Weakly-Supervised Convolutional Neural Networks for Vessel Segmentation in Cerebral Angiography

Vepa, Arvind; Choi, Andrew; Nakhaei, Noor; Lee, Wonjun; Stier, Noah; Vu, Andrew; Jenkins, Greyson; Yang, Xin; Shergill, Manjot; Desphy, Moira; Delao, Kevin; Garduno, Cristopher; Nelson, L. D.; Liu, Wandi; Hung, Fan; Scalzo, Fabien

doi:10.1109/wacv51458.2022.00328

Cited by 12 publications

(13 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Afterwards, the pseudo label quality is apparently improved to a level close to the fully-supervised method after 3 selftraining rounds. Despite that pseudo label has been previously investigated (Sivanesan et al 2021, Lyu et al 2022, Vepa et al 2022, our method is superior to these works mainly for the following reasons. Firstly, the style of all samples in the dataset used in Sivanesan et al (2021), Lyu et al (2022), Vepa et al (2022) studies is uniform.…”

Section: Conclusion and Discussionmentioning

confidence: 97%

“…Despite that pseudo label has been previously investigated (Sivanesan et al 2021, Lyu et al 2022, Vepa et al 2022, our method is superior to these works mainly for the following reasons. Firstly, the style of all samples in the dataset used in Sivanesan et al (2021), Lyu et al (2022), Vepa et al (2022) studies is uniform. However, the dataset has two types of samples in our study, the first type is normal-vessel LSCI images with clear vascular structure and low annotation difficulty, the other type is abnormal-vessel LSCI images with the fuzzy blood vessel structure and difficult to obtain accurate labels through manual annotate.…”

Section: Conclusion and Discussionmentioning

confidence: 97%

“…The second one is pseudo label generation strategy, which introduces pseudo labels for unannotated images to optimize the model parameters. For example, Vepa et al (2022) utilized an active contour model for weak supervision and a low-cost human-in-the-loop strategy to improve weak label quality. Lyu et al (2022) proposed an approach to train convolutional networks for liver tumor segmentation using Couinaud segment annotations which are image-level labels with values ranging from 1 to 8, indicating a specific region of the liver.…”

Section: Conclusion and Discussionmentioning

confidence: 99%

See 2 more Smart Citations

Robust vessel segmentation in laser speckle contrast images based on semi-weakly supervised learning

Yang,

Chang,

Yuan

et al. 2023

Phys. Med. Biol.

View full text Add to dashboard Cite

Objective. The goal of this study is to develop a robust semi-weakly supervised learning strategy for vessel segmentation in laser speckle contrast imaging (LSCI), addressing the challenges associated with the low signal-to-noise ratio, small vessel size, and irregular vascular aberration in diseased regions, while improving the performance and robustness of the segmentation method. Approach. For the training dataset, the healthy vascular images denoted as normal-vessel samples were manually labeled, while the diseased LSCI images involving tumor or embolism were denoted as abnormal-vessel samples and annotated as pseudo labels by the traditional semantic segmentation methods. In the training phase, the pseudo labels were constantly updated to improve the segmentation accuracy based on DeepLabv3+. Objective evaluation was conducted on the normal-vessel test set, while subjective evaluation was performed on the abnormal-vessel test set. Main results. The proposed method achieved an IOU of 0.8671, a Dice of 0.9288, and a mean relative percentage difference (mRPD) with supervised learning of 0.5% in the objective evaluation. In the subjective evaluation, our method significantly outperformed other methods in main vessel segmentation, tiny vessel segmentation, and blood vessel connection. Additionally, our method exhibited robustness when abnormal-vessel style noise was added to normal-vessel samples using a style translation network. Significance. The proposed semi-weakly supervised learning strategy demonstrates high efficiency and excellent robustness for vascular segmentation in LSCI, providing a potential tool for assessing the morphological and structural features of vessels in clinical applications.

show abstract

Section: Conclusion and Discussionmentioning

confidence: 97%

Section: Conclusion and Discussionmentioning

confidence: 97%

Section: Conclusion and Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Robust vessel segmentation in laser speckle contrast images based on semi-weakly supervised learning

Yang,

Chang,

Yuan

et al. 2023

Phys. Med. Biol.

View full text Add to dashboard Cite

show abstract

“…Vepa et al proposed an automated cerebral vascular segmentation by using active contour as a weak annotation generator. Their results illustrated slight lower scores but significantly reducing annotation time comparing to manual labeling for weak label [29]. Gondal et al achieved high accuracy, low false positives with high sensitivity in detecting lesion region in retinal images by using lesion-level and image-level annotation for weakly-supervised boundary localization, achieving commensurate or even better performance than fully-supervise method [30].…”

Section: B Application Of Weakly-supervise Learning In Medical Imagingmentioning

confidence: 99%

OCTAve: 2D en face Optical Coherence Tomography Angiography Vessel Segmentation in Weakly-Supervised Learning with Locality Augmentation

Chinkamol¹,

Kanjaras²,

Sawangjai³

et al. 2022

Preprint

View full text Add to dashboard Cite

While there have been increased researches using deep learning techniques for the extraction of vascular structure from the 2D en face OCTA, for such approach, it is known that the data annotation process on the curvilinear structure like the retinal vasculature is very costly and time consuming, albeit few tried to address the annotation problem. In this work, we propose the application of the scribble-base weakly-supervised learning method to automate the pixel-level annotation. The proposed method, called OCTAve, combines the weakly-supervised learning using scribble-annotated ground truth augmented with an adversarial and a novel self-supervised deep supervision. Our novel mechanism is designed to utilize the discriminative outputs from the discrimination layer of a UNet-like architecture where the Kullback-Liebler Divergence between the aggregate discriminative outputs and the segmentation map predicate is minimized during the training. This combined method leads to the better localization of the vascular structure as shown in our experiments. We validate our proposed method on the large public datasets i.e., ROSE, OCTA-500. The segmentation performance is compared against both state-of-the-art fully-supervised and scribblebased weakly-supervised approaches. The implementation of our work used in the experiments is located at [LINK].

show abstract

“…They proved that deep learning methods are capable for the task when data is sufficient. To address the problem of lack of data and the very expensive manual labeling cost, weakly supervised methods [11,43] has been adopted to mitigate the need for dense and accurate annotation. However, we believe the generation of weak labels still calls for much domain knowledge and labor.…”

Section: Introductionmentioning

confidence: 99%

Weakly-supervised Cerebrovascular Segmentation Network with Shape Prior and Model Indicator

Chen

Huang

et al. 2022

Proceedings of the 2022 International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

Labeling cerebral vessels requires domain knowledge in neurology and could be extremely laborious, and there is a scarcity of public annotated cerebrovascular datasets. Traditional machine learning or statistical models could yield decent results on thick vessels with high contrast while having poor performance on those regions of low contrast. In our work, we employ a statistic model as noisy labels and propose a Transformer-based architecture which utilizes Hessian shape prior as soft supervision. It enhances the learning ability of the network to tubular structures, so that the model can make more accurate predictions on refined cerebrovascular segmentation. Furthermore, to combat the overfitting towards noisy labels as model training, we introduce an effective label extension strategy that only calls for a few manual strokes on one sample. These supplementary labels are not used for supervision but only as an indicator to tell where the model keeps the most generalization capability, so as to further guide the model selection in validation. Our experiments are carried out on a public TOF-MRA dataset from MIDAS data platform, and the results demonstrate that our method shows superior performance on cerebrovascular segmentation which achieves Dice of 0.831±0.040 in the dataset.

show abstract

Weakly-Supervised Convolutional Neural Networks for Vessel Segmentation in Cerebral Angiography

Cited by 12 publications

References 29 publications

Robust vessel segmentation in laser speckle contrast images based on semi-weakly supervised learning

Robust vessel segmentation in laser speckle contrast images based on semi-weakly supervised learning

OCTAve: 2D en face Optical Coherence Tomography Angiography Vessel Segmentation in Weakly-Supervised Learning with Locality Augmentation

Weakly-supervised Cerebrovascular Segmentation Network with Shape Prior and Model Indicator

Contact Info

Product

Resources

About