NIPY logo

Site Navigation

NIPY Community

Table Of Contents

Previous topic

core.image.image_spaces

This Page

core.image.roi

Module: core.image.roi

Inheritance diagram for nipy.core.image.roi:

Template region of interest (ROI) module

Classes

ContinuousROI

class nipy.core.image.roi.ContinuousROI(coordinate_system, bfn, args=None, ndim=3)

Bases: nipy.core.image.roi.ROI

Create an ROI with a binary function in a given coordinate system.

Methods

todiscrete
__init__(coordinate_system, bfn, args=None, ndim=3)

Initialize continuous ROI instance

Parameters :

coordinate_system : CoordinateSystem instance

TODO

bfn : callable

binary function accepting real-value points as input, and any args in args, returning 1 at points inside the ROI and 0 for points outside the ROI.

args : sequence

arguments to be passed to bfn other then real-valued points

ndim : int

number of dimensions.

todiscrete(voxels)

Return a DiscreteROI instance at the voxels in the ROI.

Parameters :

voxels : array shape (N, 3)

voxel points in real space

Returns :

droi : DiscreteROI instance

discrete ROI where roi defined by voxels inside self

CoordinateMapROI

class nipy.core.image.roi.CoordinateMapROI(coordinate_system, voxels, coordmap)

Bases: nipy.core.image.roi.DiscreteROI

Methods

feature
mask
next
pool
__init__(coordinate_system, voxels, coordmap)

Initialize coordinate map ROI instance

Parameters :

coordinate_system : TODO

TODO

voxels : TODO

TODO

coordmap : TODO

TODO

feature(fn, **extra)

Return a feature of an image within the ROI.

Take the mean of voxel (point) features in ROI.

Parameters :

fn : callable

accepts point and kwargs **extra, returns value for that point (see pool method)

**extra : kwargs

keyword arguments to pass to fn

Returns :

val : object

result of np.mean when applied to the values output from fn

mask(img)

Return image with ones within ROI, zeros elsewhere

Returns :``numpy.ndarray`
next()

Return next point in ROI

pool(image)

Pool data from an image over the ROI

Return image value for each voxel in ROI

Parameters :

image : image.Image

or something with a get_data method

Returns :

vals : list

values in image at voxel points given by self.voxels

Raises :

ValueError: if coordinate maps of image and ROI do not match :

DiscreteROI

class nipy.core.image.roi.DiscreteROI(coordinate_system, voxels)

Bases: nipy.core.image.roi.ROI

ROI defined from discrete points

Methods

feature
next
pool
__init__(coordinate_system, voxels)

Initialize discrete ROI

Parameters :

coordinate_system : TODO

TODO

voxels : sequence

feature(fn, **extra)

Return a feature of an image within the ROI.

Take the mean of voxel (point) features in ROI.

Parameters :

fn : callable

accepts point and kwargs **extra, returns value for that point (see pool method)

**extra : kwargs

keyword arguments to pass to fn

Returns :

val : object

result of np.mean when applied to the values output from fn

next()

Return next point in ROI

pool(fn, **extra)

Pool data from an image over the ROI – return fn evaluated at each voxel.

Parameters :

fn : callable

function to apply to each voxel

**extras : kwargs

keyword arguments to pass to fn

Returns :

proc_pts : list

result of fn applied to each point within ROI

ROI

class nipy.core.image.roi.ROI(coordinate_system)

Bases: object

This is the basic ROI class, which we model as basically a function defined on Euclidean space, i.e. R^3. For practical purposes, this function is evaluated on the range of a Mapping instance.

__init__(coordinate_system)

Initialize ROI instance

Parameters :coordinate_system : CoordinateSystem instance

Functions

nipy.core.image.roi.roi_ellipse_fn(center, form, a=1.0)

Ellipse determined by regions where a quadratic form is <= a. The quadratic form is given by the inverse of the ‘form’ argument, so a sphere of radius 10 can be specified as {‘form’:10**2 * identity(3), ‘a’:1} or {‘form’:identity(3), ‘a’:100}.

Form must be positive definite.

Parameters :

form : TODO

TODO

a : float

TODO

Returns :

ellipse_fn : function

binary function of point, returning True if point is within ellipse, False otherwise

nipy.core.image.roi.roi_from_array_sampling_coordmap(data, coordmap)

Return a CoordinateMapROI from an array (data) on a coordmap. interpolation. Obvious ways to extend this.

Parameters :

data : array

Non-zero values in data define points in ROI

coordmap : CoordinateMap instance

coordinate map defining relationship of ijk(etc) indices in data and point space

Returns :

cm_roi : CoordinateMapROI

nipy.core.image.roi.roi_sphere_fn(center, radius)

Binary function for sphere with center and radius

Parameters center : sequence

real coordinates point for sphere center
radius : float
sphere radius
Returns :

sph_fn : function

binary function accepting points as input, return True if point is within sphere, False otherwise