Metadata-Version: 2.1
Name: monte-carlo-analysis
Version: 0.0.1
Summary: Package that analyses the output of a monte-carlo dropout network
Home-page: UNKNOWN
Author: Robin Camarasa
Author-email: r.camarasa@erasmusmc.nl
License: UNKNOWN
Platform: UNKNOWN
Description-Content-Type: text/markdown
Requires-Dist: numpy
Requires-Dist: pytest
Requires-Dist: joblib
Requires-Dist: tox
Requires-Dist: numba
Requires-Dist: sklearn
Requires-Dist: pandas
Requires-Dist: matplotlib
Requires-Dist: seaborn

# Monte Carlo Analysis

This python package analyses the output of a monte-carlo dropout network. The method to analyse the different uncertainty maps are described in the following article:
```
@inproceedings{camarasa2020quantitative,
    title={{Quantitative Comparison of Monte-Carlo Dropout Uncertainty Measures for Multi-Class Segmentation}},
    author={Camarasa, Robin and Bos, Daniel and Hendrikse, Jeroen and Nederkoorn, Paul and Kooi, Eline and van der Lugt, Aad and de Bruijne, Marleen},
    note={Uncertainty for Safe Utilization of Machine Learning in Medical Imaging (UNSURE) workshop of MICCAI conference},
    year={2020}
    }
```

A talk a given at [UNSURE](https://unsuremiccai.github.io/) a workshop of [MICCAI conference](https://www.miccai2020.org/en/) is available at this link on Youtube:
[![](http://img.youtube.com/vi/Uw2r1JRSxgE/0.jpg)](http://www.youtube.com/watch?v=Uw2r1JRSxgE)

# Usage

A documentation of the code can be found on the website [ReadTheDoc](https://monte-carlo-analysis.readthedocs.io/en/latest/) and a tutorial can be found on [google colab](https://colab.research.google.com/drive/1Sw5mnIiho8YdEkVykrRTH12KjD51OuNp?usp=sharing). In the process of re-writting the article the name of the classes might not match the code anymore but this will be changed as soon as possible.

Even if the presented application in the talk and the article is made on 3D MR images of carotid artery, the code can be applied to any kind of n-d imaging data as long as they are converted to numpy arrays.

# Requirements

Python 3.8

# Installation

Install via pip
``` bash
$ python3 -m pip install monte_carlo_analysis
```

# Todo

- Adapt the code to numba to decrease the execution time

# Authors

- [Robin Camarasa](https://github.com/RobinCamarasa) : [r.camarasa@erasmusmc.nl](mailto:r.camarasa@erasmusmc.nl)


