Getting Started

Neuroslice provides two interfaces for brain tumor segmentation:

  1. Command-Line Interface (CLI) - Quick processing from the terminal

  2. Python - Integration into your scripts and pipelines

Both interfaces provide access to the same core functionality, allowing you to choose based on your workflow preferences.

Quick Examples

Command-line

# Basic usage
neuroslice FLAIR.nii.gz tumor_mask.nii.gz

# With options
neuroslice FLAIR.nii.gz tumor_mask.nii.gz --axis 2 --mode cuboid --verbose

Python

from neuroslice import predict_mask
import nibabel as nib

# Generate mask
mask = predict_mask("FLAIR.nii.gz", axis=2, mode="cuboid", verbose=True)

# Save output
nifti = nib.load("FLAIR.nii.gz")
output = nib.Nifti1Image(mask.astype("uint8"), nifti.affine, nifti.header)
nib.save(output, "tumor_mask.nii.gz")

Note

  • Try different orientations if results are suboptimal

  • Use verbose during initial testing

  • Consider using cuboid mode for downstream processing

See also