Source code for neuroconv.tools.roiextractors.roiextractors

import math
from collections import defaultdict
from copy import deepcopy
from typing import Literal, Optional
from warnings import warn

import numpy as np
import psutil
from hdmf.backends.hdf5.h5_utils import H5DataIO

# from hdmf.common import VectorData
from hdmf.data_utils import DataChunkIterator
from pynwb import NWBFile
from pynwb.base import Images
from pynwb.device import Device
from pynwb.image import GrayscaleImage
from pynwb.ophys import (
    DfOverF,
    Fluorescence,
    ImageSegmentation,
    ImagingPlane,
    OnePhotonSeries,
    OpticalChannel,
    PlaneSegmentation,
    RoiResponseSeries,
    TwoPhotonSeries,
)
from roiextractors import (
    ImagingExtractor,
    MultiSegmentationExtractor,
    SegmentationExtractor,
)

from .imagingextractordatachunkiterator import ImagingExtractorDataChunkIterator
from ..hdmf import SliceableDataChunkIterator
from ..nwb_helpers import get_default_nwbfile_metadata, get_module, make_or_load_nwbfile
from ...utils import (
    DeepDict,
    OptionalFilePathType,
    calculate_regular_series_rate,
    dict_deep_update,
)


[docs]def get_default_ophys_metadata() -> DeepDict: """Fill default metadata for Device and ImagingPlane.""" metadata = get_default_nwbfile_metadata() default_device = dict(name="Microscope") default_optical_channel = dict( name="OpticalChannel", emission_lambda=np.nan, description="An optical channel of the microscope.", ) default_imaging_plane = dict( name="ImagingPlane", description="The plane or volume being imaged by the microscope.", excitation_lambda=np.nan, indicator="unknown", location="unknown", device=default_device["name"], optical_channel=[default_optical_channel], ) metadata.update( Ophys=dict( Device=[default_device], ImagingPlane=[default_imaging_plane], ), ) return metadata
[docs]def get_default_segmentation_metadata() -> DeepDict: """Fill default metadata for segmentation.""" metadata = get_default_ophys_metadata() default_fluorescence_roi_response_series = dict( name="RoiResponseSeries", description="Array of raw fluorescence traces.", unit="n.a." ) default_fluorescence = dict( name="Fluorescence", PlaneSegmentation=dict( raw=default_fluorescence_roi_response_series, ), ) default_dff_roi_response_series = dict( name="RoiResponseSeries", description="Array of df/F traces.", unit="n.a.", ) default_df_over_f = dict( name="DfOverF", PlaneSegmentation=dict( dff=default_dff_roi_response_series, ), ) default_image_segmentation = dict( name="ImageSegmentation", plane_segmentations=[ dict( name="PlaneSegmentation", description="Segmented ROIs", imaging_plane=metadata["Ophys"]["ImagingPlane"][0]["name"], ) ], ) default_segmentation_images = dict( name="SegmentationImages", description="The summary images of the segmentation.", PlaneSegmentation=dict( correlation=dict(name="correlation", description="The correlation image."), ), ) metadata["Ophys"].update( dict( Fluorescence=default_fluorescence, DfOverF=default_df_over_f, ImageSegmentation=default_image_segmentation, SegmentationImages=default_segmentation_images, ), ) return metadata
[docs]def get_nwb_imaging_metadata( imgextractor: ImagingExtractor, photon_series_type: Literal["OnePhotonSeries", "TwoPhotonSeries"] = "TwoPhotonSeries", ) -> dict: """ Convert metadata from the ImagingExtractor into nwb specific metadata. Parameters ---------- imgextractor : ImagingExtractor photon_series_type : {'OnePhotonSeries', 'TwoPhotonSeries'}, optional """ metadata = get_default_ophys_metadata() channel_name_list = imgextractor.get_channel_names() or ( ["OpticalChannel"] if imgextractor.get_num_channels() == 1 else [f"OpticalChannel{idx}" for idx in range(imgextractor.get_num_channels())] ) imaging_plane = metadata["Ophys"]["ImagingPlane"][0] for index, channel_name in enumerate(channel_name_list): if index == 0: imaging_plane["optical_channel"][index]["name"] = channel_name else: imaging_plane["optical_channel"].append( dict( name=channel_name, emission_lambda=np.nan, description="An optical channel of the microscope.", ) ) one_photon_description = "Imaging data from one-photon excitation microscopy." two_photon_description = "Imaging data from two-photon excitation microscopy." photon_series_metadata = dict( name=photon_series_type, description=two_photon_description if photon_series_type == "TwoPhotonSeries" else one_photon_description, unit="n.a.", imaging_plane=imaging_plane["name"], dimension=list(imgextractor.get_image_size()), ) metadata["Ophys"].update({photon_series_type: [photon_series_metadata]}) return metadata
[docs]def add_devices(nwbfile: NWBFile, metadata: dict) -> NWBFile: """ Add optical physiology devices from metadata. The metadata concerning the optical physiology should be stored in metadata["Ophys]["Device"] This function handles both a text specification of the device to be built and an actual pynwb.Device object. """ metadata_copy = deepcopy(metadata) default_metadata = get_default_ophys_metadata() metadata_copy = dict_deep_update(default_metadata, metadata_copy, append_list=False) device_metadata = metadata_copy["Ophys"]["Device"] for device in device_metadata: device_name = device["name"] if isinstance(device, dict) else device.name if device_name not in nwbfile.devices: device = Device(**device) if isinstance(device, dict) else device nwbfile.add_device(device) return nwbfile
[docs]def _create_imaging_plane_from_metadata(nwbfile: NWBFile, imaging_plane_metadata: dict) -> ImagingPlane: """ Private auxiliary function to create an ImagingPlane object from pynwb using the imaging_plane_metadata. Parameters ---------- nwbfile : NWBFile An previously defined -in memory- NWBFile. imaging_plane_metadata : dict The metadata to create the ImagingPlane object. Returns ------- ImagingPlane The created ImagingPlane. """ device_name = imaging_plane_metadata["device"] imaging_plane_metadata["device"] = nwbfile.devices[device_name] imaging_plane_metadata["optical_channel"] = [ OpticalChannel(**metadata) for metadata in imaging_plane_metadata["optical_channel"] ] imaging_plane = ImagingPlane(**imaging_plane_metadata) return imaging_plane
[docs]def add_imaging_plane( nwbfile: NWBFile, metadata: dict, imaging_plane_name: Optional[str] = None, imaging_plane_index: Optional[int] = None, ) -> NWBFile: """ Adds the imaging plane specified by the metadata to the nwb file. The imaging plane that is added is the one located in metadata["Ophys"]["ImagingPlane"][imaging_plane_index] Parameters ---------- nwbfile : NWBFile An previously defined -in memory- NWBFile. metadata : dict The metadata in the nwb conversion tools format. imaging_plane_name: str The name of the imaging plane to be added. Returns ------- NWBFile The nwbfile passed as an input with the imaging plane added. """ if imaging_plane_index is not None: warn( message="Keyword argument 'imaging_plane_index' is deprecated and will be removed on or after Dec 1st, 2023. " "Use 'imaging_plane_name' to specify which imaging plane to add by its name.", category=DeprecationWarning, ) imaging_plane_name = metadata["Ophys"]["ImagingPlane"][imaging_plane_index]["name"] # Set the defaults and required infrastructure metadata_copy = deepcopy(metadata) default_metadata = get_default_ophys_metadata() metadata_copy = dict_deep_update(default_metadata, metadata_copy, append_list=False) add_devices(nwbfile=nwbfile, metadata=metadata_copy) default_imaging_plane_name = default_metadata["Ophys"]["ImagingPlane"][0]["name"] imaging_plane_name = imaging_plane_name or default_imaging_plane_name existing_imaging_planes = nwbfile.imaging_planes if imaging_plane_name not in existing_imaging_planes: imaging_plane_metadata = next( ( imaging_plane_metadata for imaging_plane_metadata in metadata_copy["Ophys"]["ImagingPlane"] if imaging_plane_metadata["name"] == imaging_plane_name ), None, ) if imaging_plane_metadata is None: raise ValueError( f"Metadata for Imaging Plane '{imaging_plane_name}' not found in metadata['Ophys']['ImagingPlane']." ) imaging_plane = _create_imaging_plane_from_metadata( nwbfile=nwbfile, imaging_plane_metadata=imaging_plane_metadata ) nwbfile.add_imaging_plane(imaging_plane) return nwbfile
[docs]def add_image_segmentation(nwbfile: NWBFile, metadata: dict) -> NWBFile: """ Adds the image segmentation specified by the metadata to the nwb file. Parameters ---------- nwbfile : NWBFile The nwbfile to add the image segmentation to. metadata: dict The metadata to create the image segmentation from. Returns ------- NWBFile The NWBFile passed as an input with the image segmentation added. """ # Set the defaults and required infrastructure metadata_copy = deepcopy(metadata) default_metadata = get_default_segmentation_metadata() metadata_copy = dict_deep_update(default_metadata, metadata_copy, append_list=False) image_segmentation_metadata = metadata_copy["Ophys"]["ImageSegmentation"] image_segmentation_name = image_segmentation_metadata["name"] ophys = get_module(nwbfile, "ophys") # Check if the image segmentation already exists in the NWB file if image_segmentation_name not in ophys.data_interfaces: ophys.add(ImageSegmentation(name=image_segmentation_name)) return nwbfile
[docs]def add_photon_series( imaging: ImagingExtractor, nwbfile: NWBFile, metadata: dict, photon_series_type: Literal["TwoPhotonSeries", "OnePhotonSeries"] = "TwoPhotonSeries", photon_series_index: int = 0, parent_container: Literal["acquisition", "processing/ophys"] = "acquisition", two_photon_series_index: Optional[int] = None, # TODO: to be removed iterator_type: Optional[str] = "v2", iterator_options: Optional[dict] = None, ) -> NWBFile: """ Auxiliary static method for nwbextractor. Adds photon series from ImagingExtractor to NWB file object. The photon series can be added to the NWB file either as a TwoPhotonSeries or OnePhotonSeries object. Parameters ---------- imaging : ImagingExtractor The imaging extractor to get the data from. nwbfile : NWBFile The nwbfile to add the photon series to. metadata: dict The metadata for the photon series. photon_series_type: {'OnePhotonSeries', 'TwoPhotonSeries'}, optional The type of photon series to add, default is TwoPhotonSeries. photon_series_index: int, default: 0 The metadata for the photon series is a list of the different photon series to add. Specify which element of the list with this parameter. parent_container: {'acquisition', 'processing/ophys'}, optional The container where the photon series is added, default is nwbfile.acquisition. When 'processing/ophys' is chosen, the photon series is added to ``nwbfile.processing['ophys']``. Returns ------- NWBFile The NWBFile passed as an input with the photon series added. """ if two_photon_series_index: warn( "Keyword argument 'two_photon_series_index' is deprecated and it will be removed on 2024-04-16. Use 'photon_series_index' instead." ) photon_series_index = two_photon_series_index iterator_options = iterator_options or dict() metadata_copy = deepcopy(metadata) assert photon_series_type in [ "OnePhotonSeries", "TwoPhotonSeries", ], "'photon_series_type' must be either 'OnePhotonSeries' or 'TwoPhotonSeries'." metadata_copy = dict_deep_update( get_nwb_imaging_metadata(imaging, photon_series_type=photon_series_type), metadata_copy, append_list=False ) if photon_series_type == "TwoPhotonSeries": assert ( "OnePhotonSeries" not in metadata_copy["Ophys"] ), "Received metadata for 'OnePhotonSeries' but `photon_series_type` was not explicitly specified." assert parent_container in [ "acquisition", "processing/ophys", ], "'parent_container' must be either 'acquisition' or 'processing/ophys'." # Tests if TwoPhotonSeries//OnePhotonSeries already exists in acquisition photon_series_metadata = metadata_copy["Ophys"][photon_series_type][photon_series_index] photon_series_name = photon_series_metadata["name"] if parent_container == "acquisition" and photon_series_name in nwbfile.acquisition: raise ValueError(f"{photon_series_name} already added to nwbfile.acquisition.") elif parent_container == "processing/ophys": ophys = get_module(nwbfile, name="ophys") if photon_series_name in ophys.data_interfaces: raise ValueError(f"{photon_series_name} already added to nwbfile.processing['ophys'].") # Add the image plane to nwb imaging_plane_name = photon_series_metadata["imaging_plane"] add_imaging_plane(nwbfile=nwbfile, metadata=metadata_copy, imaging_plane_name=imaging_plane_name) imaging_plane = nwbfile.get_imaging_plane(name=imaging_plane_name) photon_series_kwargs = deepcopy(photon_series_metadata) photon_series_kwargs.update(imaging_plane=imaging_plane) # Add the data frames_to_iterator = _imaging_frames_to_hdmf_iterator( imaging=imaging, iterator_type=iterator_type, iterator_options=iterator_options, ) data = H5DataIO(data=frames_to_iterator, compression=True) photon_series_kwargs.update(data=data) # Add dimension photon_series_kwargs.update(dimension=imaging.get_image_size()) # Add timestamps or rate if imaging.has_time_vector(): timestamps = imaging.frame_to_time(np.arange(imaging.get_num_frames())) estimated_rate = calculate_regular_series_rate(series=timestamps) if estimated_rate: photon_series_kwargs.update(starting_time=timestamps[0], rate=estimated_rate) else: photon_series_kwargs.update(timestamps=H5DataIO(data=timestamps, compression="gzip"), rate=None) else: rate = float(imaging.get_sampling_frequency()) photon_series_kwargs.update(rate=rate) # Add the photon series to the nwbfile (either as OnePhotonSeries or TwoPhotonSeries) photon_series = dict( OnePhotonSeries=OnePhotonSeries, TwoPhotonSeries=TwoPhotonSeries, )[ photon_series_type ](**photon_series_kwargs) if parent_container == "acquisition": nwbfile.add_acquisition(photon_series) elif parent_container == "processing/ophys": ophys = get_module(nwbfile, name="ophys") ophys.add(photon_series) return nwbfile
[docs]def check_if_imaging_fits_into_memory(imaging: ImagingExtractor) -> None: """ Raise an error if the full traces of an imaging extractor are larger than available memory. Parameters ---------- imaging : ImagingExtractor An imaging extractor object from roiextractors. Raises ------ MemoryError """ element_size_in_bytes = imaging.get_dtype().itemsize image_size = imaging.get_image_size() num_frames = imaging.get_num_frames() traces_size_in_bytes = num_frames * math.prod(image_size) * element_size_in_bytes available_memory_in_bytes = psutil.virtual_memory().available if traces_size_in_bytes > available_memory_in_bytes: message = ( f"Memory error, full TwoPhotonSeries data is {round(traces_size_in_bytes/1e9, 2)} GB) but only" f"({round(available_memory_in_bytes/1e9, 2)} GB are available! Please use iterator_type='v2'." ) raise MemoryError(message)
[docs]def _imaging_frames_to_hdmf_iterator( imaging: ImagingExtractor, iterator_type: Optional[str] = "v2", iterator_options: Optional[dict] = None, ): """ Private auxiliary method to wrap frames from an ImagingExtractor into a DataChunkIterator. Parameters ---------- imaging : ImagingExtractor The imaging extractor to get the data from. iterator_type : {"v2", "v1", None}, default: 'v2' The type of DataChunkIterator to use. 'v1' is the original DataChunkIterator of the hdmf data_utils. 'v2' is the locally developed SpikeInterfaceRecordingDataChunkIterator, which offers full control over chunking. None: write the TimeSeries with no memory chunking. iterator_options : dict, optional Dictionary of options for the iterator. For 'v1' this is the same as the options for the DataChunkIterator. For 'v2', see https://hdmf.readthedocs.io/en/stable/hdmf.data_utils.html#hdmf.data_utils.GenericDataChunkIterator for the full list of options. Returns ------- DataChunkIterator The frames of the imaging extractor wrapped in an iterator object. """ def data_generator(imaging): for i in range(imaging.get_num_frames()): yield imaging.get_frames(frame_idxs=[i]).squeeze().T assert iterator_type in ["v1", "v2", None], "'iterator_type' must be either 'v1', 'v2' (recommended), or None." iterator_options = dict() if iterator_options is None else iterator_options if iterator_type is None: check_if_imaging_fits_into_memory(imaging=imaging) return imaging.get_video().transpose((0, 2, 1)) if iterator_type == "v1": if "buffer_size" not in iterator_options: iterator_options.update(buffer_size=10) return DataChunkIterator(data=data_generator(imaging), **iterator_options) return ImagingExtractorDataChunkIterator(imaging_extractor=imaging, **iterator_options)
def add_imaging( imaging: ImagingExtractor, nwbfile: NWBFile, metadata: Optional[dict] = None, photon_series_type: Literal["TwoPhotonSeries", "OnePhotonSeries"] = "TwoPhotonSeries", photon_series_index: int = 0, iterator_type: Optional[str] = "v2", iterator_options: Optional[dict] = None, parent_container: Literal["acquisition", "processing/ophys"] = "acquisition", ): add_devices(nwbfile=nwbfile, metadata=metadata) add_photon_series( imaging=imaging, nwbfile=nwbfile, metadata=metadata, photon_series_type=photon_series_type, photon_series_index=photon_series_index, iterator_type=iterator_type, iterator_options=iterator_options, parent_container=parent_container, )
[docs]def write_imaging( imaging: ImagingExtractor, nwbfile_path: OptionalFilePathType = None, nwbfile: Optional[NWBFile] = None, metadata: Optional[dict] = None, overwrite: bool = False, verbose: bool = True, iterator_type: str = "v2", iterator_options: Optional[dict] = None, photon_series_type: Literal["TwoPhotonSeries", "OnePhotonSeries"] = "TwoPhotonSeries", buffer_size: Optional[int] = None, # TODO: to be removed ): """ Primary method for writing an ImagingExtractor object to an NWBFile. Parameters ---------- imaging: ImagingExtractor The imaging extractor object to be written to nwb nwbfile_path: FilePathType Path for where to write or load (if overwrite=False) the NWBFile. If specified, the context will always write to this location. nwbfile: NWBFile, optional If passed, this function will fill the relevant fields within the NWBFile object. E.g., calling write_recording(recording=my_recording_extractor, nwbfile=my_nwbfile) will result in the appropriate changes to the my_nwbfile object. If neither 'nwbfile_path' nor 'nwbfile' are specified, an NWBFile object will be automatically generated and returned by the function. metadata: dict, optional Metadata dictionary with information used to create the NWBFile when one does not exist or overwrite=True. overwrite: bool, optional Whether to overwrite the NWBFile if one exists at the nwbfile_path. The default is False (append mode). verbose: bool, optional If 'nwbfile_path' is specified, informs user after a successful write operation. The default is True. iterator_type: {"v2", "v1", None}, default: 'v2' The type of DataChunkIterator to use. 'v1' is the original DataChunkIterator of the hdmf data_utils. 'v2' is the locally developed SpikeInterfaceRecordingDataChunkIterator, which offers full control over chunking. None: write the TimeSeries with no memory chunking. iterator_options : dict, optional Dictionary of options for the iterator. For 'v1' this is the same as the options for the DataChunkIterator. For 'v2', see https://hdmf.readthedocs.io/en/stable/hdmf.data_utils.html#hdmf.data_utils.GenericDataChunkIterator for the full list of options. """ assert ( nwbfile_path is None or nwbfile is None ), "Either pass a nwbfile_path location, or nwbfile object, but not both!" if nwbfile is not None: assert isinstance(nwbfile, NWBFile), "'nwbfile' should be of type pynwb.NWBFile" iterator_options = iterator_options or dict() if buffer_size: warn( "Keyword argument 'buffer_size' is deprecated and will be removed on or after September 1st, 2022." "Specify as a key in the new 'iterator_options' dictionary instead." ) if metadata is None: metadata = dict() if hasattr(imaging, "nwb_metadata"): metadata = dict_deep_update(imaging.nwb_metadata, metadata, append_list=False) with make_or_load_nwbfile( nwbfile_path=nwbfile_path, nwbfile=nwbfile, metadata=metadata, overwrite=overwrite, verbose=verbose ) as nwbfile_out: add_imaging( imaging=imaging, nwbfile=nwbfile, metadata=metadata, photon_series_type=photon_series_type, iterator_type=iterator_type, iterator_options=iterator_options, ) return nwbfile_out
[docs]def get_nwb_segmentation_metadata(sgmextractor: SegmentationExtractor) -> dict: """ Convert metadata from the segmentation into nwb specific metadata. Parameters ---------- sgmextractor: SegmentationExtractor """ metadata = get_default_segmentation_metadata() # Optical Channel name: for i in range(sgmextractor.get_num_channels()): ch_name = sgmextractor.get_channel_names()[i] if i == 0: metadata["Ophys"]["ImagingPlane"][0]["optical_channel"][i]["name"] = ch_name else: metadata["Ophys"]["ImagingPlane"][0]["optical_channel"].append( dict( name=ch_name, emission_lambda=np.nan, description=f"{ch_name} description", ) ) plane_segmentation_name = metadata["Ophys"]["ImageSegmentation"]["plane_segmentations"][0]["name"] for trace_name, trace_data in sgmextractor.get_traces_dict().items(): # raw traces have already default name ("RoiResponseSeries") if trace_name in ["raw", "dff"]: continue if trace_data is not None and len(trace_data.shape) != 0: metadata["Ophys"]["Fluorescence"][plane_segmentation_name][trace_name] = dict( name=trace_name.capitalize(), description=f"description of {trace_name} traces", ) return metadata
[docs]def add_plane_segmentation( segmentation_extractor: SegmentationExtractor, nwbfile: NWBFile, metadata: Optional[dict], plane_segmentation_name: Optional[str] = None, plane_segmentation_index: Optional[int] = None, # TODO: to be removed include_roi_centroids: bool = True, include_roi_acceptance: bool = True, mask_type: Optional[str] = "image", # Optional[Literal["image", "pixel"]] iterator_options: Optional[dict] = None, compression_options: Optional[dict] = None, ) -> NWBFile: """ Adds the plane segmentation specified by the metadata to the image segmentation. If the plane segmentation already exists in the image segmentation, it is not added again. Parameters ---------- segmentation_extractor : SegmentationExtractor The segmentation extractor to get the results from. nwbfile : NWBFile The NWBFile to add the plane segmentation to. metadata : dict, optional The metadata for the plane segmentation. plane_segmentation_name : str, optional The name of the plane segmentation to be added. include_roi_centroids : bool, default: True Whether to include the ROI centroids on the PlaneSegmentation table. If there are a very large number of ROIs (such as in whole-brain recordings), you may wish to disable this for faster write speeds. include_roi_acceptance : bool, default: True Whether to include if the detected ROI was 'accepted' or 'rejected'. If there are a very large number of ROIs (such as in whole-brain recordings), you may wish to disable this for faster write speeds. mask_type : {'image', 'pixel', 'voxel'}, optional There are two types of ROI masks in NWB: ImageMasks and PixelMasks. Image masks have the same shape as the reference images the segmentation was applied to, and weight each pixel by its contribution to the ROI (typically boolean, with 0 meaning 'not in the ROI'). Pixel masks are instead indexed by ROI, with the data at each index being the shape of the image by the number of pixels in each ROI. Voxel masks are instead indexed by ROI, with the data at each index being the shape of the volume by the number of voxels in each ROI. Specify your choice between these three as mask_type='image', 'pixel', 'voxel', or None. If None, the mask information is not written to the NWB file. Defaults to 'image'. iterator_options : dict, optional The options to use when iterating over the image masks of the segmentation extractor. compression_options : dict, optional The options to use when compressing the image masks of the segmentation extractor. Returns ------- NWBFile The nwbfile passed as an input with the plane segmentation added. """ assert mask_type in ["image", "pixel", "voxel", None], ( "Keyword argument 'mask_type' must be one of either 'image', 'pixel', 'voxel', " f"or None (to not write any masks)! Received '{mask_type}'." ) iterator_options = iterator_options or dict() compression_options = compression_options or dict(compression="gzip") # Set the defaults and required infrastructure metadata_copy = deepcopy(metadata) default_metadata = get_default_segmentation_metadata() metadata_copy = dict_deep_update(default_metadata, metadata_copy, append_list=False) image_segmentation_metadata = metadata_copy["Ophys"]["ImageSegmentation"] plane_segmentation_name = ( plane_segmentation_name or default_metadata["Ophys"]["ImageSegmentation"]["plane_segmentations"][0]["name"] ) if plane_segmentation_index: warn( "Keyword argument 'plane_segmentation_index' is deprecated and it will be removed on 2024-04-16. Use 'plane_segmentation_name' instead." ) plane_segmentation_name = image_segmentation_metadata["plane_segmentations"][plane_segmentation_index]["name"] plane_segmentation_metadata = next( ( plane_segmentation_metadata for plane_segmentation_metadata in image_segmentation_metadata["plane_segmentations"] if plane_segmentation_metadata["name"] == plane_segmentation_name ), None, ) if plane_segmentation_metadata is None: raise ValueError( f"Metadata for Plane Segmentation '{plane_segmentation_name}' not found in metadata['Ophys']['ImageSegmentation']['plane_segmentations']." ) imaging_plane_name = plane_segmentation_metadata["imaging_plane"] add_imaging_plane(nwbfile=nwbfile, metadata=metadata_copy, imaging_plane_name=imaging_plane_name) add_image_segmentation(nwbfile=nwbfile, metadata=metadata_copy) ophys = get_module(nwbfile, "ophys") image_segmentation_name = image_segmentation_metadata["name"] image_segmentation = ophys.get_data_interface(image_segmentation_name) # Check if the plane segmentation already exists in the image segmentation if plane_segmentation_name not in image_segmentation.plane_segmentations: roi_ids = segmentation_extractor.get_roi_ids() if include_roi_acceptance: accepted_ids = [int(roi_id in segmentation_extractor.get_accepted_list()) for roi_id in roi_ids] rejected_ids = [int(roi_id in segmentation_extractor.get_rejected_list()) for roi_id in roi_ids] imaging_plane = nwbfile.imaging_planes[imaging_plane_name] plane_segmentation_kwargs = deepcopy(plane_segmentation_metadata) plane_segmentation_kwargs.update(imaging_plane=imaging_plane) if mask_type is None: plane_segmentation = PlaneSegmentation(id=roi_ids, **plane_segmentation_kwargs) elif mask_type == "image": plane_segmentation = PlaneSegmentation(id=roi_ids, **plane_segmentation_kwargs) plane_segmentation.add_column( name="image_mask", description="Image masks for each ROI.", data=H5DataIO(segmentation_extractor.get_roi_image_masks().T, **compression_options), ) elif mask_type == "pixel" or mask_type == "voxel": pixel_masks = segmentation_extractor.get_roi_pixel_masks() num_pixel_dims = pixel_masks[0].shape[1] assert num_pixel_dims in [3, 4], ( "The segmentation extractor returned a pixel mask that is not 3- or 4- dimensional! " "Please open a ticket with https://github.com/catalystneuro/roiextractors/issues" ) if mask_type == "pixel" and num_pixel_dims == 4: warn( "Specified mask_type='pixel', but ROIExtractors returned 4-dimensional masks. " "Using mask_type='voxel' instead." ) mask_type = "voxel" if mask_type == "voxel" and num_pixel_dims == 3: warn( "Specified mask_type='voxel', but ROIExtractors returned 3-dimensional masks. " "Using mask_type='pixel' instead." ) mask_type = "pixel" mask_type_kwarg = f"{mask_type}_mask" plane_segmentation = PlaneSegmentation(**plane_segmentation_kwargs) for roi_id, pixel_mask in zip(roi_ids, pixel_masks): plane_segmentation.add_roi(**{"id": roi_id, mask_type_kwarg: [tuple(x) for x in pixel_mask]}) if include_roi_centroids: # ROIExtractors uses height x width x (depth), but NWB uses width x height x depth tranpose_image_convention = (1, 0) if len(segmentation_extractor.get_image_size()) == 2 else (1, 0, 2) roi_locations = segmentation_extractor.get_roi_locations()[tranpose_image_convention, :].T plane_segmentation.add_column( name="ROICentroids", description="The x, y, (z) centroids of each ROI.", data=H5DataIO(roi_locations, **compression_options), ) if include_roi_acceptance: plane_segmentation.add_column( name="Accepted", description="1 if ROI was accepted or 0 if rejected as a cell during segmentation operation.", data=H5DataIO(accepted_ids, **compression_options), ) plane_segmentation.add_column( name="Rejected", description="1 if ROI was rejected or 0 if accepted as a cell during segmentation operation.", data=H5DataIO(rejected_ids, **compression_options), ) image_segmentation.add_plane_segmentation(plane_segmentations=[plane_segmentation]) return nwbfile
[docs]def add_fluorescence_traces( segmentation_extractor: SegmentationExtractor, nwbfile: NWBFile, metadata: Optional[dict], plane_segmentation_name: Optional[str] = None, plane_index: Optional[int] = None, # TODO: to be removed iterator_options: Optional[dict] = None, compression_options: Optional[dict] = None, ) -> NWBFile: """ Adds the fluorescence traces specified by the metadata to the nwb file. The fluorescence traces that are added are the one located in metadata["Ophys"]["Fluorescence"]. The df/F traces that are added are the one located in metadata["Ophys"]["DfOverF"]. Parameters ---------- segmentation_extractor : SegmentationExtractor The segmentation extractor to get the traces from. nwbfile : NWBFile The nwbfile to add the fluorescence traces to. metadata : dict The metadata for the fluorescence traces. plane_segmentation_name : str, optional The name of the plane segmentation that identifies which plane to add the fluorescence traces to. iterator_options : dict, optional compression_options : dict, optional Returns ------- NWBFile The nwbfile passed as an input with the fluorescence traces added. """ iterator_options = iterator_options or dict() compression_options = compression_options or dict(compression="gzip") # Set the defaults and required infrastructure metadata_copy = deepcopy(metadata) default_metadata = get_default_segmentation_metadata() metadata_copy = dict_deep_update(default_metadata, metadata_copy, append_list=False) if plane_index: warn( "Keyword argument 'plane_index' is deprecated and it will be removed on 2024-04-16. Use 'plane_segmentation_name' instead." ) plane_segmentation_name = metadata_copy["Ophys"]["ImageSegmentation"]["plane_segmentations"][plane_index][ "name" ] plane_segmentation_name = ( plane_segmentation_name or default_metadata["Ophys"]["ImageSegmentation"]["plane_segmentations"][0]["name"] ) # df/F metadata df_over_f_metadata = metadata_copy["Ophys"]["DfOverF"] df_over_f_name = df_over_f_metadata["name"] # Fluorescence traces metadata fluorescence_metadata = metadata_copy["Ophys"]["Fluorescence"] fluorescence_name = fluorescence_metadata["name"] # Get traces from the segmentation extractor traces_to_add = segmentation_extractor.get_traces_dict() # Filter empty data traces_to_add = { trace_name: trace for trace_name, trace in traces_to_add.items() if trace is not None and trace.size != 0 } # Filter all zero data # traces_to_add = { # trace_name: trace for trace_name, trace in traces_to_add.items() if any(x != 0 for x in np.ravel(trace)) # } # Early return if there is nothing to add if not traces_to_add: return nwbfile # Create a reference for ROIs from the plane segmentation roi_table_region = _create_roi_table_region( segmentation_extractor=segmentation_extractor, nwbfile=nwbfile, metadata=metadata_copy, plane_segmentation_name=plane_segmentation_name, ) roi_response_series_kwargs = dict(rois=roi_table_region, unit="n.a.") # Add timestamps or rate if segmentation_extractor.has_time_vector(): timestamps = segmentation_extractor.frame_to_time(np.arange(segmentation_extractor.get_num_frames())) estimated_rate = calculate_regular_series_rate(series=timestamps) if estimated_rate: roi_response_series_kwargs.update(starting_time=timestamps[0], rate=estimated_rate) else: roi_response_series_kwargs.update(timestamps=H5DataIO(data=timestamps, compression="gzip"), rate=None) else: rate = float(segmentation_extractor.get_sampling_frequency()) roi_response_series_kwargs.update(rate=rate) trace_to_data_interface = defaultdict() traces_to_add_to_fluorescence_data_interface = [ trace_name for trace_name in traces_to_add.keys() if trace_name != "dff" ] if traces_to_add_to_fluorescence_data_interface: fluorescence_data_interface = _get_segmentation_data_interface( nwbfile=nwbfile, data_interface_name=fluorescence_name ) trace_to_data_interface.default_factory = lambda: fluorescence_data_interface if "dff" in traces_to_add: df_over_f_data_interface = _get_segmentation_data_interface(nwbfile=nwbfile, data_interface_name=df_over_f_name) trace_to_data_interface.update(dff=df_over_f_data_interface) for trace_name, trace in traces_to_add.items(): # Decide which data interface to use based on the trace name data_interface = trace_to_data_interface[trace_name] data_interface_metadata = df_over_f_metadata if isinstance(data_interface, DfOverF) else fluorescence_metadata # Extract the response series metadata # the name of the trace is retrieved from the metadata, no need to override it here # trace_name = "RoiResponseSeries" if trace_name in ["raw", "dff"] else trace_name.capitalize() assert plane_segmentation_name in data_interface_metadata, ( f"Plane segmentation '{plane_segmentation_name}' not found in " f"{data_interface_metadata} metadata." ) trace_metadata = data_interface_metadata[plane_segmentation_name][trace_name] if trace_metadata is None: raise ValueError(f"Metadata for '{trace_name}' trace not found in {data_interface_metadata}.") if trace_metadata["name"] in data_interface.roi_response_series: continue # Pop the rate from the metadata if irregular time series if "timestamps" in roi_response_series_kwargs and "rate" in trace_metadata: trace_metadata.pop("rate") # Build the roi response series roi_response_series_kwargs.update( data=H5DataIO(SliceableDataChunkIterator(trace, **iterator_options), **compression_options), rois=roi_table_region, **trace_metadata, ) roi_response_series = RoiResponseSeries(**roi_response_series_kwargs) # Add trace to the data interface data_interface.add_roi_response_series(roi_response_series) return nwbfile
[docs]def _create_roi_table_region( segmentation_extractor: SegmentationExtractor, nwbfile: NWBFile, metadata: dict, plane_segmentation_name: Optional[str] = None, plane_index: Optional[int] = None, ): """Private method to create ROI table region. Parameters ---------- segmentation_extractor : SegmentationExtractor The segmentation extractor to get the results from. nwbfile : NWBFile The NWBFile to add the plane segmentation to. metadata : dict, optional The metadata for the plane segmentation. plane_segmentation_name : str, optional The name of the plane segmentation that identifies which plane to add the ROI table region to. """ image_segmentation_metadata = metadata["Ophys"]["ImageSegmentation"] if plane_index: warn( "Keyword argument 'plane_index' is deprecated and it will be removed on 2024-04-16. Use 'plane_segmentation_name' instead." ) plane_segmentation_name = image_segmentation_metadata["plane_segmentations"][plane_index]["name"] add_plane_segmentation( segmentation_extractor=segmentation_extractor, nwbfile=nwbfile, metadata=metadata, plane_segmentation_name=plane_segmentation_name, ) image_segmentation_name = image_segmentation_metadata["name"] ophys = get_module(nwbfile, "ophys") image_segmentation = ophys.get_data_interface(image_segmentation_name) # Get plane segmentation from the image segmentation plane_segmentation = image_segmentation.plane_segmentations[plane_segmentation_name] # Create a reference for ROIs from the plane segmentation id_list = list(plane_segmentation.id) imaging_plane_name = plane_segmentation.imaging_plane.name roi_table_region = plane_segmentation.create_roi_table_region( region=[id_list.index(id) for id in segmentation_extractor.get_roi_ids()], description=f"The ROIs for {imaging_plane_name}.", ) return roi_table_region
[docs]def _get_segmentation_data_interface(nwbfile: NWBFile, data_interface_name: str): """Private method to get the container for the segmentation data. If the container does not exist, it is created.""" ophys = get_module(nwbfile, "ophys") if data_interface_name in ophys.data_interfaces: return ophys.get(data_interface_name) if data_interface_name == "DfOverF": data_interface = DfOverF(name=data_interface_name) else: data_interface = Fluorescence(name=data_interface_name) # Add the data interface to the ophys module ophys.add(data_interface) return data_interface
[docs]def add_summary_images( nwbfile: NWBFile, segmentation_extractor: SegmentationExtractor, metadata: Optional[dict] = None, plane_segmentation_name: Optional[str] = None, images_set_name: Optional[str] = None, # TODO: to be removed ) -> NWBFile: """ Adds summary images (i.e. mean and correlation) to the nwbfile using an image container object pynwb.Image Parameters ---------- nwbfile : NWBFile An previously defined -in memory- NWBFile. segmentation_extractor : SegmentationExtractor A segmentation extractor object from roiextractors. metadata: dict, optional The metadata for the summary images is located in metadata["Ophys"]["SegmentationImages"]. plane_segmentation_name: str, optional The name of the plane segmentation that identifies which images to add. Returns ------- NWBFile The nwbfile passed as an input with the summary images added. """ if metadata is None: metadata = dict() # Set the defaults and required infrastructure metadata_copy = deepcopy(metadata) default_metadata = get_default_segmentation_metadata() metadata_copy = dict_deep_update(default_metadata, metadata_copy, append_list=False) segmentation_images_metadata = metadata_copy["Ophys"]["SegmentationImages"] if images_set_name is not None: warn( "Keyword argument 'images_set_name' is deprecated it will be removed on 2024-04-16." "Specify the name of the Images container in metadata['Ophys']['SegmentationImages'] instead." ) segmentation_images_metadata["name"] = images_set_name images_container_name = segmentation_images_metadata["name"] images_dict = segmentation_extractor.get_images_dict() images_to_add = {img_name: img for img_name, img in images_dict.items() if img is not None} if not images_to_add: return nwbfile ophys = get_module(nwbfile=nwbfile, name="ophys", description="contains optical physiology processed data") image_collection_does_not_exist = images_container_name not in ophys.data_interfaces if image_collection_does_not_exist: ophys.add(Images(name=images_container_name, description=segmentation_images_metadata["description"])) image_collection = ophys.data_interfaces[images_container_name] plane_segmentation_name = ( plane_segmentation_name or default_metadata["Ophys"]["ImageSegmentation"]["plane_segmentations"][0]["name"] ) assert ( plane_segmentation_name in segmentation_images_metadata ), f"Plane segmentation '{plane_segmentation_name}' not found in metadata['Ophys']['SegmentationImages']" images_metadata = segmentation_images_metadata[plane_segmentation_name] for img_name, img in images_to_add.items(): image_kwargs = dict(name=img_name, data=img.T) image_metadata = images_metadata.get(img_name, None) if image_metadata is not None: image_kwargs.update(image_metadata) # Note that nwb uses the conversion width x height (columns, rows) and roiextractors uses the transpose image_collection.add_image(GrayscaleImage(**image_kwargs)) return nwbfile
def add_segmentation( segmentation_extractor: SegmentationExtractor, nwbfile: NWBFile, metadata: Optional[dict] = None, plane_segmentation_name: Optional[str] = None, plane_num: Optional[int] = None, # TODO: to be removed include_roi_centroids: bool = True, include_roi_acceptance: bool = True, mask_type: Optional[str] = "image", # Literal["image", "pixel"] iterator_options: Optional[dict] = None, compression_options: Optional[dict] = None, ): # Add device: add_devices(nwbfile=nwbfile, metadata=metadata) # `add_imaging_plane` is also called from `add_plane_segmentation` so no need to call it explicitly here # Add PlaneSegmentation: add_plane_segmentation( segmentation_extractor=segmentation_extractor, nwbfile=nwbfile, metadata=metadata, plane_segmentation_name=plane_segmentation_name, include_roi_centroids=include_roi_centroids, include_roi_acceptance=include_roi_acceptance, mask_type=mask_type, iterator_options=iterator_options, compression_options=compression_options, ) # Add fluorescence traces: add_fluorescence_traces( segmentation_extractor=segmentation_extractor, nwbfile=nwbfile, metadata=metadata, plane_segmentation_name=plane_segmentation_name, iterator_options=iterator_options, compression_options=compression_options, ) # Adding summary images (mean and correlation) add_summary_images( nwbfile=nwbfile, segmentation_extractor=segmentation_extractor, metadata=metadata, plane_segmentation_name=plane_segmentation_name, )
[docs]def write_segmentation( segmentation_extractor: SegmentationExtractor, nwbfile_path: OptionalFilePathType = None, nwbfile: Optional[NWBFile] = None, metadata: Optional[dict] = None, overwrite: bool = False, verbose: bool = True, include_roi_centroids: bool = True, include_roi_acceptance: bool = True, mask_type: Optional[str] = "image", # Literal["image", "pixel"] iterator_options: Optional[dict] = None, compression_options: Optional[dict] = None, ) -> NWBFile: """ Primary method for writing an SegmentationExtractor object to an NWBFile. Parameters ---------- segmentation_extractor: SegmentationExtractor The segmentation extractor object to be written to nwb nwbfile_path: FilePathType Path for where to write or load (if overwrite=False) the NWBFile. If specified, the context will always write to this location. nwbfile: NWBFile, optional If passed, this function will fill the relevant fields within the NWBFile object. E.g., calling write_recording(recording=my_recording_extractor, nwbfile=my_nwbfile) will result in the appropriate changes to the my_nwbfile object. If neither 'nwbfile_path' nor 'nwbfile' are specified, an NWBFile object will be automatically generated and returned by the function. metadata: dict, optional Metadata dictionary with information used to create the NWBFile when one does not exist or overwrite=True. overwrite: bool, default: False Whether to overwrite the NWBFile if one exists at the nwbfile_path. verbose: bool, default: True If 'nwbfile_path' is specified, informs user after a successful write operation. buffer_size : int, default: 10 The buffer size in GB. plane_num : int, default: 0 The plane number to be extracted. include_roi_centroids : bool, default: True Whether to include the ROI centroids on the PlaneSegmentation table. If there are a very large number of ROIs (such as in whole-brain recordings), you may wish to disable this for faster write speeds. Defaults to True. include_roi_acceptance : bool, default: True Whether to include if the detected ROI was 'accepted' or 'rejected'. If there are a very large number of ROIs (such as in whole-brain recordings), you may wish to ddisable this for faster write speeds. Defaults to True. mask_type : str, optional There are two types of ROI masks in NWB: ImageMasks and PixelMasks. Image masks have the same shape as the reference images the segmentation was applied to, and weight each pixel by its contribution to the ROI (typically boolean, with 0 meaning 'not in the ROI'). Pixel masks are instead indexed by ROI, with the data at each index being the shape of the image by the number of pixels in each ROI. Voxel masks are instead indexed by ROI, with the data at each index being the shape of the volume by the number of voxels in each ROI. Specify your choice between these two as mask_type='image', 'pixel', 'voxel', or None. If None, the mask information is not written to the NWB file. Defaults to 'image'. """ assert ( nwbfile_path is None or nwbfile is None ), "Either pass a nwbfile_path location, or nwbfile object, but not both!" iterator_options = iterator_options or dict() compression_options = compression_options or dict(compression="gzip") # parse metadata correctly considering the MultiSegmentationExtractor function: if isinstance(segmentation_extractor, MultiSegmentationExtractor): segmentation_extractors = segmentation_extractor.segmentations if metadata is not None: assert isinstance(metadata, list), ( "For MultiSegmentationExtractor enter 'metadata' as a list of " "SegmentationExtractor metadata" ) assert len(metadata) == len(segmentation_extractor), ( "The 'metadata' argument should be a list with the same " "number of elements as the segmentations in the " "MultiSegmentationExtractor" ) else: segmentation_extractors = [segmentation_extractor] if metadata is not None and not isinstance(metadata, list): metadata = [metadata] metadata_base_list = [ get_nwb_segmentation_metadata(segmentation_extractor) for segmentation_extractor in segmentation_extractors ] # updating base metadata with new: for num, data in enumerate(metadata_base_list): metadata_input = metadata[num] if metadata else {} metadata_base_list[num] = dict_deep_update(metadata_base_list[num], metadata_input, append_list=False) metadata_base_common = metadata_base_list[0] with make_or_load_nwbfile( nwbfile_path=nwbfile_path, nwbfile=nwbfile, metadata=metadata_base_common, overwrite=overwrite, verbose=verbose ) as nwbfile_out: _ = get_module(nwbfile=nwbfile_out, name="ophys", description="contains optical physiology processed data") for plane_no_loop, (segmentation_extractor, metadata) in enumerate( zip(segmentation_extractors, metadata_base_list) ): add_segmentation( segmentation_extractor=segmentation_extractor, nwbfile=nwbfile_out, metadata=metadata, plane_num=plane_no_loop, include_roi_centroids=include_roi_centroids, include_roi_acceptance=include_roi_acceptance, mask_type=mask_type, iterator_options=iterator_options, compression_options=compression_options, ) return nwbfile_out