A Comprehensive Study on Colorectal Polyp Segmentation With ResUNet++, Conditional Random Field and Test-Time Augmentation

@article{Jha2021ACS,
  title={A Comprehensive Study on Colorectal Polyp Segmentation With ResUNet++, Conditional Random Field and Test-Time Augmentation},
  author={Debesh Jha and Pia Helen Smedsrud and Dag Johansen and Thomas de Lange and H{\aa}vard Dagenborg Johansen and P. Halvorsen and M. Riegler},
  journal={IEEE Journal of Biomedical and Health Informatics},
  year={2021},
  volume={25},
  pages={2029-2040}
}
Colonoscopy is considered the gold standard for detection of colorectal cancer and its precursors. Existing examination methods are, however, hampered by high overall miss-rate, and many abnormalities are left undetected. Computer-Aided Diagnosis systems based on advanced machine learning algorithms are touted as a game-changer that can identify regions in the colon overlooked by the physicians during endoscopic examinations, and help detect and characterize lesions. In previous work, we have… Expand
Advances in Artificial Intelligence to Reduce Polyp Miss Rates during Colonoscopy
TLDR
The Focus U-Net, a new Attention-based deep neural network architecture, which uses a Focus Gate to encourage learning of polyp location and associated features is introduced, showing the potential for Artificial Intelligence to provide fast and accurate polyp segmentation results for use during colonoscopy. Expand
Improving Generalizability in Polyp Segmentation using Ensemble Convolutional Neural Network
TLDR
This work investigates the feasibility of building a generalizable model for polyp segmentation using an ensemble of four MultiResUNet architectures, each trained on the combination of the different centered datasets provided by the challenge organizers. Expand
A comprehensive analysis of classification methods in gastrointestinal endoscopy imaging
TLDR
A comprehensive analysis of the Medico GI challenges is presented, revealing that the participants achieved an improvement on maximum Mathew correlation coefficient (MCC) from 82.68% in 2017 to 93.98% in 2018 and 95.20% in 2019 challenges, and a significant increase in computational speed over consecutive years. Expand
NanoNet: Real-Time Polyp Segmentation in Video Capsule Endoscopy and Colonoscopy
TLDR
This work proposes NanoNet, a novel architecture for the segmentation of video capsule endoscopy and colonoscopy images that allows real-time performance and has higher segmentation accuracy compared to other more complex ones. Expand
Polyp-PVT: Polyp Segmentation with Pyramid Vision Transformers
TLDR
The proposed model, named PolypPVT, effectively suppresses noises in the features and significantly improves their expressive capabilities, and achieves the new state-of-the-art performance. Expand
MSRF-Net: A Multi-Scale Residual Fusion Network for Biomedical Image Segmentation
TLDR
A novel architecture called MSRF-Net is proposed, which is specially designed for medical image segmentation tasks and provides improved results on different biomedical datasets, and is more generalizable as compared to state-of-the-art methods. Expand
Multi-Compound Transformer for Accurate Biomedical Image Segmentation
TLDR
A unified transformer network, termed Multi-Compound Transformer (MCTrans), which incorporates rich feature learning and semantic structure mining into a unified framework, and performs intraand inter-scale self-attention, rather than single-scale attention in previous works. Expand
Focus U-Net: A novel dual attention-gated CNN for polyp segmentation during colonoscopy
TLDR
The Focus U-Net is introduced, a novel dual attention-gated deep neural network, which combines efficient spatial and channel-based attention into a single Focus Gate module to encourage selective learning of polyp features and shows the potential for deep learning to provide fast and accurate polyp segmentation results for use during colonoscopy. Expand
Pyramidal Segmentation of Medical Images using Adversarial Training
TLDR
A novel approach based on learning to segment within several grids, which is introduced to U-Net and Pix2Pix architectures and suggests that segmentation at lower resolutions produces better results at the cost of less precision. Expand
...
1
2
...

References

SHOWING 1-10 OF 62 REFERENCES
ResUNet++: An Advanced Architecture for Medical Image Segmentation
TLDR
ResUNet++ is proposed, which is an improved ResUNet architecture for colonoscopic image segmentation, which significantly outperforms U-Net and Res UNet, two key state-of-the-art deep learning architectures, by achieving high evaluation scores. Expand
Training Data Enhancements for Robust Polyp Segmentation in Colonoscopy Images
TLDR
This work proposes a novel method to increase the quantity and variability of training images from a publicly available colonoscopy dataset by enrich the training data adding polyps to regions of nonpolypoid samples, creating automatically new data with their appropriate labels. Expand
Comparative Validation of Polyp Detection Methods in Video Colonoscopy: Results From the MICCAI 2015 Endoscopic Vision Challenge
TLDR
Results show that convolutional neural networks are the state of the art in polyp detection and it is also demonstrated that combining different methodologies can lead to an improved overall performance. Expand
A Benchmark for Endoluminal Scene Segmentation of Colonoscopy Images
TLDR
A comparative study is performed to show that FCNs significantly outperform, without any further postprocessing, prior results in endoluminal scene segmentation, especially with respect to polyp segmentation and localization. Expand
Ensemble of Instance Segmentation Models for Polyp Segmentation in Colonoscopy Images
TLDR
The results show that the transfer learning-based ensemble model significantly outperforms state-of-the-art methods for polyps diagnosis in colonoscopy images. Expand
Colorectal Segmentation Using Multiple Encoder-Decoder Network in Colonoscopy Images
TLDR
Evaluation of the proposed polyp segmentation method using the ETIS-LariPolypDB database shows that the proposed method outperforms state-of-the-art results. Expand
Development and validation of a deep-learning algorithm for the detection of polyps during colonoscopy
TLDR
A deep-learning algorithm can detect polyps in the colon in real time and with high sensitivity and specificity, according to validation studies with prospectively collected images and videos from colonoscopies performed in 1,138 patients. Expand
Colorectal Polyp Segmentation by U-Net with Dilation Convolution
TLDR
This paper proposes a novel end-to-end deep learning framework for the colorectal polyp segmentation and achieves state-of-the-art results on CVC-ClinicDB and ETIS-Larib Polyp DB. Expand
Fully convolutional neural networks for polyp segmentation in colonoscopy
TLDR
This work converts three established networks into a fully convolution architecture and fine-tuned their learned representations to the polyp segmentation task, achieving high segmentation accuracy and a detection precision and recall of 73.61% and 86.31%, respectively. Expand
Colorectal polyp segmentation using a fully convolutional neural network
  • Qiaoliang Li, Guangyao Yang, +7 authors Tianfu Wang
  • Computer Science
  • 2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)
  • 2017
TLDR
A new, end-to-end fully convolutional neural network structure for segmenting colorectal polyps, which can directly output a prediction map of the same size as the original image of the input network is proposed. Expand
...
1
2
3
4
5
...