2. Build a SegmentationExtractor:#

To build a custom SegmentationExtractor that interfaces with the output of a custom segmentation pipeline used for post processing the raw image data.

  • get_accepted_list(): returns a list of accepted ROIs’ id numbers

  • get_frame_shape(): the (height, width) dimensions of the image

from roiextractors import SegmentationExtractor
from roiextractors.segmentationextractor import _RoiResponse

class MyFormatSegmentationExtractor(SegmentationExtractor):
    def __init__(self, file_path):
        RecordingExtractor.__init__(self)

        ## All file specific initialization code can go here.

        self._sampling_frequency = # logic to extract sampling frequency here
        self._channel_names = ['OpticalChannel']
        self._num_planes = 1
        self._image_masks = self._load_rois()
        cell_ids = list(range(self._image_masks.shape[2]))
        raw_traces = self._load_traces(name="raw")  # define a method to extract fluorescence traces
        if raw_traces is not None:
            self._roi_responses.append(_RoiResponse("raw", raw_traces, cell_ids))
        dff_traces = self._load_traces(name="dff")  # define a method to extract dF/F traces if any else None
        if dff_traces is not None:
            self._roi_responses.append(_RoiResponse("dff", dff_traces, cell_ids))
        neuropil_traces = self._load_traces(name="neuropil")  # define a method to extract neuropil info if any else None
        if neuropil_traces is not None:
            self._roi_responses.append(_RoiResponse("neuropil", neuropil_traces, cell_ids))
        denoised_traces = self._load_traces(name="denoised")  # define a method to extract denoised traces if any else None
        if denoised_traces is not None:
            self._roi_responses.append(_RoiResponse("denoised", denoised_traces, cell_ids))
        deconvolved_traces = self._load_traces(name="deconvolved")  # define a method to extract deconvolved traces if any else None
        if deconvolved_traces is not None:
            self._roi_responses.append(_RoiResponse("deconvolved", deconvolved_traces, cell_ids))
        self._image_correlation = self._load_summary_images()# define method to extract a correlation image else None
        self._image_mean = self._load_summary_images() # define method to extract a mean image else None

    def _load_traces(self, name):

        # define your logic to extract roi time traces
        # return a np.ndarray, None

    def _load_rois(self):

        # define logic to extract the image masks used to define the extracted regions of interest
        # will return a np.ndarray with 0,1 values. 1: the ROI. Shape as image shape.

    def _load_summary_images(self):

        # define method to extract and return summary images like mean/correlation etc.
        # return np.ndarray , shape: x,y pixels

    def get_accepted_list(self):

        # define method to get all the accepted ROIs after the segmentation operation of pipeline

    def get_rejected_list(self):

        # define method to get all the rejected ROIs after the segmentation operation of pipeline
        # this can also be the compliment of accepted list

    def get_sampling_frequency(self):

        return self._sampling_frequency

    def get_roi_ids(self):

        # define logic to get the numerical id of the ROI output by the segmentation pipeline.

    def get_frame_shape(self):

        # returns the (height, width) dimensions as a tuple
        # e.g., self._image_mean.shape
        return self._image_mean.shape

This Page