StrokeSegLab.preprocessing.resampling.Resampler Class Reference

Public Member Functions
None	__init__ (self)
np.ndarray	run (self, np.ndarray data, tuple[float, float, float] current_spacing, tuple[float, float, float] new_spacing)

Public Attributes
float	separate_z_anisotropy_threshold = 3.0
	force_separate_z = None
bool	is_seg = False
int	order = 1
int	order_z = 0
np.ndarray	force_separate_z = self._get_lowres_axis(current_spacing)

Protected Member Functions
tuple[bool, int]	_determine_do_sep_z_and_axis (self, tuple[float, float, float] current_spacing, tuple[float, float, float] new_spacing)
np.ndarray	_get_lowres_axis (self, tuple[float, float, float] new_spacing)
bool	_get_do_separate_z (self, tuple[float, float, float] spacing)
np.ndarray	_compute_new_shape (self, np.ndarray old_shape, tuple[float, float, float] old_spacing, tuple[float, float, float] new_spacing)
np.ndarray	_resample_data_or_seg (self, np.ndarray data, np.ndarray new_shape, int axis, bool do_separate_z, np.dtype dtype_out=None)

Detailed Description

This class handle the resampling during the preprocessing and postprocessing

Constructor & Destructor Documentation

__init__()

None StrokeSegLab.preprocessing.resampling.Resampler.__init__

(

self

)

Initialize the resampler class

Member Function Documentation

_compute_new_shape()

np.ndarray StrokeSegLab.preprocessing.resampling.Resampler._compute_new_shape ( self, np.ndarray old_shape, tuple[float,float,float] old_spacing, tuple[float,float,float] new_spacing )

protected

@public
Computes the new shape of data after resampling to a new spacing
Scales each dimension according to the ratio of old and new spacing values

Args:
    old_shape (np.ndarray): Current shape of the data
    old_spacing (tuple[float,float,float]): Current spacing in mm (sx, sy, sz) 
    new_spacing (tuple[float,float,float]): Desired spacing in mm (sx, sy, sz)

Returns:
    np.ndarray: Future shape of the resempled data

_determine_do_sep_z_and_axis()

tuple[bool,int] StrokeSegLab.preprocessing.resampling.Resampler._determine_do_sep_z_and_axis ( self, tuple[float,float,float] current_spacing, tuple[float,float,float] new_spacing )

protected

@public
Decides if resampling should be done separately along the z-axis and finds the right axis
The choice is based on the force_separate_z attribute and the spacing values. If force_separate_z is set, it is used first. If not, the method checks if the current or new spacing needs separate resampling in the z direction

Args:
    current_spacing (tuple[float,float,float]): Current spacing in mm (sx, sy, sz)
    new_spacing (tuple[float,float,float]): Desired spacing in mm (sx, sy, sz)

Returns:
    tuple[bool,int]:
    - True if separate resampling along the z-axis should be done
    - The axis to resample separately, or None if not needed

_get_do_separate_z()

bool StrokeSegLab.preprocessing.resampling.Resampler._get_do_separate_z ( self, tuple[float,float,float] spacing )

protected

@public
 It compares the ratio between the max and min spacing values to a threshold (separate_z_anisotropy_threshold)

Args:
    spacing (tuple[float,float,float]): Spacing in mm (sx, sy, sz)

Returns:
    bool: True if the ratio is higher than the threshold

_get_lowres_axis()

np.ndarray StrokeSegLab.preprocessing.resampling.Resampler._get_lowres_axis ( self, tuple[float,float,float] new_spacing )

protected

@public
Finds which axis has the biggest spacing (=the lowest resolution)

Args:
    new_spacing (tuple[float,float,float]): Spacing in mm (sx, sy, sz)

Returns:
    np.ndarray: The indices of the axis with the biggest spacing

_resample_data_or_seg()

np.ndarray StrokeSegLab.preprocessing.resampling.Resampler._resample_data_or_seg ( self, np.ndarray data, np.ndarray new_shape, int axis, bool do_separate_z, np.dtype dtype_out = None )

protected

@public
Resample image or segmentation data to a new shape, optionally resampling separately along an anisotropic axis

Args:
    data (np.ndarray): Input array of shape (c, x, y, z)
    new_shape (np.ndarray): Future shape of the resempled data
    axis (int): The axis to resample separately, or None if not needed
    do_separate_z (bool): True if separate resampling along the z-axis should be done
    dtype_out (np.dtype, optional): Output data type. Defaults to None.

Returns:
    np.ndarray: Output data array of shape (c, x, y, z)

run()

np.ndarray StrokeSegLab.preprocessing.resampling.Resampler.run	(		self,
		np.ndarray	data,
		tuple[float,float,float]	current_spacing,
		tuple[float,float,float]	new_spacing )

Resamples the input data to a new spacing

Args:
    data (np.ndarray): Input array of shape (c, x, y, z)
    current_spacing (tuple[float,float,float]): Current spacing in mm (sx, sy, sz)
    new_spacing (tuple[float,float,float]): Desired spacing in mm (sx, sy, sz)

Returns:
    np.ndarray: The resampled data 

---

The documentation for this class was generated from the following file:

• /home/ymahe/Desktop/StrokeSeg/StrokeSegLab/preprocessing/resampling.py