Skip to content

Latest commit

 

History

History
56 lines (38 loc) · 3.25 KB

README.md

File metadata and controls

56 lines (38 loc) · 3.25 KB
Raw

KairoSight

Python 3.7 package and user interface to analyze videos produced in cardiac optical mapping experiments.

Such optocardiography studies record fluorescent emissions from cardiac tissue preparations to capture transient transmembrane voltage (Vm) and/or intracellular calcium (Ca) activity.

Consisting of 3 stages (Preparation, Processing, Analysis), KairoSight enables the calculation of common cardiac mapping results:

  • Start time (ms) : the time of a transient's inflection from baseline (max of the 2nd derivative)
  • Activation time (ms) : the time of a transient's most significant rise toward peak fluorescence (max of the 1st derivative)
  • Duration time (ms) : the span of time from activation to a given percentage of peak fluorescence (often 80% or 90%)

This project started as a python port of Camat (cardiac mapping analysis tool) and is inspired by design cues and algorithms from RHYTHM and Fiji.

Setup

  • Clone or download repository.

  • From your directory, use pip to install required packages (ensure pip uses python3, or use pip3 if necessary):

    pip install -r requirements.txt

Usage

Running KairoSight

From /src/ start the user interface by running the following script:

python kairosight.py

Any .tif/.tiff image stack (video) can then be imported using the menu bar at the top of the KairoSight window.

Once a video is imported, the following stages can be completed and consist of distinct steps:

1) Preparation steps

  • Properties - enter frames-per-second, image scale (px/cm), etc.
  • Bin - apply spatial binning to reduce the number of pixels
  • Mask - apply a semi-automatically generated mask to eliminate background pixels

2) Processing steps

  • Normalize - convert each pixel from their arbitrary fluorescence units to a range of 0 to 1
  • Filter - apply a gaussian spatial filter to remove high-frequency noise (<3mm kernel size recommended)
  • Signal-to-Noise (SNR) - generate a map of SNR for each pixel with significant transient activity

3) Analysis steps

  • Time Crop - limit the frames of the video used for the ensuing analysis
  • Analyze - calculate and export trace-wise results, maps of a certain results across the entire FOV, or just the current signal displayed in the UI

Editing

User Interface (UI)

The UI is built with Qt Designer (Version 5.13.0) which, once all packages are installed, can be found in the interpreter's directory at:

\Lib\site-packages\pyqt5_tools\Qt\bin\designer.exe

The resulting .ui file must be converted into a .py file by using the puic5 command from the project's \ui\ directory. For example:

pyuic5 KairoSight_WindowMain.ui > KairoSight_WindowMain.py

The primary UI file, KairoSight_WindowMain.py, contains all of the analysis components. KairoSight_WindowMDI.py is a Multiple-document Interface (MDI) which can contain multiple primary UI windows.

Program logic

The primary logic file, kairosight.py, connects UI elements to their functions and provides feedback as a user proceeds through each stage/step.