CheerL/MPRI

MPRI

Overview

This project is a fast algorithm to calculate MPRI, which is inspired by this paper[1]

Requirements

It is requires Python3 and is compatible both in Windows and Linux.

Install

1. The most recommanded way is using pipenv, which is a modern virtualenv and package management tool.

   pip install pipenv

   After pipenv installed, you can use that to install required packages and a pure virtual enviroment.

   pipenv install

   Then entry this new virtual enviroment.

   pipenv shell

   Note that if you intall in this way, you need to entry the virtual enviroment by pipenv shell + python *** or pipenv run python *** every time you run it.

2. If you don't want to learn to use pipenv, you can also install packages in the traditional way.

   pip install -r requirements.txt

Usage

python main.py MODE --image/-i IMAGE
                    --lable/-l LABLE
                    [--output/-o OUTPUT]
                    [--data/d DATA]

Mode:
    calc:                   Calculate and output the pons area, midbrain area, MCP
                            width, SCP width and MPRI.
    seg:                    Segment MCP, SCP and save to output image but not
                            calculate MPRI, etc. Note that --output OUTPUT must be
                            given in this mode.
    both:                   Equal to runing both calc mode and seg mode.

Required arguments:
    --image/-i IMAGE:       The image path IMAGE to be segmented or calculated.
    --lable/-l LABLE:       The lable path LABLE of correspounding image
                            from AccuBrain.

Optional arguments:
    --output/-o OUTPUT:     The output path OUTPUT to save segmentation results.
                            It must be given if the mode is seg or both and does
                            not work if the mode is calc.
    --data/-d DATA:         The output path DATA to save calculation results. If
                            given, calculation results would be saved to DATA,
                            otherwise shown in the screen.

Reference

[1] Nigro S, Arabia G, Antonini A, Weis L, Marcante A, Tessitore A et al. Magnetic Resonance Parkinsonism Index: diagnostic accuracy of a fully automated algorithm in comparison with the manual measurement in a large Italian multicentre study in patients with progressive supranuclear palsy. European Radiology. 2016 Oct 19;1-11. https://doi.org/10.1007/s00330-016-4622-x