Skip to content

Prostate lesion classification using Deep Convolutional Neural Networks

License

Notifications You must be signed in to change notification settings

piotrsobecki/PCa-CNNs

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Prostate lesion classification using Deep Convolutional Neural Networks

We propose three novel VGG based architectures for prostate lesion classification (see Architectures module):

  • CNN_VGG_SIMPLE with three subnetworks for mpMRI modalities
  • CNN_VGG_MODALITIES with mixture of experts architecture (each modality is capable of making predictions)
  • CNN_VGG_PIRADS with a-priori knowledge embedding in mixture of experts architecture

This project uses data from ProstateX1 competition.

Directory data/ProstateX contains trail dataset, a preselected cases from ProstateX1 Challenge. In order to perform analysis on bigger dataset - download the data from original sources and extract them to directory following the trail directory as an example.

Installation

Install Anaconda 4.3.1 (with Python 3.6.0) - https://repo.continuum.io/archive/ Execute setup.bat (or copy content to terminal on unix) Note that requirements.txt may not be complete - if so, please add missing requirements to the text file and make pull request

Running

After installation, to run the tests:

  • Run augment_data.py to augment and save locally the data
  • Modify constants.py to match your experiment parameters
  • Run test.py to train the model using CV and verbose logging

About

Prostate lesion classification using Deep Convolutional Neural Networks

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages