how to create a SLR pulse by pypulseq

[qizixinge]

I want to replace original sinc pulse by SLR pulse, but I don't know how to. I made 2 attempts on write_gre.py, but neither was successful.

So how can I get SLR pulse with 3mm slice thickness by code? It would be even better if apodization=0.42 could be set

[btasdelen]

Hi @qizixinge. If you want to use sigpy_n_seq function of pypulseq, here is an example code from the RF pulse test:

    time_bw_product = 4
    slice_thickness = 3e-3  # Slice thickness
    flip_angle = np.pi / 2
    # Set system limits
    system = Opts(
        max_grad=32,
        grad_unit='mT/m',
        max_slew=130,
        slew_unit='T/m/s',
        rf_ringdown_time=30e-6,
        rf_dead_time=100e-6,
    )
    pulse_cfg = SigpyPulseOpts(
        pulse_type='slr',
        ptype='st',
        ftype='ls',
        d1=0.01,
        d2=0.01,
        cancel_alpha_phs=False,
        n_bands=3,
        band_sep=20,
        phs_0_pt='None',
    )
    rfp, gz, gzr = sigpy_n_seq(
        flip_angle=flip_angle,
        system=system,
        duration=3e-3,
        slice_thickness=slice_thickness,
        time_bw_product=4,
        return_gz=True,
        pulse_cfg=pulse_cfg,
        plot=True,
    )

    seq = pp.Sequence()
    seq.add_block(rfp)

There is no apodization input to this function. The documentation for this function is incorrect, I will update it soon.

If this function is not sufficient for your use case (e.g. you really need apodization), you can design your pulse externally as a waveform, and use make_arbitrary_rf function instead. Here is an SLR example:

Trfmin       = 2e-3          # duration of one pulse segment [sec]
slice_thickness = 4e-3 # [m]
Nz = 20

n = ceil((Trfmin/dt)/4)*4
Trf = n*dt
tb = 8
bw = tb/Trf

signal = slr.dzrf(n=n, tb=tb, ptype='st', ftype='ls', d1=0.01, d2=0.01, cancel_alpha_phs=False)

# Create an RF event
rf_res, gz = make_arbitrary_rf(
    signal=signal, slice_thickness=slice_thickness*Nz,
    bandwidth=bw, flip_angle=pi/3, 
    system=system, return_gz=True, use="excitation"
    )

[btasdelen]

You seem to be using an old version of PyPulseq. Please update to the latest version from PyPI and try again.

[qizixinge]

writegreSLR.txt
Thank you. But how can i update to the latest version from PyPI when running google colab? I tried "pip install git+https://github.com/imr-framework/pypulseq@dev", but still don't work. My original code on Google Colab is in the above txt file

[qizixinge]

And for the earliest version I used "pip install pypulseq"

[btasdelen]

Okay, I found the issue, but this is not an issue with pypulseq, but probably an issue with sigpy. Setting time_bw_product=1.5 causes a negative value in a call to firls function, which, as stated in the error, only supports values between 0 and 1. In fact, you can only set the time_bw_product to 2 or higher.

I don't know enough RF pulse design to say not having TBW less than 2 for SLR pulses is reasonable or not. I think you would have a better luck discussing this over sigpy. Overall, your code works if you set time_bw_product=2.