# vnet.pytorch

**Repository Path**: frankytom/vnet.pytorch

## Basic Information

- **Project Name**: vnet.pytorch
- **Description**: A PyTorch implementation for V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation
- **Primary Language**: Python
- **License**: BSD-3-Clause
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 1
- **Forks**: 0
- **Created**: 2020-05-15
- **Last Updated**: 2023-04-11

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# A PyTorch implementation of V-Net

Vnet is a [PyTorch](http://pytorch.org/) implementation of the paper
[V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation](https://arxiv.org/abs/1606.04797)
by Fausto Milletari, Nassir Navab, and Seyed-Ahmad Ahmadi. Although this implementation is still a work  in progress, 
I'm seeing a respectable 0.355% test error rate and a Dice coefficient of .9825 segmenting lungs from the LUNA16 data
set after 249 epochs. The official implementation is available in the [faustomilletari/VNet](https://github.com/faustomilletari/VNet)
repo on GitHub.

![](images/diagram.png)

This implementation relies on the LUNA16 loader and dice loss function from
the [Torchbiomed](https://github.com/mattmacy/torchbiomed) package.

## Differences with the official version
This version uses batch normalization and dropout. Lung volumes in CTs are ~10% of the scan volume - a not too
unreasonable class balance. For this particular test application I've added the option of using NLLoss instead
of the Dice Coefficient.

![](images/lung-loss-error.png)


## What does the PyTorch compute graph of Vnet look like?

You can see the compute graph [here](images/vnet.png),
which I created with [make_graph.py](https://github.com/mattmacy/vnet.pytorch/blob/master/make_graph.py),
which I copied from [densenet.pytorch](https://github.com/bamos/densenet.pytorch) which in turn was
copied from [Adam Paszke's gist](https://gist.github.com/apaszke/01aae7a0494c55af6242f06fad1f8b70).

### Credits

The train.py script was derived from the one in the [densenet.pytorch](https://github.com/bamos/densenet.pytorch)
repo.