This project is an open-source API of a GUI for visualization of multimodal medical images and segmentations viewing. It was produced in a scenario of need for the assistance of graphical tools for the study of deep-learning neural networks for segmenting medical images, and also as the final project of my Bachelor's degree on Computer Engineering in FEEC-Unicamp. The project was made using Tkinter, and designed in a MVC software architecture (Model View Controller).
The software was developed and tested in Ubuntu 22.04, and tested on Windows. The minimum required version of Python is 3.8, others required libs are described in requirements
It is suggested to create a Python environment with Conda:
$ conda create -n micview python=3.8
$ conda activate micview
To download the package, do:
$ pip install MICView
Or fork this GitHub Repository:
$ git clone https://github.com/MICLab-Unicamp/micview
$ cd micview
$ pip install .
You can use the API calling inside a terminal, and calling functions of micview inside your Python code, to call micview inside a terminal do:
micview
You can also pass arguments as the file path, the mask path, and if you want to see the image in the original shape or in a squared shape.
micview --input 'filepath' --mask 'maskpath' --resized
In the above example the GUI will open with an image and segmentation loaded, and in a squared shape. To call the GUI inside your Python code, you can use the functions:
micview.open() #Opens the GUI
micview.openfile(file='filepath', mask: optional ='maskpath', resized: optional = False (default)) #Opens the GUI with an image
Pay attention that if you call this functions inside a Python code, the programm will block in this line until you closes the MICView window, it occurs because the Tkinter mainloop is a blocking process.