Automated Design of Deep Learning Methods for Biomedical Image Segmentation

By Fabian Isensee, Paul F. Jager, Simon A. A. Kohl, Jens Petersen, Klaus Hermann Maier-Hein. 2019

325 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/4692b1eefa9bc43c23075412aeb4181a81e0bbe3

How researchers see this paper

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

1. Deng-Ping Fan
2. Tao Zhou
3. Ling Shao
4. Wenguan Wang
5. Ge-Peng Ji
6. Jianbing Shen
7. Huan Fu
8. Hong-Yu Zhou

Citations in the corpus

• TransFuse: Fusing Transformers and CNNs for Medical Image Segmentation (2021, methodology): “Following the setting in [12], we reshape all MRI slices to 320× 320, and independently normalize each volume using z-score normalization.”
• The state of the art in kidney and kidney tumor segmentation in contrast-enhanced CT imaging: Results of the KiTS19 challenge (2019, methodology): “These preprocessing constants were selected through a meta-learning process described in Isensee et al. (2019a).”
• CoTr: Efficiently Bridging CNN and Transformer for 3D Medical Image Segmentation (2021, methodology): “Following [12], we first truncated the HU values of each scan using the range of [ − 958 , 327] In the training stage, we randomly cropped sub-volumes of size 48 × 192 × 192 from CT scans as the input.”
• nnFormer: Interleaved Transformer for Volumetric Segmentation (2021, background): “The numbers of heads of multi-head self-attention used in different encoder stages are [6, 12, 24, 48] and [3, 6, 12, 24] on Synapse and ACDC, respectively.”
• Inf-Net: Automatic COVID-19 Lung Infection Segmentation From CT Images (2020, background): “Parallel Partial Decoder: Several existing medical image segmentation networks segment interested organs/lesions using all high- and low-level features in the encoder branch [57], [58], [62]–[65].”
• AbdomenCT-1K: Is Abdominal Organ Segmentation a Solved Problem? (2020, methodology): “In particular, nnU-Net (no-new-U-Net) [33] has been proposed to automatically adapt preprocessing strategies and network architectures (i.”
• A review of deep learning in medical imaging: Imaging traits, technology trends, case studies with progress highlights, and future promises (2020, methodology): “Abnormal areas in the lung were segmented using a 3-D U-net built with the nnU-Net framework [119] in a cross-validated fashion with 108 scans and corresponding reference delineations to segment ground-glass opacities and consolidation in the lungs.”
• VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-detector CT Images (2020, methodology): “Second, which form the majority, are the approaches that perform patch-wise segmentation in 3D using architrectures such as 3D U-Net (Çiçek et al., 2016), V-Net (Milletari et al., 2016), or nnU-Net (Isensee et al., 2019).”