PLN: Parasitic-Like Network for Barely Supervised Medical Image Segmentation

Li, Shumeng; Cai, H.; Qi, Lei; Yu, Qian; Shi, Yinghuan

doi:10.1109/tmi.2022.3211188

Cited by 10 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Semi-supervise learning (SSL) relies on leveraging the unlabeled samples to boost the performance of model trained on the labeled data (Yang et al 2019;Van Engelen and Hoos 2020;Li et al 2022;Xing et al 2022Xing et al , 2023Zhao et al 2023;Duan et al 2023;Yang et al 2023). For semi-supervised image classification, recently proposed SSL approaches unify pseudo-labeling technique to assign pseudo-labels to unlabeled data, enabling their use in the training.…”

Section: Related Workmentioning

confidence: 99%

Roll with the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning

Duan,

Zhao,

et al. 2024

AAAI

View full text Add to dashboard Cite

While semi-supervised learning (SSL) has yielded promising results, the more realistic SSL scenario remains to be explored, in which the unlabeled data exhibits extremely high recognition difficulty, e.g., fine-grained visual classification in the context of SSL (SS-FGVC). The increased recognition difficulty on fine-grained unlabeled data spells disaster for pseudo-labeling accuracy, resulting in poor performance of the SSL model. To tackle this challenge, we propose Soft Label Selection with Confidence-Aware Clustering based on Class Transition Tracking (SoC) by reconstructing the pseudo-label selection process by jointly optimizing Expansion Objective and Shrinkage Objective, which is based on a soft label manner. Respectively, the former objective encourages soft labels to absorb more candidate classes to ensure the attendance of ground-truth class, while the latter encourages soft labels to reject more noisy classes, which is theoretically proved to be equivalent to entropy minimization. In comparisons with various state-of-the-art methods, our approach demonstrates its superior performance in SS-FGVC. Checkpoints and source code are available at https://github.com/NJUyued/SoC4SS-FGVC.

show abstract

Section: Related Workmentioning

confidence: 99%

Roll with the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning

Duan,

Zhao,

et al. 2024

AAAI

View full text Add to dashboard Cite

show abstract

“…In fact, segmentation targets in adjacent slices of 3D volume are highly similar in both appearance and location, leading it redundant to label every slice. Although the sparse annotation is discussed in recent work [18], we notice these conventional methods still neglect the complementary views between different directions in 3D volume.…”

Section: Introductionmentioning

confidence: 99%

“…Weakly-supervised segmentation methods usually utilize weak annotations, e.g., image-level label [16,17], scribble [20,21], point [3] or partial slices [5,18]. Unfortunately, most of them are either difficult to distinguish some fuzzy boundaries or with additional large computational burden [15].…”

Section: Introductionmentioning

confidence: 99%

“…We believe our annotation way has two merits: 1) it could largely force the model to learn from complementary views with two diversely initialized labeled slices, 2) it helps greatly reduce the label costs with fully utilizing the inter-slice similarity. Following very recent work [18], we name the setting as Barely-supervised Segmentation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Orthogonal Annotation Benefits Barely-supervised Medical Image Segmentation

Cai¹,

Li²,

Liu³

et al. 2023

Preprint

View full text Add to dashboard Cite

Recent trends in semi-supervised learning have significantly boosted the performance of 3D semi-supervised medical image segmentation. Compared with 2D images, 3D medical volumes involve information from different directions, e.g., transverse, sagittal, and coronal planes, so as to naturally provide complementary views. These complementary views and the intrinsic similarity among adjacent 3D slices inspire us to develop a novel annotation way and its corresponding semi-supervised model for effective segmentation. Specifically, we firstly propose the orthogonal annotation by only labeling two orthogonal slices in a labeled volume, which significantly relieves the burden of annotation. Then, we perform registration to obtain the initial pseudo labels for sparsely labeled volumes. Subsequently, by introducing unlabeled volumes, we propose a dual-network paradigm named Dense-Sparse Cotraining (DeSCO) that exploits dense pseudo labels in early stage and sparse labels in later stage and meanwhile forces consistent output of two networks. Experimental results on three benchmark datasets validated our effectiveness in performance and efficiency in annotation. For example, with only 10 annotated slices, our method reaches a Dice up to 86.93% on KiTS19 dataset. Our code and models are available at https://github.com/HengCai-NJU/DeSCO.

show abstract

“…Weak annotations 12 aim at alleviating these issues. Li et al 13 propose sparse annotations only labeling one slice per volume. Cai et al 14 propose annotating multiple perpendicular slices to maximize the data distribution coverage.…”

Section: Introductionmentioning

confidence: 99%

Learning 3D aortic root assessment based on sparse annotations

Brosig,

Krüger,

Wamala

et al. 2024

Medical Imaging 2024: Computer-Aided Diagnosis

View full text Add to dashboard Cite

Analyzing the anatomy of the aorta and left ventricular outflow tract (LVOT) is crucial for risk assessment and planning of transcatheter aortic valve implantation (TAVI). We propose 2D cross-sectional annotation and point cloud-based surface reconstruction to train a fully automatic 3D segmentation network for the aortic root and the LVOT. Our sparse annotation scheme enables easy and fast training data generation for tubular structures like the aortic root. Based on this annotation concept, we trained a 3D segmentation model that achieves a Dice similarity coefficient (DSC) of 0.9 and an average surface distance (ASD) of 0.96 mm. In addition, we show that our fully automatic segmentation approach facilitates reproducible and quantifiable measurements for TAVI planning. Our approach achieves an aortic annulus maximum diameter difference between prediction and annotation of 0.45 mm (inter-observer variance: 0.25 mm).

show abstract

PLN: Parasitic-Like Network for Barely Supervised Medical Image Segmentation

Cited by 10 publications

References 33 publications

Roll with the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning

Roll with the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning

Orthogonal Annotation Benefits Barely-supervised Medical Image Segmentation

Learning 3D aortic root assessment based on sparse annotations

Contact Info

Product

Resources

About