compressShape

Compresses a gradient or pulse waveform using a run-length encoding scheme applied to its derivative. This efficiently represents waveforms with constant or linearly changing segments using significantly fewer samples. The function returns a structure containing the compressed waveform and the original number of samples.

Syntax

function s=compressShape(w, forceCompression)

Calling Pattern

mr.compressShape(...)

Parameters

Required Parameters

Name	Type	Description	Example	Units
w	double	The input waveform (gradient or pulse shape) to be compressed. This is a vector of waveform amplitude values.	[1, 1, 1, 2, 2, 3, 3, 3, 3]

Optional Parameters

Name	Type	Default	Description	Example
forceCompression	logical	false	A flag indicating whether to force compression even if the input waveform is short. If false, waveforms with 4 or fewer samples are returned uncompressed. Valid values: true, false	true

Returns

Output	Type	Description
s	struct	A structure containing the compressed waveform data. The structure has two fields:
- num_samples: The number of samples in the original, uncompressed waveform.
- data: A vector containing the compressed waveform data in a format suitable for Pulseq.

Examples

% Compress gradient waveform for storage efficiency
compressed = mr.compressShape(gradientWaveform);

% Force compression of short waveforms
compressed = mr.compressShape(shortWaveform, true);

% Compress RF magnitude data for sequence storage
mag_compressed = mr.compressShape(rfMagnitude);

See Also

decompressShape