Scribble-Supervised Medical Image Segmentation via Dual-Branch Network and Dynamically Mixed Pseudo Labels Supervision

Luo, Xiangde; Hu, Minhao; Liao, Wenjun; Zhai, Shuwei; Song, Tao; Wang, Guotai; Zhang, Shaoting

doi:10.1007/978-3-031-16431-6_50

Cited by 54 publications

(19 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Weakly-supervised segmentation aims to reduce the labeling costs by training segmentation models on data annotated with coarse granularity [18]. Among various formats of sparse annotations, scribble is recognized as the most flexible and versatile one that can be used to annotate complex structures [23], [27]. Existing scribble-supervised segmentation methods fall into two main categories.…”

Section: B Weakly-supervised Segmentationmentioning

confidence: 99%

“…Weakly supervised learning (WSL) attempts to alleviate the annotation issue from another perspective by performing sparsely-grained (i.e., point-, scribble-, bounding boxwise) supervision and attains promising performance [18]- [22]. Compared with either point or bounding box, scribble is a relatively more flexible and generalizable form of sparse annotation that can be used to annotate complex structures [23]. Existing scribble-supervised segmentation methods mainly fall into two categories.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

YoloCurvSeg: You Only Label One Noisy Skeleton for Vessel-style Curvilinear Structure Segmentation

Li¹,

Peng²,

He³

et al. 2023

Preprint

View full text Add to dashboard Cite

<p>Weakly-supervised learning (WSL) has been proposed to alleviate the conflict between data annotation cost and model performance through employing sparsely-grained (i.e., point-, box-, scribble-wise) supervision and has shown promising performance, particularly in the image segmentation field. However, it is still a very challenging problem due to the limited supervision, especially when only a small number of labeled samples are available. Additionally, almost all existing WSL segmentation methods are designed for star-convex structures which are very different from curvilinear structures such as vessels and nerves. In this paper, we propose a novel sparsely annotated segmentation framework for curvilinear structures, named YoloCurvSeg, based on image synthesis. A background generator delivers image backgrounds that closely match real distributions through inpainting dilated skeletons. The extracted backgrounds are then combined with randomly emulated curves generated by a Space Colonization Algorithm-based foreground generator and through a multilayer patch-wise contrastive learning synthesizer. In this way, a synthetic dataset with both images and curve segmentation labels is obtained, at the cost of only one or a few noisy skeleton annotations.</p> <p>Finally, a segmenter is trained with the generated dataset and possibly an unlabeled dataset. The proposed YoloCurvSeg is evaluated on four publicly available datasets (OCTA500, CORN, DRIVE and CHASEDB1) and the results show that YoloCurvSeg outperforms state-of-the-art WSL segmentation methods by large margins. With only one noisy skeleton annotation (respectively 0.14%, 0.03%, 1.40%, and 0.65% of the full annotation), YoloCurvSeg achieves more than 97% of the fully-supervised performance on each dataset. Code and datasets will be released at https://github.com/llmir/YoloCurvSeg.</p>

show abstract

Section: B Weakly-supervised Segmentationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

YoloCurvSeg: You Only Label One Noisy Skeleton for Vessel-style Curvilinear Structure Segmentation

Li¹,

Peng²,

He³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…To obtain high-quality pseudo labels and update it throughout the training process, Luo et al [29] proposed to mix the predictions from dual-branch network as auxiliary pseudo label. This approach has achieved promising results on cardiac segmentation, but still susceptible to inaccurate supervisions, especially on more challenging tasks with irregular objects.…”

Section: Weakly Supervised Segmentationmentioning

confidence: 99%

“…We first implemented the PCE loss (L pce ) as a baseline method (referred to PCE). Then, we implemented four state-of-the-art (SOTA) scribble supervised segmentation methods, i.e., WSL4 [29], GatedCRF [31], CycleMix [52], and ShapePU [53] to run the same experiments. We cited the ACDC and PPSS results reported in [42] for the MAAG method, which is also a SOTA method for this task.…”

Section: Performance and Comparisonsmentioning

confidence: 99%

“…Existing scribble learning approaches mainly aim to reconstruct complete labels from scribbles, and use the generated pseudo labels for model training. These works include 1) label expansion strategies that assume the pixels with similar features are likely to be in the same category [16,27], and 2) ensemble methods that generate labels by fusing several independent predictions [29]. These methods could be susceptible to the label noises introduced by imprecise segmentation proposals.…”

mentioning

confidence: 99%

See 1 more Smart Citation

ZScribbleSeg: Zen and the Art of Scribble Supervised Medical Image Segmentation

Zhang¹,

Zhuang²

2023

Preprint

View full text Add to dashboard Cite

Curating a large scale fully-annotated dataset can be both labour-intensive and expertise-demanding, especially for medical images.To alleviate this problem, we propose to utilize solely scribble annotations for weakly supervised segmentation. Existing solutions mainly leverage selective losses computed solely on annotated areas and generate pseudo gold standard segmentation by propagating labels to adjacent areas. However, these methods could suffer from the inaccurate and sometimes unrealistic pseudo segmentation due to the insufficient supervision and incomplete shape features. Different from previous efforts, we first investigate the principle of "good scribble annotations", which leads to efficient scribble forms via supervision maximization and randomness simulation. Furthermore, we introduce regularization terms to encode the spatial relationship and shape prior, where a new formulation is developed to estimate the mixture ratios of label classes. These ratios are critical in identifying the unlabeled pixels for each class and correcting erroneous predictions, thus the accurate estimation lays the foundation for the incorporation of spatial prior. Finally, we integrate the efficient scribble supervision with the prior into a unified framework, denoted as ZScribbleSeg, and apply the method to multiple scenarios. Leveraging only scribble annotations, ZScribbleSeg set new state-of-thearts on four segmentation tasks using ACDC, MSCMRseg, MyoPS and PPSS datasets.

show abstract

Light mixed‐supervised segmentation for 3D medical image data

Yang,

Tan,

Tegzes

et al. 2023

Medical Physics

View full text Add to dashboard Cite

BackgroundAccurate 3D semantic segmentation models are essential for many clinical applications. To train a model for 3D segmentation, voxel‐level annotation is necessary, which is expensive to obtain due to laborious work and privacy protection. To accurately annotate 3D medical data, such as MRI, a common practice is to annotate the volumetric data in a slice‐by‐slice contouring way along principal axes.PurposeIn order to reduce the annotation effort in slices, weakly supervised learning with a bounding box (Bbox) was proposed to leverage the discriminating information via a tightness prior assumption. Nevertheless, this method requests accurate and tight Bboxes, which will significantly drop the performance when tightness is not held, that is when a relaxed Bbox is applied. Therefore, there is a need to train a stable model based on relaxed Bbox annotation.MethodsThis paper presents a mixed‐supervised training strategy to reduce the annotation effort for 3D segmentation tasks. In the proposed approach, a fully annotated contour is only required for a single slice of the volume. In contrast, the rest of the slices with targets are annotated with relaxed Bboxes. This mixed‐supervised method adopts fully supervised learning, relaxed Bbox prior, and contrastive learning during the training, which ensures the network exploits the discriminative information of the training volumes properly. The proposed method was evaluated on two public 3D medical imaging datasets (MRI prostate dataset and Vestibular Schwannoma [VS] dataset).ResultsThe proposed method obtained a high segmentation Dice score of 85.3% on an MRI prostate dataset and 83.3% on a VS dataset with relaxed Bbox annotation, which are close to a fully supervised model. Moreover, with the same relaxed Bbox annotations, the proposed method outperforms the state‐of‐the‐art methods. More importantly, the model performance is stable when the accuracy of Bbox annotation varies.ConclusionsThe presented study proposes a method based on a mixed‐supervised learning method in 3D medical imaging. The benefit will be stable segmentation of the target in 3D images with low accurate annotation requirement, which leads to easier model training on large‐scale datasets.

show abstract

Scribble-Supervised Medical Image Segmentation via Dual-Branch Network and Dynamically Mixed Pseudo Labels Supervision

Cited by 54 publications

References 27 publications

YoloCurvSeg: You Only Label One Noisy Skeleton for Vessel-style Curvilinear Structure Segmentation

YoloCurvSeg: You Only Label One Noisy Skeleton for Vessel-style Curvilinear Structure Segmentation

ZScribbleSeg: Zen and the Art of Scribble Supervised Medical Image Segmentation

Light mixed‐supervised segmentation for 3D medical image data

Contact Info

Product

Resources

About