AFQ.nn.multiaxial#

Functions#

`multiaxial`(ort, img, model_sagittal, model_axial, ...)	Perform multiaxial segmentation using three ONNX models
`run_multiaxial`(ort, t1_img, onnx_kwargs)	Run the multiaxial model.
`extract_brain_mask`(predictions)	Extract brain mask from multiaxial predictions.

Module Contents#

AFQ.nn.multiaxial.multiaxial(ort, img, model_sagittal, model_axial, model_coronal, consensus_model, onnx_kwargs)[source]#

Perform multiaxial segmentation using three ONNX models and a consensus model [1].

Parameters:

imgndarray: 3D T1 image to segment.
model_sagittalstr: Path to sagittal ONNX model.
model_axialstr: Path to axial ONNX model.
model_coronalstr: Path to coronal ONNX model.
consensus_modelstr: Path to consensus ONNX model.
onnx_kwargsdict: ONNX kwargs to use for inference.

Returns:

predndarray: Segmentation labels for each coordinate.

References

[1] Birnbaum, Andrew M., et al. “Full-head segmentation of MRI with abnormal brain anatomy: model and data release.” Journal of Medical Imaging 12.5 (2025): 054001-054001.

AFQ.nn.multiaxial.run_multiaxial(ort, t1_img, onnx_kwargs)[source]#: Run the multiaxial model.

AFQ.nn.multiaxial.extract_brain_mask(predictions)[source]#

Extract brain mask from multiaxial predictions.

Parameters:

predictionsNifti1Image: Multiaxial segmentation predictions.

Returns:

bm_datandarray: Brain mask data.