Preprocessing

Preprocessing tools to modify signals before reconstruction.

Filtering

You need to filter signals before reconstruction. The bandpass filter helps to clip frequencies that you want to work with. The cutoff frequencies can be passed to reconstruction class as properties. Then the filtering will be applied inside the reconstruction module.

preprocessing.filterBandPass.sigMatFilter(sigMat, lowCutOff=500000.0, highCutOff=6000000.0, fSampling=40000000.0, fOrder=3, conRatio=0.5)

Band pass filtering for signals

Parameters

• sigMat – 3D array (samples x channels x repetition) of signals

• lowCutOff – Low cut off frequency of bandpass filter

• highCutOff – High cut off frequency of bandpass filter

• fSampling – Sampling frequency of signals

• fOrder – Butterworth filter order

• conRatio – Nyquist ratio percentage

Returns

3D filtered signal array

Normalization

Normalization of signals over each channel of transducer array helps to improve image quality by scaling the signal intensities at the same range. It is highly recommended to normalize signals using module provided in preprocessing tools. Reconstruction functions involve the normalization step inside. There is no need to apply it beforehand.

preprocessing.normalizeUS.sigMatNormalizeUS(sigMatIn)

Band pass filtering for signals

Parameters

sigMatIn – 3D array (samples x channels x repetition) of signals

Returns

3D array normalized around mean value of the channels