Stack Overflow
Stack Overflow

Questions tagged [niftynet]

NiftyNet is a TensorFlow-based open-source convolutional neural networks (CNNs) platform for research in medical image analysis and image-guided therapy.

NiftyNet is a TensorFlow-based open-source convolutional neural networks (CNNs) platform for research in medical image analysis and image-guided therapy.

The code is available via GitLab, or you can quickly get started with the PyPI module available here.

Features

NiftyNet currently supports medical image segmentation and generative adversarial networks. NiftyNet is not intended for clinical use. Other features of NiftyNet include:

  • Easy-to-customise interfaces of network components
  • Sharing networks and pre-trained models
  • Support for 2-D, 2.5-D, 3-D, 4-D inputs
  • Efficient discriminative training with multiple-GPU support
  • Implementation of recent networks (HighRes3DNet, 3D U-net, V-net, DeepMedic)
  • Comprehensive evaluation metrics for medical image segmentation

NiftyNet is released under the Apache License, Version 2.0. Please see the LICENSE file in the NiftyNet source code repository for details.

If you use NiftyNet in your work, please cite Gibson, Li et al. 2017. The NiftyNet platform originated in software developed for Li et. al. 2017.

55 questions
0
votes
1 answer

Trouble using pre-trained model from niftynet

I want to do a 2 class segmentation using dense_vnet model available on niftynet which originally does a 9 class segmentation I tried to retrain only the last layer by making changes in config file according to this suggestion: HOw to fine tune…
seek
  • 3
  • 4
0
votes
1 answer

How to fine tune niftynet pre trained model for custom data

I want to use niftynet pretrained segmentation model for segmenting custom data. I downloaded the pre trained weights and and modified model_dir path to downloaded one. However when I run python3 net_segment.py train -c…
user1996075
  • 11
  • 1
  • 4
0
votes
1 answer

niftynet filename in T1 is empty

ValueError raised I've been using NiftyNet and it has been working before, but recently I've encountered this problem with it not recognizing the T1 or T2 images I think? There is a valueerror raised because the filename list column passed to [T1]…
bujian
  • 152
  • 8
0
votes
1 answer

output of inference section in niftynet run

I can't seem to find where my inference models are, I have the .csv files and can see the loss but I can't find the label files or images themselves. This is my inference section: border = (44,44,44) output_interp_order = 0 By the documentation I…
bujian
  • 152
  • 8
0
votes
1 answer

NiftyNet Interpolation order

In NiftyNet framework - what is the purpose of interp_order param in config file? There is very short explanation in documentation which is not clear enough. Why there is usually set to 0 in label section, and to 3 in image section.
manza
  • 322
  • 1
  • 11
0
votes
1 answer

niftynet multi-class 3D segmentation with dense vnet

Neural network newbie here. I've been testing Niftynet and achieved decent single-class 3D segmentation predictions on an own MRI data set with dense_vnet. However, I ran out of luck when I tried to add a second label. The network seems to spot the…
fraxyv
  • 1
0
votes
1 answer

How to change output format for Niftynet?

Is it possible to set output format in configuration file for Niftynet? e.g. I would like to have my output as png. My output is currently in .nii.gz format. Thank you.
manza
  • 322
  • 1
  • 11
0
votes
1 answer

Error in importing NiftyNet with Tensorflow 1.9

I installed the package NiftyNet 0.3.0 with Python 2.7.5 on CentOS Linux 7.5. Tensorflow 1.9 was installed a priori. When I import NiftyNet, I got the following error message. $ python Python 2.7.5 (default) [GCC 4.8.5 20150623 (Red Hat 4.8.5-28)]…
Lucas
  • 920
  • 1
  • 6
  • 9
0
votes
1 answer

Don't know how to generate sampling locations:

How are the dimensions generated in the uniform sampler? I tried debugging the image size it seems it works for some of the iterations it does not for others. Any idea how to fix this. My configuration is given below : [CUSTOM] num_classes:…
Yashovardhan Chaturvedi
  • 3
  • 1
0
votes
1 answer

Generalised dice loss only learning one class

I am experimenting with the generalised dice loss implemented in niftynet to segment MRI volumes containing 4 classes (1 background 3 regions of interest) using the V-Net. I tried to format the labels in 2 ways: spatial dimensions only with 0 being…
LRier
  • 5
  • 3
0
votes
1 answer

NiftyNet net_segment has no attribute '_logger'

When I try to do the quickstart instructions for NiftyNet installed from PyPI I get the following error when running net_segment: net_segment inference -c ~/niftynet/extensions/dense_vnet_abdominal_ct/config.ini WARNING:tensorflow:SimpleITK adapter…
tubbytoad
  • 1
  • 2
0
votes
1 answer

NiftyNet ValueError: Dimension -1 must be >= 0

I want to train a VNet for a segmentation task. For a simple "getting started" example, I have one CT and its corresponding segmentation image. I get a ValueError when running the train command: $ net_segment -c ~/git/vesseg/learning/config.ini…
fepegar
  • 595
  • 5
  • 15
0
votes
1 answer

I am struggling to implement multiple channel images

I am trying to train a vnet architecture on a set of 3D MRI volumes and I would like to try adding a channel with a different echo time to each image, effectively making it 4D. I saw that niftynet allows for 4D image input but I received error…
LRier
  • 5
  • 3
0
votes
1 answer

Error running NiftyNet Promise12 demo on multiple GPUs

I am trying to duplicate the the Promise12 demo in NiftyNet. Installed NiftyNet 0.2.2 with TensorFlow 1.3 GPU version. After downloading the PROMISE12 data and unzipping, I modified NiftyNet's promise12_demo_train_config.ini to reflect the path to…
Daniel Blezek
  • 4,539
  • 1
  • 19
  • 20
0
votes
1 answer

NiftyNet Input/Output Data Type

What data type does NiftyNet expect as input? I'm aware that NiftyNet reads and outputs data in the NIfTI file format, but I'm unclear about the typing of the individual data (e.g. 16 sint or 32 sint). Is it documented anywhere? I only ask because…
Alex
  • 301
  • 3
  • 7
1 2
3
4