Scipediacontent: Scipediacontent moved page Draft Content 683881866 to Dam et al 2019a

2021-01-21T14:43:03Z

Scipediacontent moved page Draft Content 683881866 to Dam et al 2019a

Scipediacontent: Created page with " == Abstract == Many recent medical segmentation systems rely on powerful deep learning models to solve highly specific tasks. To maximize performance, it is standard practic..."

2021-01-21T14:43:00Z

Created page with " == Abstract == Many recent medical segmentation systems rely on powerful deep learning models to solve highly specific tasks. To maximize performance, it is standard practic..."

New page

== Abstract ==

Many recent medical segmentation systems rely on powerful deep learning models to solve highly specific tasks. To maximize performance, it is standard practice to evaluate numerous pipelines with varying model topologies, optimization parameters, pre- & postprocessing steps, and even model cascades. It is often not clear how the resulting pipeline transfers to different tasks. We propose a simple and thoroughly evaluated deep learning framework for segmentation of arbitrary medical image volumes. The system requires no task-specific information, no human interaction and is based on a fixed model topology and a fixed hyperparameter set, eliminating the process of model selection and its inherent tendency to cause method-level over-fitting. The system is available in open source and does not require deep learning expertise to use. Without task-specific modifications, the system performed better than or similar to highly specialized deep learning methods across 3 separate segmentation tasks. In addition, it ranked 5-th and 6-th in the first and second round of the 2018 Medical Segmentation Decathlon comprising another 10 tasks. The system relies on multi-planar data augmentation which facilitates the application of a single 2D architecture based on the familiar U-Net. Multi-planar training combines the parameter efficiency of a 2D fully convolutional neural network with a systematic train- and test-time augmentation scheme, which allows the 2D model to learn a representation of the 3D image volume that fosters generalization.

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1911.01764 http://arxiv.org/abs/1911.01764]

* [http://arxiv.org/pdf/1911.01764 http://arxiv.org/pdf/1911.01764]

* [http://link.springer.com/content/pdf/10.1007/978-3-030-32245-8_4 http://link.springer.com/content/pdf/10.1007/978-3-030-32245-8_4],
: [http://dx.doi.org/10.1007/978-3-030-32245-8_4 http://dx.doi.org/10.1007/978-3-030-32245-8_4] under the license http://www.springer.com/tdm

* [https://dblp.uni-trier.de/db/journals/corr/corr1911.html#abs-1911-01764 https://dblp.uni-trier.de/db/journals/corr/corr1911.html#abs-1911-01764],
: [https://link.springer.com/chapter/10.1007%2F978-3-030-32245-8_4 https://link.springer.com/chapter/10.1007%2F978-3-030-32245-8_4],
: [https://arxiv.org/abs/1911.01764 https://arxiv.org/abs/1911.01764],
: [https://arxiv.org/pdf/1911.01764v1 https://arxiv.org/pdf/1911.01764v1],
: [https://doi.org/10.1007/978-3-030-32245-8_4 https://doi.org/10.1007/978-3-030-32245-8_4],
: [https://www.arxiv-vanity.com/papers/1911.01764 https://www.arxiv-vanity.com/papers/1911.01764],
: [https://rd.springer.com/chapter/10.1007/978-3-030-32245-8_4 https://rd.springer.com/chapter/10.1007/978-3-030-32245-8_4],
: [https://academic.microsoft.com/#/detail/2979649449 https://academic.microsoft.com/#/detail/2979649449]

← Older revision	Revision as of 14:43, 21 January 2021
(No difference)

Dam et al 2019a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 683881866 to Dam et al 2019a

Scipediacontent: Created page with " == Abstract == Many recent medical segmentation systems rely on powerful deep learning models to solve highly specific tasks. To maximize performance, it is standard practic..."