NiftyNet: a deep-learning platform for medical imaging

By Eli Gibson, Wenqi Li, C. Sudre, Lucas Fidon, D. I. Shakir, Guotai Wang, Z. Eaton-Rosen, Robert J. Gray, Tom Doel, Yipeng Hu and 7 more. 2017

565 citations recorded by the source corpus.

Highlights • An open-source platform is implemented based on TensorFlow APIs for deep learning in medical imaging domain.• A modular implementation of the typical medical imaging machine learning pipeline facilitates (1) warm starts with established pre-trained networks, (2) adapting existing neural network architectures to new problems, and (3) rapid prototyping of new solutions.• Three deep-learning applications, including segmentation, regression, image generation and representation learning, are presented as concrete examples illustrating the platform’s key features.

Subfields: Biomedical Engineering, Biophysics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Biophysics

In Computer Vision and Pattern Recognition

Citations in the corpus

An overview of deep learning in medical imaging focusing on MRI (2018, background): “for medical image analysis and image-guided therapy [320,321] http://niftynet.”
Deep Learning Applications in Medical Image Analysis (2018, methodology): “io) [18] is a useful open source framework that contains many machine learning algorithms, released under an Apache License.”
TorchIO: A Python library for efficient loading, preprocessing, augmentation and patch-based sampling of medical images in deep learning (2020, methodology): “Similar to NiftyNet and DLTK, they include implementation of network architectures, transforms, and higher-level features to perform training and inference.”
GANs for Medical Image Analysis (2018, background): “The interested reader is also refered to the NiftyNet framework [87], in which this conditional GAN is contained.”
Models Genesis: Generic Autodidactic Models for 3D Medical Image Analysis (2019, methodology): “vised manner with abundant unlabeled image datasets, demanding zero expert annotation eort. Consequently, Models Genesis are very dierent from traditional supervised transfer learning from ImageNet [7,9], which oers 4 Appendix can be found in the full version attinyurl.com/ModelsGenesisFullVersion 4”
Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernels (2019, methodology): “We used Tensorﬂow and implemented our models within the NiftyNet framework [41].”
Transformers in Medical Imaging: A Survey (2022, background): “, GPUs) and purpose-built open source software libraries [64]–[66].”
Automatic Brain Tumor Segmentation Using Cascaded Anisotropic Convolutional Neural Networks (2017, methodology): “Our networks were implemented in Tensorflow(2) [1] using NiftyNet(34) [10].”
MONAI: An open-source framework for deep learning in healthcare (2022, background): “With the increasing popularity of AI in healthcare, several open-source healthcare-speciﬁc frameworks have also been developed by academic groups (NiftyNet [8], DLTK [9], DeepNeuro [10]) and industry teams (NVIDIA Clara, Microsoft Project InnerEye).”
DLTK: State of the Art Reference Implementations for Deep Learning on Medical Images (2017, background): “While other packages, such as NiftyNet [3], build on a predeﬁned application structure, we emphasise an API level prioritising experimentation by exposing access to low-level operations.”
Short Acquisition Time PET/MR Pharmacokinetic Modelling Using CNNs (2018, methodology): “The network was implemented in NiftyNet [4] using the ‘highresnet’ convolutional neural network with 20 convolutional layers [5], which uses a stack of residual dilated convolutions with increasingly large dilation factors.”
Deep learning in generating radiology reports: A survey (2020, unknown): “To allow practitioners to rapidly implement DL solutions for image analysis tasks, NiftyNet4 [49] features an open source framework for many medical imaging CNN algorithms under the 1 https://www.nlm.nih.gov/research/umls/knowledge_sources/ metathesaurus.”