+TITLE: Unsupervised lesion detection with multi view MRI and autoencoders

+AUTHOR: Linas Vidziunas & Ørjan Thoresen

+EMAIL: linasvidz@gmail.com

+DESCRIPTION: Using UIS's GPU cluster for training autoencoders to detect and classify cancer tumors

+KEYWORDS: tensorflow, lesion detection, autoencoder

+LANGUAGE: en

** Running on GPU cluster using SLURM * Files ** ~setup.sh~ Is responsible to /create an enviroment/ with the correct requirements without using GPUs.

*** ~run.sh~ Is responsible to activate the previously created enviroment and run the python script using GPUs*.

**** ~requirements.txt~ Generated using [[https://github.com/Damnever/pigar][pigar]]. Includes a list of dependencies for this project.. This file will get used, in the previously mentioned ~setup.sh~, to install the required dependencies on the server.

*** Usage Install dependencies on the remote server

+begin_src sh

sbatch setup.sh tail -f setup.out

+end_src

Run the deep neural network

+begin_src sh

sbatch run.sh tail -f cAE_01.out

+end_src

*** Resources

• [[https://github.com/tlinjordet/gpu_ux_uis][gpu_ux_uis]]