Support digital channels on NIDQ interface and use TimeSeries instead…

… of ElecricalSeries for analog channels (#1152)

Co-authored-by: Ben Dichter <ben.dichter@gmail.com>

CHANGELOG.md

@@ -9,6 +9,8 @@
 * Make `NWBMetaDataEncoder` public again [PR #1142](https://github.com/catalystneuro/neuroconv/pull/1142)
 * Fix a bug where data in `DeepLabCutInterface` failed to write when `ndx-pose` was not imported. [#1144](https://github.com/catalystneuro/neuroconv/pull/1144)
 * `SpikeGLXConverterPipe` converter now accepts multi-probe structures with multi-trigger and does not assume a specific folder structure [#1150](https://github.com/catalystneuro/neuroconv/pull/1150)
+* `SpikeGLXNIDQInterface` is no longer written as an ElectricalSeries [#1152](https://github.com/catalystneuro/neuroconv/pull/1152)
+
 
 ## Features
 * Propagate the `unit_electrode_indices` argument from the spikeinterface tools to `BaseSortingExtractorInterface`. This allows users to map units to the electrode table when adding sorting data [PR #1124](https://github.com/catalystneuro/neuroconv/pull/1124)
@@ -17,7 +19,10 @@
 * `SpikeGLXRecordingInterface` now also accepts `folder_path` making its behavior equivalent to SpikeInterface [#1150](https://github.com/catalystneuro/neuroconv/pull/1150)
 * Added the `rclone_transfer_batch_job` helper function for executing Rclone data transfers in AWS Batch jobs. [PR #1085](https://github.com/catalystneuro/neuroconv/pull/1085)
 * Added the `deploy_neuroconv_batch_job` helper function for deploying NeuroConv AWS Batch jobs. [PR #1086](https://github.com/catalystneuro/neuroconv/pull/1086)
-* YAML specification files now accept an outer keyword `upload_to_dandiset="< six-digit ID >"` to automatically upload the produced NWB files to the DANDI archive [PR #1089](https://github.com/catalystneuro/neuroconv/pull/1089)
+* YAML specification files now accepts an outer keyword `upload_to_dandiset="< six-digit ID >"` to automatically upload the produced NWB files to the DANDI archive [PR #1089](https://github.com/catalystneuro/neuroconv/pull/1089)
+*`SpikeGLXNIDQInterface` now handdles digital demuxed channels (`XD0`) [#1152](https://github.com/catalystneuro/neuroconv/pull/1152)
+
+
 
 
 ## Improvements

docs/conversion_examples_gallery/recording/spikeglx.rst

@@ -24,7 +24,7 @@ We can easily convert all data stored in the native SpikeGLX folder structure to
     >>>
     >>> folder_path = f"{ECEPHY_DATA_PATH}/spikeglx/Noise4Sam_g0"
     >>> converter = SpikeGLXConverterPipe(folder_path=folder_path)
-    >>>
+    Source data is valid!
     >>> # Extract what metadata we can from the source files
     >>> metadata = converter.get_metadata()
     >>> # For data provenance we add the time zone information to the conversion

src/neuroconv/datainterfaces/ecephys/spikeglx/spikeglxdatainterface.py

@@ -66,18 +66,22 @@ def __init__(
             Folder path containing the binary files of the SpikeGLX recording.
         stream_id: str, optional
             Stream ID of the SpikeGLX recording.
-            Examples are 'nidq', 'imec0.ap', 'imec0.lf', 'imec1.ap', 'imec1.lf', etc.
+            Examples are 'imec0.ap', 'imec0.lf', 'imec1.ap', 'imec1.lf', etc.
         file_path : FilePathType
             Path to .bin file. Point to .ap.bin for SpikeGLXRecordingInterface and .lf.bin for SpikeGLXLFPInterface.
         verbose : bool, default: True
             Whether to output verbose text.
         es_key : str, the key to access the metadata of the ElectricalSeries.
         """
 
+        if stream_id == "nidq":
+            raise ValueError(
+                "SpikeGLXRecordingInterface is not designed to handle nidq files. Use SpikeGLXNIDQInterface instead"
+            )
+
         if file_path is not None and stream_id is None:
             self.stream_id = fetch_stream_id_for_spikelgx_file(file_path)
             self.folder_path = Path(file_path).parent
-
         else:
             self.stream_id = stream_id
             self.folder_path = Path(folder_path)

src/neuroconv/datainterfaces/ecephys/spikeglx/spikeglxnidqinterface.py

@@ -1,39 +1,36 @@
+import warnings
 from pathlib import Path
-from typing import Optional
+from typing import Literal, Optional
 
 import numpy as np
 from pydantic import ConfigDict, DirectoryPath, FilePath, validate_call
+from pynwb import NWBFile
+from pynwb.base import TimeSeries
 
 from .spikeglx_utils import get_session_start_time
-from ..baserecordingextractorinterface import BaseRecordingExtractorInterface
+from ....basedatainterface import BaseDataInterface
 from ....tools.signal_processing import get_rising_frames_from_ttl
-from ....utils import get_json_schema_from_method_signature
+from ....tools.spikeinterface.spikeinterface import _recording_traces_to_hdmf_iterator
+from ....utils import (
+    calculate_regular_series_rate,
+    get_json_schema_from_method_signature,
+)
 
 
-class SpikeGLXNIDQInterface(BaseRecordingExtractorInterface):
+class SpikeGLXNIDQInterface(BaseDataInterface):
     """Primary data interface class for converting the high-pass (ap) SpikeGLX format."""
 
     display_name = "NIDQ Recording"
-    keywords = BaseRecordingExtractorInterface.keywords + ("Neuropixels",)
+    keywords = ("Neuropixels", "nidq", "NIDQ", "SpikeGLX")
     associated_suffixes = (".nidq", ".meta", ".bin")
     info = "Interface for NIDQ board recording data."
 
-    ExtractorName = "SpikeGLXRecordingExtractor"
-    stream_id = "nidq"
-
     @classmethod
     def get_source_schema(cls) -> dict:
         source_schema = get_json_schema_from_method_signature(method=cls.__init__, exclude=["x_pitch", "y_pitch"])
         source_schema["properties"]["file_path"]["description"] = "Path to SpikeGLX .nidq file."
         return source_schema
 
-    def _source_data_to_extractor_kwargs(self, source_data: dict) -> dict:
-
-        extractor_kwargs = source_data.copy()
-        extractor_kwargs["folder_path"] = self.folder_path
-        extractor_kwargs["stream_id"] = self.stream_id
-        return extractor_kwargs
-
     @validate_call(config=ConfigDict(arbitrary_types_allowed=True))
     def __init__(
         self,
@@ -56,12 +53,18 @@ def __init__(
             Path to .nidq.bin file.
         verbose : bool, default: True
             Whether to output verbose text.
-        load_sync_channel : bool, default: False
-            Whether to load the last channel in the stream, which is typically used for synchronization.
-            If True, then the probe is not loaded.
         es_key : str, default: "ElectricalSeriesNIDQ"
         """
 
+        if load_sync_channel:
+
+            warnings.warn(
+                "The 'load_sync_channel' parameter is deprecated and will be removed in June 2025. "
+                "The sync channel data is only available the raw files of spikeglx`.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
+
         if file_path is None and folder_path is None:
             raise ValueError("Either 'file_path' or 'folder_path' must be provided.")
 
@@ -72,18 +75,36 @@ def __init__(
         if folder_path is not None:
             self.folder_path = Path(folder_path)
 
+        from spikeinterface.extractors import SpikeGLXRecordingExtractor
+
+        self.recording_extractor = SpikeGLXRecordingExtractor(
+            folder_path=self.folder_path,
+            stream_id="nidq",
+            all_annotations=True,
+        )
+
+        channel_ids = self.recording_extractor.get_channel_ids()
+        analog_channel_signatures = ["XA", "MA"]
+        self.analog_channel_ids = [ch for ch in channel_ids if "XA" in ch or "MA" in ch]
+        self.has_analog_channels = len(self.analog_channel_ids) > 0
+        self.has_digital_channels = len(self.analog_channel_ids) < len(channel_ids)
+        if self.has_digital_channels:
+            import ndx_events  # noqa: F401
+            from spikeinterface.extractors import SpikeGLXEventExtractor
+
+            self.event_extractor = SpikeGLXEventExtractor(folder_path=self.folder_path)
+
         super().__init__(
             verbose=verbose,
             load_sync_channel=load_sync_channel,
             es_key=es_key,
+            folder_path=self.folder_path,
+            file_path=file_path,
         )
-        self.source_data.update(file_path=str(file_path))
 
-        self.recording_extractor.set_property(
-            key="group_name", values=["NIDQChannelGroup"] * self.recording_extractor.get_num_channels()
-        )
+        self.subset_channels = None
 
-        signal_info_key = (0, self.stream_id)  # Key format is (segment_index, stream_id)
+        signal_info_key = (0, "nidq")  # Key format is (segment_index, stream_id)
         self._signals_info_dict = self.recording_extractor.neo_reader.signals_info_dict[signal_info_key]
         self.meta = self._signals_info_dict["meta"]
 
@@ -101,24 +122,206 @@ def get_metadata(self) -> dict:
             manufacturer="National Instruments",
         )
 
-        # Add groups metadata
-        metadata["Ecephys"]["Device"] = [device]
+        metadata["Devices"] = [device]
 
-        metadata["Ecephys"]["ElectrodeGroup"][0].update(
-            name="NIDQChannelGroup", description="A group representing the NIDQ channels.", device=device["name"]
-        )
-        metadata["Ecephys"]["Electrodes"] = [
-            dict(name="group_name", description="Name of the ElectrodeGroup this electrode is a part of."),
-        ]
-        metadata["Ecephys"]["ElectricalSeriesNIDQ"][
-            "description"
-        ] = "Raw acquisition traces from the NIDQ (.nidq.bin) channels."
         return metadata
 
     def get_channel_names(self) -> list[str]:
         """Return a list of channel names as set in the recording extractor."""
         return list(self.recording_extractor.get_channel_ids())
 
+    def add_to_nwbfile(
+        self,
+        nwbfile: NWBFile,
+        metadata: Optional[dict] = None,
+        stub_test: bool = False,
+        starting_time: Optional[float] = None,
+        write_as: Literal["raw", "lfp", "processed"] = "raw",
+        write_electrical_series: bool = True,
+        iterator_type: Optional[str] = "v2",
+        iterator_opts: Optional[dict] = None,
+        always_write_timestamps: bool = False,
+    ):
+        """
+        Add NIDQ board data to an NWB file, including both analog and digital channels if present.
+
+        Parameters
+        ----------
+        nwbfile : NWBFile
+            The NWB file to which the NIDQ data will be added
+        metadata : Optional[dict], default: None
+            Metadata dictionary with device information. If None, uses default metadata
+        stub_test : bool, default: False
+            If True, only writes a small amount of data for testing
+        starting_time : Optional[float], default: None
+            DEPRECATED: Will be removed in June 2025. Starting time offset for the TimeSeries
+        write_as : Literal["raw", "lfp", "processed"], default: "raw"
+            DEPRECATED: Will be removed in June 2025. Specifies how to write the data
+        write_electrical_series : bool, default: True
+            DEPRECATED: Will be removed in June 2025. Whether to write electrical series data
+        iterator_type : Optional[str], default: "v2"
+            Type of iterator to use for data streaming
+        iterator_opts : Optional[dict], default: None
+            Additional options for the iterator
+        always_write_timestamps : bool, default: False
+            If True, always writes timestamps instead of using sampling rate
+        """
+
+        if starting_time is not None:
+            warnings.warn(
+                "The 'starting_time' parameter is deprecated and will be removed in June 2025. "
+                "Use the time alignment methods for modifying the starting time or timestamps "
+                "of the data if needed: "
+                "https://neuroconv.readthedocs.io/en/main/user_guide/temporal_alignment.html",
+                DeprecationWarning,
+                stacklevel=2,
+            )
+
+        if write_as != "raw":
+            warnings.warn(
+                "The 'write_as' parameter is deprecated and will be removed in June 2025. "
+                "NIDQ should always be written in the acquisition module of NWB. "
+                "Writing data as LFP or processed data is not supported.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
+
+        if write_electrical_series is not True:
+            warnings.warn(
+                "The 'write_electrical_series' parameter is deprecated and will be removed in June 2025. "
+                "The option to skip the addition of the data is no longer supported. "
+                "This option was used in ElectricalSeries to write the electrode and electrode group "
+                "metadata without the raw data.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
+
+        if stub_test or self.subset_channels is not None:
+            recording = self.subset_recording(stub_test=stub_test)
+        else:
+            recording = self.recording_extractor
+
+        if metadata is None:
+            metadata = self.get_metadata()
+
+        # Add devices
+        device_metadata = metadata.get("Devices", [])
+        for device in device_metadata:
+            if device["name"] not in nwbfile.devices:
+                nwbfile.create_device(**device)
+
+        # Add analog and digital channels
+        if self.has_analog_channels:
+            self._add_analog_channels(
+                nwbfile=nwbfile,
+                recording=recording,
+                iterator_type=iterator_type,
+                iterator_opts=iterator_opts,
+                always_write_timestamps=always_write_timestamps,
+            )
+
+        if self.has_digital_channels:
+            self._add_digital_channels(nwbfile=nwbfile)
+
+    def _add_analog_channels(
+        self,
+        nwbfile: NWBFile,
+        recording,
+        iterator_type: Optional[str],
+        iterator_opts: Optional[dict],
+        always_write_timestamps: bool,
+    ):
+        """
+        Add analog channels from the NIDQ board to the NWB file.
+
+        Parameters
+        ----------
+        nwbfile : NWBFile
+            The NWB file to add the analog channels to
+        recording : BaseRecording
+            The recording extractor containing the analog channels
+        iterator_type : Optional[str]
+            Type of iterator to use for data streaming
+        iterator_opts : Optional[dict]
+            Additional options for the iterator
+        always_write_timestamps : bool
+            If True, always writes timestamps instead of using sampling rate
+        """
+        analog_recorder = recording.select_channels(channel_ids=self.analog_channel_ids)
+        channel_names = analog_recorder.get_property(key="channel_names")
+        segment_index = 0
+        analog_data_iterator = _recording_traces_to_hdmf_iterator(
+            recording=analog_recorder,
+            segment_index=segment_index,
+            iterator_type=iterator_type,
+            iterator_opts=iterator_opts,
+        )
+
+        name = "TimeSeriesNIDQ"
+        description = f"Analog data from the NIDQ board. Channels are {channel_names} in that order."
+        time_series_kwargs = dict(name=name, data=analog_data_iterator, unit="a.u.", description=description)
+
+        if always_write_timestamps:
+            timestamps = recording.get_times(segment_index=segment_index)
+            shifted_timestamps = timestamps
+            time_series_kwargs.update(timestamps=shifted_timestamps)
+        else:
+            recording_has_timestamps = recording.has_time_vector(segment_index=segment_index)
+            if recording_has_timestamps:
+                timestamps = recording.get_times(segment_index=segment_index)
+                rate = calculate_regular_series_rate(series=timestamps)
+                recording_t_start = timestamps[0]
+            else:
+                rate = recording.get_sampling_frequency()
+                recording_t_start = recording._recording_segments[segment_index].t_start or 0
+
+            if rate:
+                starting_time = float(recording_t_start)
+                time_series_kwargs.update(starting_time=starting_time, rate=recording.get_sampling_frequency())
+            else:
+                shifted_timestamps = timestamps
+                time_series_kwargs.update(timestamps=shifted_timestamps)
+
+        time_series = TimeSeries(**time_series_kwargs)
+        nwbfile.add_acquisition(time_series)
+
+    def _add_digital_channels(self, nwbfile: NWBFile):
+        """
+        Add digital channels from the NIDQ board to the NWB file as events.
+
+        Parameters
+        ----------
+        nwbfile : NWBFile
+            The NWB file to add the digital channels to
+        """
+        from ndx_events import LabeledEvents
+
+        event_channels = self.event_extractor.channel_ids
+        for channel_id in event_channels:
+            events_structure = self.event_extractor.get_events(channel_id=channel_id)
+            timestamps = events_structure["time"]
+            labels = events_structure["label"]
+
+            # Some channels have no events
+            if timestamps.size > 0:
+
+                # Timestamps are not ordered, the ones for off are first and then the ones for on
+                ordered_indices = np.argsort(timestamps)
+                ordered_timestamps = timestamps[ordered_indices]
+                ordered_labels = labels[ordered_indices]
+
+                unique_labels = np.unique(ordered_labels)
+                label_to_index = {label: index for index, label in enumerate(unique_labels)}
+                data = [label_to_index[label] for label in ordered_labels]
+
+                channel_name = channel_id.split("#")[-1]
+                description = f"On and Off Events from channel {channel_name}"
+                name = f"EventsNIDQDigitalChannel{channel_name}"
+                labeled_events = LabeledEvents(
+                    name=name, description=description, timestamps=ordered_timestamps, data=data, labels=unique_labels
+                )
+                nwbfile.add_acquisition(labeled_events)
+
     def get_event_times_from_ttl(self, channel_name: str) -> np.ndarray:
         """
         Return the start of event times from the rising part of TTL pulses on one of the NIDQ channels.