Class: Resampler

Defined in: src/resampling/Resampler.ts:21

Resampling implementation for NeuroVol

Implements

• IResampler

Constructors

Constructor

new Resampler(volume): Resampler;

Defined in: src/resampling/Resampler.ts:24

Parameters

volume

NeuroVol

Returns

Resampler

Methods

resample()

resample(targetSpace, options?): NeuroVol;

Defined in: src/resampling/Resampler.ts:31

Resample volume to new space/resolution

Parameters

targetSpace

NeuroSpace

options?

ResampleOptions = {}

Returns

NeuroVol

Implementation of

IResampler.resample

---

resampleToDimensions()

resampleToDimensions(dimensions, options?): NeuroVol;

Defined in: src/resampling/Resampler.ts:79

Resample to specific voxel dimensions

Parameters

dimensions

[number, number, number]

options?

ResampleOptions

Returns

NeuroVol

Implementation of

IResampler.resampleToDimensions

---

resampleToVoxelSize()

resampleToVoxelSize(voxelSize, options?): NeuroVol;

Defined in: src/resampling/Resampler.ts:103

Resample to specific voxel size

Parameters

voxelSize

[number, number, number]

options?

ResampleOptions

Returns

NeuroVol

Implementation of

IResampler.resampleToVoxelSize

---

transform()

transform(options, resampleOptions?): NeuroVol;

Defined in: src/resampling/Resampler.ts:139

Apply an affine transformation to the volume, resampling onto a grid of the same dimensions/spacing as the input.

The transform is expressed in voxel (grid) coordinates. The forward map sends an input voxel p to an output voxel q via

q = M * (p - center) + center + translation

where M = R * S (rotation composed with scale). Resampling fills each OUTPUT voxel q by mapping it back into the INPUT volume with the inverse transform and sampling there:

p = M^-1 * (q - center - translation) + center

Parameters

options

TransformOptions

resampleOptions?

ResampleOptions = {}

Returns

NeuroVol

Implementation of

IResampler.transform

---

interpolateAt()

interpolateAt(
   x, 
   y, 
   z, 
   method?): number;

Defined in: src/resampling/Resampler.ts:180

Interpolate value at continuous coordinate

Parameters

x

number

y

number

z

number

method?

InterpolationMethod = 'linear'

Returns

number

Implementation of

IResampler.interpolateAt

---

downsample()

downsample(factor, options?): NeuroVol;

Defined in: src/resampling/Resampler.ts:187

Downsample by integer factor

Parameters

factor

number

options?

ResampleOptions

Returns

NeuroVol

Implementation of

IResampler.downsample

---

upsample()

upsample(factor, options?): NeuroVol;

Defined in: src/resampling/Resampler.ts:204

Upsample by integer factor

Parameters

factor

number

options?

ResampleOptions

Returns

NeuroVol

Implementation of

IResampler.upsample

---

resampleClustered()

static resampleClustered(source, targetSpace): ClusteredNeuroVol;

Defined in: src/resampling/Resampler.ts:225

Resample a ClusteredNeuroVol to a new target space.

Uses nearest-neighbour interpolation to preserve integer labels. The algorithm:

1. Resample labels (dense Int32 representation) with nearest-neighbour
2. Resample the mask with nearest-neighbour
3. Keep voxels where both resampled mask and label are non-zero
4. Build a new ClusteredNeuroVol with the surviving labels

Parameters

source

ClusteredNeuroVol

targetSpace

NeuroSpace

Returns

ClusteredNeuroVol