xcp_d.utils.qcmetrics module

Quality control metrics.

xcp_d.utils.qcmetrics.compute_dvars(datat)[source]

Compute standard DVARS.

Parameters: datat (numpy.ndarray) – The data matrix from which to calculate DVARS. Ordered as vertices by timepoints.
Returns: The calculated DVARS array. A (timepoints,) array.
Return type: numpy.ndarray

xcp_d.utils.qcmetrics.compute_registration_qc(bold2t1w_mask, t1w_mask, bold2template_mask, template_mask)[source]

Compute quality of registration metrics.

This function will calculate a series of metrics, including Dice’s similarity index, Jaccard’s coefficient, cross-correlation, and coverage, between the BOLD-to-T1w brain mask and the T1w mask, as well as between the BOLD-to-template brain mask and the template mask.

Parameters

bold2t1w_mask (str) – Path to the BOLD mask in T1w space.
t1w_mask (str) – Path to the T1w mask.
bold2template_mask (str) – Path to the BOLD mask in template space.
template_mask (str) – Path to the template mask.

Returns

reg_qc – Quality control measures between different inputs.

Return type

dict

xcp_d.utils.qcmetrics.coverage(input1, input2)[source]

Estimate the coverage between two masks.

Parameters: input1/input2 (str) – Path to a NIFTI image. Can be any type but will be converted into binary: False where 0, True everywhere else.
Returns: cov – Coverage between two images.
Return type: float

xcp_d.utils.qcmetrics.crosscorr(input1, input2)[source]

Calculate cross correlation between two images.

NOTE: TS- This appears to be Pearson’s correlation, not cross-correlation.

Parameters: input1/input2 (str) – Path to a NIFTI image. Can be any type but will be converted into binary: False where 0, True everywhere else.
Returns: cc – Correlation between the two images.
Return type: float

xcp_d.utils.qcmetrics.dc(input1, input2)[source]

Calculate Dice coefficient between two images.

Computes the Dice coefficient (also known as Sorensen index) between the binary objects in twom j images.

The metric is defined as .. math:

DC=\frac{2|A\cap B|}{|A|+|B|}

, where \(A\) is the first and \(B\) the second set of samples (here: binary objects).

Parameters: input1/input2 (str) – Path to a NIFTI image. Can be any type but will be converted into binary: False where 0, True everywhere else.
Returns: dc – The Dice coefficient between the object(s) in `input1` and the object(s) in `input2`. It ranges from 0 (no overlap) to 1 (perfect overlap).
Return type: float

Notes

This is a real metric.

xcp_d.utils.qcmetrics.jc(input1, input2)[source]

Calculate Jaccard coefficient between two images.

Computes the Jaccard coefficient between the binary objects in two images.

Parameters: input1/input2 (str) – Path to a NIFTI image. Can be any type but will be converted into binary: False where 0, True everywhere else.
Returns: jc – The Jaccard coefficient between the object(s) in input1 and the object(s) in input2. It ranges from 0 (no overlap) to 1 (perfect overlap).
Return type: float

Notes

This is a real metric.