum structure segmentation in brain magnetic resonance imaging (MRI).

deepNeuroSeg

There are two different segmentation tasks you can perform: White Matter Lesions (WMH - White Matter Hyperintensities) and Claustrum Segmentation. The required models will be downloaded to ~/.deepNeuroSeg as soon as needed. WMH segmentation can be performed either using FLAIR only or both FLAIR and T1. The results would be similar to our experience [1]. The default is using both of them.

Fig.1: Segmentation result on Singapore 34 [1]. From left to right: FLAIR MR image, the associated ground truth, segmentation result using FLAIR modality only and segmentation result using FLAIR and T1 modalities. In column SegF LAIR and SegF LAIR+T1, the green area is the overlap between the segmentation maps and the ground-truth, the red pixels are the false negatives and the black ones are the false positives.

Claustrum segmentation is still in development.

How to:

Run deepNeuroSeg either in command line or Python.

Command line

deepNeuroSeg performs two different segmentation tasks. The desired task must be specified with --type that is either equals to wmh (White Matter Hyperintensities (Lesions)) or c (Claustrum). For example:

deepNeuroSeg --type wmh --flair YOUR_PATH.nii.gz --o YOUR_PATH

For more details see:

deepNeuroSeg --help
Options:
  --type [wmh|c]  Either 'wmh' (White Matter Hyperintensities) or 'c'
                  (Claustrum)
  --flair TEXT    Path to .nii.gz file of a FLAIR scan.  [required]
  --t1 TEXT       Path to .nii.gz file of a T1 scan. [optional]
  --o TEXT        Directory path where to save the resulting segmentation.
                  [required]

The resulting mask will be saved in the user-specified directory under the name out_mask.nii.gz.

Python

In Python user will have to follow the following steps by creating a SegmentationFactory object that can have a segmentation type either SegmentationType.WMH or SegmentationType.Claustrum. An example for WMH Segmentation with both FLAIR and T1 modalities:

from deepNeuroSeg import SegmentationFactory, SegmentationType

segmenter = SegmentationFactory.create_segmenter(SegmentationType.WMH, 
                                                    FLAIR_path='YOUR_PATH',
                                                    T1_path='YOUR_PATH')

An example for WMH Segmentation with FLAIR only:

from deepNeuroSeg import SegmentationFactory, SegmentationType

segmenter = SegmentationFactory.create_segmenter(SegmentationType.WMH, 
                                                    FLAIR_path='YOUR_PATH')

Then the segmentation can be performed. Here user can specify the output directory where the segmentation mask will be saved as out_mask.nii.gz:

prediction = segmenter.perform_segmentation(outputDir='YOUR_PATH')

or inspect the numpy array yourself and save it with save_segmentation method.

prediction = segmenter.perform_segmentation()
segmenter.save_segmentation(mask=prediction, outputDir='YOUR_PATH')

In both cases, the prediction can be saved in the output directory desired by the user under the name out_mask.nii.gz.

[1]: Li, Hongwei, et al. "Fully convolutional network ensembles for white matter hyperintensities segmentation in MR images." NeuroImage 183 (2018): 650-665.