Use Koma in Notebooks

Run the same brain/EPI example as the UI and Julia script guides in Pluto or Jupyter.

Pluto

First, install the Pluto module in your Julia environment. Remember to press the ] button to open the Package Manager Session:

julia> ]

(@v1.13) pkg> add Pluto

Afterward, return to the Julia Session by pressing the backspace button, and then execute the Pluto.run() function:

julia> using Pluto

julia> Pluto.run()

This should automatically open the Pluto dashboard in your default web browser:

Next, create a new notebook by clicking on + Create a new notebook:

Run the basic workflow below, with each code block in its own cell. Pluto installs the required packages and stores their environment in the notebook.

Jupyter

Ensure you have Jupyter installed on your computer. Follow this tutorial for installation using Anaconda.

Next, install the IJulia module in your Julia environment. Remember to press the ] key to open the Package Manager Session:

julia>

(@v1.13) pkg> add IJulia

Install the packages used by the workflow:

(@v1.13) pkg> add KomaMRI MAT

Next, open Jupyter, navigate to a working folder, and create a new notebook by clicking on New, then selecting your installed Julia kernel (e.g. Julia 1.13.1).

A new, empty notebook will appear:

Run the basic workflow below, with each code block in its own cell. Plotly figures use the notebook's standard HTML display; no WebIO extension is needed.

Basic Workflow

Loading Simulation Inputs

Use the default scanner, brain phantom, and EPI sequence. Zero off-resonance to match the UI's default phantom:

using KomaMRI, MAT

sys = Scanner()

begin
    obj = brain_phantom2D()
    obj.Δw .= 0
end

seq = PulseDesigner.EPI_example(; sys)

Display the phantom, sequence, and trajectory in separate cells:

plot_phantom_map(obj, :ρ)

plot_seq(seq)

plot_kspace(seq)

Running Simulation

sim_params = KomaMRICore.default_sim_params()

raw = simulate(obj, seq, sys; sim_params)

plot_signal(raw)

Reconstructing Image using MRIReco

rec_params = Dict{Symbol,Any}(:reco => "direct")

result = reconstruct_with_labels(raw; rec_params)

begin
    Nx, Ny = raw.params["reconSize"][1:2]
    image = reshape(first(result.images).image, Nx, Ny)
    plot_image(image * (Nx * Ny))
end

Exporting Results

begin
    save(ISMRMRDFile(joinpath(tempdir(), "data_raw.mrd")), raw)
    MAT.matwrite(joinpath(tempdir(), "data_image.mat"), Dict("image" => image))
end

For the same optional sphere and spiral examples, run the code in Phantom and Sequence in separate cells. These use sphere and spiral, leaving the brain/EPI inputs unchanged.