ResUNet++: An Advanced Architecture for Medical Image Segmentation

By Debesh Jha, Pia Helen Smedsrud, Michael Alexander Riegler, Dag Johansen, Thomas de Lange, Pål Halvorsen, Haavard D. Johansen. IEEE International Symposium on Multimedia, 2019

1,345 citations recorded by the source corpus.

Accurate computer-aided polyp detection and segmentation during colonoscopy examinations can help endoscopists resect abnormal tissue and thereby decrease chances of polyps growing into cancer. Towards developing a fully automated model for pixel-wise polyp segmentation, we propose ResUNet++, which is an improved ResUNet architecture for colonoscopic image segmentation. Our experimental evaluations show that the suggested architecture produces good segmentation results on publicly available datasets. Furthermore, ResUNet++ significantly outperforms U-Net and ResUNet, two key state-of-the-art deep learning architectures, by achieving high evaluation scores with a dice coefficient of 81.33%, and a mean Intersection over Union (mIoU) of 79.27% for the Kvasir-SEG dataset and a dice coefficient of 79.55%, and a mIoU of 79.62% with CVC-612 dataset.

Subfields: Computer Vision and Pattern Recognition, Oncology

View Paper: https://www.semanticscholar.org/paper/80f4c7c360d1150ba58c3bacf5c35718ebdd0c10

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Oncology

Citations in the corpus

Pattern Recognition (2018, positive): “In [14] , generative adversarial network (GAN) was used to generate invariant side-view gait images in normal clothing.”
Cross-level Feature Aggregation Network for Polyp Segmentation (2023, methodology): “Moreover, two variants of the UNet architecture, including ResUNet++ [9] , and UNet++ [10] , have been proposed for polyp segmentation which led to a promising performance.”
TransAttUnet: Multi-Level Attention-Guided U-Net With Transformer for Medical Image Segmentation (2021, neutral): “(cid:2) Multi-scalecontextapproaches: Meanwhile, severalmulti-scale context models are used as the major contenders, including Unet++ [9], R2U-Net [38] ResUNet [41], ResUNet++ [42], BCDU-Net [43], KiU-Net [37], and DoubleU-Net [36].”
DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentation (2022, methodology): “…0 .855±0.142 0.876±0.141 0.857±0.127 0.770±0.163 Attention-UNet [28] 0.940±0.048 0 .845±0.143 0.866±0.125 0.849±0.117 0.757±0.147 ResUNet++ [24] 0.934±0.051 0 .838±0.118 0.858±0.101 0 .840±0.086 0 .736±0.121 R2U-Net [ 44] 0.933±0.056 0 .852±0.122 0.831±0.136 0.834±0.112 0.744±0.128…”
Automatic Polyp Segmentation via Multi-scale Subtraction Network (2021, methodology): “Later, ResUNet++ [12] combines many advanced techniques such as residual computation [9], squeeze and excitation [10], atrous spatial pyramidal pooling [4], and attention mechanism to further improve performance.”
DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation (2020, background): “The accuracy achieved is now is high for both natural imaging [13] and medical imaging [17], [18], [20].”
Shallow Attention Network for Polyp Segmentation (2021, methodology): “U-Net++ [25] and ResUNet++ [11] further improve original U-Net by dense connection and better pretrained backbone, achieving promising segmentation performance.”
A Comprehensive Study on Colorectal Polyp Segmentation With ResUNet++, Conditional Random Field and Test-Time Augmentation (2021, methodology): “As the model trained on Kvasir-SEG and CVC-ClinicDB have perfect annotations, the model is good at predicting a perfect shaped mask.”
UACANet: Uncertainty Augmented Context Attention for Polyp Segmentation (2021, methodology): “ResUNet++ [14] construct a U-Net shape network with famous CNN modules including residual blocks from ResNet, Atrous Spatial Pyramid Pooling (ASPP) module from Deeplab and Squeeze and Excitation mechanism from SENet [12].”
TransFuse: Fusing Transformers and CNNs for Medical Image Segmentation (2021, methodology): “Following [5,10], we use mean Dice and mean IoU for quantitative evaluation.”
TransMorph: Transformer for unsupervised medical image registration (2021, neutral): “This limits the U-Net’s ability to perceive semantic information and model long-range relationships between points.”
Polyp-PVT: Polyp Segmentation with Pyramid Vision Transformers (2021, methodology): “ResUNet++ [24], ACSNet [44], PraNet [5], PNS-Net [59], in Tab.”
PraNet: Parallel Reverse Attention Network for Polyp Segmentation (2020, methodology): “, mean Dice and mean IoU) for quantitative evaluation, similar to [16,15].”