NaviNIBS.Navigator.Model.SubjectRegistration module

class NaviNIBS.Navigator.Model.SubjectRegistration.Fiducial(key, plannedCoord=None, sampledCoords=None, sampledCoord=None, alignmentWeight=1.0)

Bases: GenericCollectionDictItem[str]

Parameters:

• _key

• _plannedCoord – Note that these (planned) coordinates are in MRI space.

• _sampledCoords – Note that these (sampled) coordinates are in tracker space, not in MRI space. This is to not require modification when using other methods besides fiducials for defining tracker->MRI coordinate transform. Is of size Nx3, where N is the number of repeated samples of the same coordinate.

• _sampledCoord – Is of size (3,). Can be used without sampledCoords to specify single sampled coordinate. Or can be used with sampledCoords to override the default sampledCoords->sampledCoord aggregation (a simple mean) with some other result (e.g. a trimmed mean) while still preserving all original samples. Note that this is cleared whenever sampledCoords is set, to avoid persisting an out-of-date aggregated sample.

• _alignmentWeight – Can be used to specify a relative positive weight for how this fiducial will be incorporated into registration. For example, if the NAS fiducial is given weight 100, and the LPA and RPA fiducials have default weights of 1, then alignment will always very closely match measured NAS to planned NAS location, and allow larger deviations for LPA and RPA for the sake of closely matching NAS.

_plannedCoord: tp.Optional[np.ndarray]

Note that these (planned) coordinates are in MRI space.

_sampledCoords: tp.Optional[np.ndarray]

Note that these (sampled) coordinates are in tracker space, not in MRI space. This is to not require modification when using other methods besides fiducials for defining tracker->MRI coordinate transform.

Is of size Nx3, where N is the number of repeated samples of the same coordinate.

_sampledCoord: tp.Optional[np.ndarray]

Is of size (3,).

Can be used without sampledCoords to specify single sampled coordinate. Or can be used with sampledCoords to override the default sampledCoords->sampledCoord aggregation (a simple mean) with some other result (e.g. a trimmed mean) while still preserving all original samples.

Note that this is cleared whenever sampledCoords is set, to avoid persisting an out-of-date aggregated sample.

_alignmentWeight: float

Can be used to specify a relative positive weight for how this fiducial will be incorporated into registration.

For example, if the NAS fiducial is given weight 100, and the LPA and RPA fiducials have default weights of 1, then alignment will always very closely match measured NAS to planned NAS location, and allow larger deviations for LPA and RPA for the sake of closely matching NAS.

property plannedCoord

property sampledCoord

Always returns single vector coordinate of size (3,), or None if the fiducial has not been sampled. If multiple repeats of the same fiducial are sampled (in sampledCoords), and no separate sampledCoord has been set, this returns an average of the samples.

property sampledCoords

property alignmentWeight

asDict()

Return type:

dict[str, Any]

classmethod fromDict(d)

class NaviNIBS.Navigator.Model.SubjectRegistration.Fiducials(items=NOTHING)

Bases: GenericCollection[str, Fiducial]

Parameters:

_items

property plannedFiducials: dict[str, ndarray | None]

Return type:

dict[str, Optional[ndarray]]

property sampledFiducials: dict[str, ndarray | None]

Note that this returns a dict where each value is a sampledCoord (always None or shape (3,)), even if multiple samples exist for a given fiducial. For full access, use allSampledFiducials instead.

Return type:

dict[str, Optional[ndarray]]

property allSampledFiducials: dict[str, ndarray | None]

Return type:

dict[str, Optional[ndarray]]

property alignmentWeights: ndarray

Return type:

ndarray

classmethod fromList(itemList)

Return type:

Fiducials

class NaviNIBS.Navigator.Model.SubjectRegistration.HeadPoints(headPoints=NOTHING, alignmentWeights=None)

Bases: object

Parameters:

• _headPoints – Note that these are coordinates are in tracker space, not in MRI space.

• _alignmentWeights – Optional weights used for headpoint-based registration refinement, in format expected by simpleicp’s rbp_observation_weights argument (i.e. rot_x, rot_y, rot_z, t_x, t_y, t_z). Can alternatively specify as a single scalar to apply the same weight to all terms, or as two scalars to apply the first weight to all rotation term, the second weights to all translation terms. Note that values will be inverted (weightArg = 1 / weight) so that higher input values correspond to greater weighting of head points, rather than greater weighting of fiducials. Note that these weights are different in definition and usage than Fiducial.alignmentWeight.

_headPoints: list[ndarray]

Note that these are coordinates are in tracker space, not in MRI space.

_alignmentWeights: ndarray | None

Optional weights used for headpoint-based registration refinement, in format expected by simpleicp’s rbp_observation_weights argument (i.e. rot_x, rot_y, rot_z, t_x, t_y, t_z). Can alternatively specify as a single scalar to apply the same weight to all terms, or as two scalars to apply the first weight to all rotation term, the second weights to all translation terms.

Note that values will be inverted (weightArg = 1 / weight) so that higher input values correspond to greater weighting of head points, rather than greater weighting of fiducials.

Note that these weights are different in definition and usage than Fiducial.alignmentWeight.

sigHeadpointsAboutToChange: Signal

This signal includes list of indices of headpoints about to change.

sigHeadpointsChanged: Signal

” This signal includes list of indices of headpoints that changed.

sigAttribsAboutToChange: Signal

This signal includes keys and incoming values of attributes (besides main headPoints list) about to change.

sigAttribsChanged: Signal

This signal includes keys of attributes (besides main headPoints list) that changed.

property alignmentWeights

append(point)

extend(points)

remove(indexOrIndices)

replace(points)

Clear all previous points, set to new points.

clear()

asNDArray()

asList()

classmethod fromList(l)

class NaviNIBS.Navigator.Model.SubjectRegistration.SubjectRegistration(fiducials=NOTHING, sampledHeadPoints=NOTHING, trackerToMRITransf=None, fiducialsHistory=NOTHING, trackerToMRITransfHistory=NOTHING)

Bases: object

Parameters:

• _fiducials

• _sampledHeadPoints

• _trackerToMRITransf

• _fiducialsHistory

• _trackerToMRITransfHistory

_fiducials: Fiducials

_sampledHeadPoints: HeadPoints

_trackerToMRITransf: ndarray | None

_fiducialsHistory: Dict[str, Fiducials]

_trackerToMRITransfHistory: Dict[str, ndarray | None]

sigTrackerToMRITransfAboutToChange: Signal

sigTrackerToMRITransfChanged: Signal

property fiducials

property fiducialsHistory

Result should not be modified

property hasMinimumPlannedFiducials: bool

Return type:

bool

property hasMinimumSampledFiducials: bool

Return type:

bool

property sampledHeadPoints: HeadPoints

Return type:

HeadPoints

property trackerToMRITransf

property trackerToMRITransfHistory

Result should not be modified

property isRegistered

property approxHeadCenter: ndarray | None

Return type:

Optional[ndarray]

asDict()

Return type:

Dict[str, Any]

classmethod fromDict(d)

Return type:

SubjectRegistration

static getDatetimeFromTimestamp(ts)

Return type:

datetime

static fiducialsEqual(fidsA, fidsB)

Return type:

bool