neurotorchmz.utils.signal_detection

neurotorchmz.utils.signal_detection#

Module to detect signals in an ImageObject

Classes

ISignalDetectionAlgorithm()

Abstract base class for a detection algorithm for signals.

SigDetect_DiffMax()

SigDetect_DiffStd()

SignalObject(imgObj)

A SignalObject holds a) references to the detected signal peaks of an ImageObject (for example stimulation) b) provides a signal used for determing those peaks and c) provides a sliced image including only / no peak frames.
class neurotorchmz.utils.signal_detection.ISignalDetectionAlgorithm#

Bases: object

Abstract base class for a detection algorithm for signals. Must only implement a get_signal method

get_signal(imgObj: ImageObject) → ndarray | None#

This method should return an 1D array (t,) interpretated as signal of the image

class neurotorchmz.utils.signal_detection.SigDetect_DiffMax#

Bases: ISignalDetectionAlgorithm

get_signal(imgObj: ImageObject) → ndarray | None#

This method should return an 1D array (t,) interpretated as signal of the image

class neurotorchmz.utils.signal_detection.SigDetect_DiffStd#

Bases: ISignalDetectionAlgorithm

get_signal(imgObj: ImageObject) → ndarray | None#

This method should return an 1D array (t,) interpretated as signal of the image

class neurotorchmz.utils.signal_detection.SignalObject(imgObj: ImageObject)#

Bases: object

A SignalObject holds a) references to the detected signal peaks of an ImageObject (for example stimulation) b) provides a signal used for determing those peaks and c) provides a sliced image including only / no peak frames. It is bound to an ImageObject and calculates its values lazy. The ImageObject is reponsible to call clear() when the underlying images (e.g. through a convolution of loading a new image) have changed.

Variables:

  • peakWidth_L (int) – When providing the sliced image without the signal peaks, n frames to the left are also excluded

  • peakWidth_R (int) – When providing the sliced image without the signal peaks, n frames to the right are also excluded

ALGORITHM: ISignalDetectionAlgorithm = <neurotorchmz.utils.signal_detection.ISignalDetectionAlgorithm object>#
PEAK_WIDTH_LEFT: int = 1#
PEAK_WIDTH_RIGHT: int = 6#
clear()#

Calling clear() will forget the current signal and found peaks. Should be called, when the underlying images of the ImageObject change

clear_peaks()#

Clears the peaks and the cached sliced images.

export_img_only_signal(path: Path) → None#

Export the current img

export_img_without_signal(path: Path) → None#

Export the current img

get_view(img_type: Literal['img', 'img_diff'], slice_type: Literal['only_signal', 'without_signal'], mode: ImageView) → AxisImage#

Internal function to get a view for every combination of img/img_diff and with or without signal frames

img_diff_only_signal_view(mode: ImageView) → AxisImage#
property img_diff_props_only_signal: ImageProperties#
property img_diff_props_without_signal: ImageProperties#
img_diff_without_signal_view(mode: ImageView) → AxisImage#
img_only_signal_view(mode: ImageView) → AxisImage#
property img_props_only_signal: ImageProperties#
property img_props_without_signal: ImageProperties#
img_without_signal_view(mode: ImageView) → AxisImage#
classmethod load_settings() → None#
property peaks: list[int] | None#

Returns the peaks given the current signal and prominence as sorted interger list (ascending). The peaks are calculated on first call

property prominence_factor: float#

The prominence factor is used to find peaks in the signal, as scipy.signal.find_peaks with prominence = factor*(max(signal)-min(signal)) is used

classmethod set_settings(peak_width_left: int | None = None, peak_width_right: int | None = None) → None#
property signal: ndarray | None#

Returns the signal from the image by calculating it using SignalObject.ALGORITHM on the first call

Contents