Malaria stage classifier

This is the documentation of a Python package to classify RBCs in microscopy images. It includes:

  • a package for the stage-specific classification of RBCs (Malaria_stage_classifier) with four folders:

    • Code which contains four modules and the code file:

      • NN.py which initialises the neural network and trains the data

      • classes.py which contains classes for evaluating the properties of each RBC

      • contours.py which provides functions for the detection of RBCs in an image

      • extractCuts.py which provides functions for extracting the most characteristic profiles in the RBC

      • Malaria_stage_classifier.py

    • Logo which contains the logo for the pop up windows

    • Neural_networks which contains the pre-trained neural networks

    • Sample_images which contains sample images for the implemented imaging techniques (Note: for the text files, the number of header lines should be set to 3)

Download

  1. Download the Malaria_stage_classifier.zip folder which contains the code, the logo for the pop up windows, the pre-trained neural networks, and sample images from: https://github.com/KatharinaPreissinger/Malaria_stage_classifier.git

  2. If you want to retrain the neural network, please download the Datasets_for_NN (this requires at least 250 MB of free space)

  3. Read the documentation for more information about the classes and modules

How to use the package

  1. Run the code file Malaria_stage_classifier.py

  2. Tab Settings:

    • select the file you want to analyse

    • choose the type, in case of header lines in the text file choose the number of header lines

    • select the folder to save your output

  3. Tab Show image:

    • displays the image

  4. Tab Threshold image:

    • use the Set threshold button to show the thresholded image

    • optionally, the preset threshold value can be changed manually

  5. Tab Detect cells:

    • the cell detection parameters can be changed manually and set back to default

  6. Tab Classify cells:

    • Load NN loads the neural network

    • Predict stages predicts the cell stage

    • Change predictions provides the possibility to change the prediction by clicking on the cell

    • Add data to NN offers the option to add new data to the NN and retrains the NN

      • to use this option, please download the training data from: Datasets for NN

    • Save predictions saves the predictions in a text or csv file and offers the option to analyse new data