cMRF

Dockerfile

@@ -5,24 +5,23 @@ FROM nvidia/cuda:${CUDA_VERSION}-base-ubuntu${UBUNTU_VERSION}
 
 # Python
 RUN apt-get update && apt-get install --no-install-recommends -y \
-        python3-pip && \
-    apt-get clean && rm -rf /var/lib/apt/lists/*
+        python3-pip git && \
+        apt-get clean && rm -rf /var/lib/apt/lists/*
 
 # Torch
 ARG TORCH_VERSION=2.5
 ARG TORCH_VARIANT=cpu
 RUN python3 -m pip --no-cache-dir install \
         --index-url https://download.pytorch.org/whl/${TORCH_VARIANT}\
         torch==${TORCH_VERSION}
-       
+
 # MRpro    
 ARG MRPRO_VERSION=0.250107
-RUN python3 -m pip --no-cache-dir install \
-        mrpro==${MRPRO_VERSION}
+RUN python3 -m pip --no-cache-dir install git+https://github.com/PTB-MR/mrpro@cMRF
 
 # MRpro Server
 WORKDIR /home
-COPY *.py /home/mrpro_server/
+COPY * /home/mrpro_server/
 CMD ["python3", "/home/mrpro_server/server.py"]
 
 # Metadata

mrpro_server/combine_mrd_header.py

@@ -0,0 +1,152 @@
+"""Functions to combine ismrmrd data files and meta files."""
+
+from pathlib import Path
+
+import ismrmrd
+
+
+def read_ismrmrd_dataset(fname: Path) -> tuple:
+    """Read ismrmrd dataset from file.
+
+    Parameters
+    ----------
+    fname
+        file path to the ismrmrd dataset.
+
+    Returns
+    -------
+        ismrmrd header and list of acquisitions.
+    """
+    with ismrmrd.File(str(fname), "r") as file:
+        ds = file[list(file.keys())[-1]]
+        header = ds.header
+        acqs = ds.acquisitions[:]
+
+    return header, acqs
+
+
+def insert_traj_from_meta(
+    data_acqs: list[ismrmrd.acquisition.Acquisition],
+    meta_acqs: list[ismrmrd.acquisition.Acquisition],
+) -> list[ismrmrd.acquisition.Acquisition]:
+    """
+    Insert trajectory information from the meta file into the data file.
+
+    Parameters
+    ----------
+    data_acqs : list
+        list of acquisitions from the data file.
+    meta_acqs : list
+        list of acquisitions from the meta file.
+
+    Returns
+    -------
+    list of acquisitions with the trajectory information from the meta file.
+    """
+    if not len(data_acqs) == len(meta_acqs):
+        raise ValueError("Number of acquisitions in data and meta file do not match.")
+
+    for i, (acq_d, acq_m) in enumerate(zip(data_acqs, meta_acqs, strict=False)):
+        if not acq_d.number_of_samples == acq_m.number_of_samples:
+            raise ValueError(f"Number of samples in acquisition {i} do not match.")
+
+        # insert trajectory information from meta file
+        acq_d.resize(
+            number_of_samples=acq_d.number_of_samples,
+            active_channels=acq_d.active_channels,
+            trajectory_dimensions=acq_m.trajectory_dimensions,
+        )
+        acq_d.traj[:] = acq_m.traj[:]
+        data_acqs[i] = acq_d
+
+    return data_acqs
+
+
+def update_header_from_meta(
+    data_header: ismrmrd.xsd.ismrmrdHeader,
+    meta_header: ismrmrd.xsd.ismrmrdHeader,
+    enc_idx: int = 0,
+) -> ismrmrd.xsd.ismrmrdHeader:
+    """Update the header of the data file with the information from the meta file.
+
+    Parameters
+    ----------
+    data_header : ismrmrd.xsd.ismrmrdHeader
+        Header of the ISMRMRD data file.
+    meta_header : ismrmrd.xsd.ismrmrdHeader
+        Header of the ISMRMRD meta file created with the seq-file.
+    enc_idx : int, optional
+        Encoding index, by default 0
+
+    Returns
+    -------
+    ismrmrd.xsd.ismrmrdHeader
+        Updated header.
+    """
+
+    # Helper function to copy attributes if they are not None
+    def copy_attributes(source, target, attributes):
+        for attr in attributes:
+            value = getattr(source, attr, None)
+            if value is not None:
+                setattr(target, attr, value)
+
+    # Define the attributes to update for encodedSpace and reconSpace
+    attributes_to_update = ["matrixSize", "fieldOfView_mm"]
+
+    # Update encodedSpace
+    copy_attributes(
+        meta_header.encoding[enc_idx].encodedSpace,
+        data_header.encoding[enc_idx].encodedSpace,
+        attributes_to_update,
+    )
+
+    # Update reconSpace
+    copy_attributes(
+        meta_header.encoding[enc_idx].reconSpace,
+        data_header.encoding[enc_idx].reconSpace,
+        attributes_to_update,
+    )
+
+    # Update trajectory type
+    if meta_header.encoding[enc_idx].trajectory is not None:
+        data_header.encoding[enc_idx].trajectory = meta_header.encoding[enc_idx].trajectory
+
+    return data_header
+
+
+def combine_ismrmrd_files(data_file: Path, meta_file: Path, filename_ext: str = "_with_traj.h5") -> ismrmrd.Dataset:
+    """Combine ismrmrd data file and meta file.
+
+    Parameters
+    ----------
+    data_file
+        path to the ismrmrd data file
+    meta_file
+        path to the ismrmrd meta file
+    filename_ext, optional
+        filename extension of the output file, by default '_with_traj.h5'
+
+    Returns
+    -------
+        combined ismrmrd file from data and meta file.
+    """
+    filename_out = data_file.parent / (data_file.stem + filename_ext)
+
+    data_header, data_acqs = read_ismrmrd_dataset(data_file)
+    meta_header, meta_acqs = read_ismrmrd_dataset(meta_file)
+
+    new_acqs = insert_traj_from_meta(data_acqs, meta_acqs)
+    new_header = update_header_from_meta(data_header, meta_header)
+
+    # Create new file
+    ds = ismrmrd.Dataset(filename_out)
+    ds.write_xml_header(new_header.toXML())
+
+    # add acquisitions with trajectory information
+    for acq in new_acqs:
+        ds.append_acquisition(acq)
+
+    ds.close()
+
+    return ds