CR-Unet: A Composite Network for Ovary and Follicle Segmentation in Ultrasound Images

Li, Haoming; Fang, Jinghui; Liu, Shengfeng; Liang, Xiaowen; Yang, Xin; Mai, Zixin; Van, Manh The; Wang, Tianfu; Chen, Zhiyi; Ni, Dong

doi:10.1109/jbhi.2019.2946092

Cited by 84 publications

(37 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a novel biomarker, we validated the clinical value of follicular area. Two specific comparisons would be made: the correlation between the mean diameter of follicles (including the mean diameter measured by the senior sonographer and CR-Unet), the follicular area (including the area of manual tacking by the senior sonographer and automated measurement by CR-Unet) and 3-D volumes (Li et al 2020). In this section, 3-D volumes calculated by experienced senior sonographers, using VOCAL software, were regarded as the gold standard, and we compared the clinical application of diameter and area biomarkers.…”

Section: Methodsmentioning

confidence: 99%

“…1). With the input of semi-structured and unstructured data, several indicators (e.g., diameter of each follicle, DSC of the segmentation of different sonographers) of ovary and follicles would be given as output rapidly (Li et al 2020).…”

Section: Modeling Proceduresmentioning

confidence: 99%

“…In our previous study (Li et al 2020), we developed a deep learning network, namely CR-Unet, to simultaneously segment the ovary and follicles based on 2-D ultrasound images in transvaginal ultrasound (TVUS). The results indicated that CR-Unet achieved the best segmentation performance compared with other state-ofthe-art methods for both the ovary and follicles, with Dice similarity coefficients (DSCs) of 0.912 and 0.858, respectively.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

CR-Unet-Based Ultrasonic Follicle Monitoring to Reduce Diameter Variability and Generate Area Automatically as a Novel Biomarker for Follicular Maturity

Liang

Fang

et al. 2020

Ultrasound in Medicine & Biology

Self Cite

View full text Add to dashboard Cite

Follicle size is closely related to ovarian function and is an important biomarker in transvaginal ultrasound examinations for assessing follicular maturity during an assisted reproduction cycle. However, manual measurement is time consuming and subject to high inter-and intra-observer variability. Based on the deep learning model CR-Unet described in our previous study, the aim of our present study was to investigate further the feasibility of using this model in clinical practice by validating its performance in reducing the inter-and intra-observer variability of follicle diameter measurement. This study also investigated whether follicular area is a better biomarker than diameter in assessing follicular maturity. Data on 106 ovaries and 230 follicles collected from 80 cases of single follicular cycles and 26 cases of multiple follicular cycles constituted the validation set. Intra-observer variability was 0.973 and 0.982 for the senior sonographer and junior sonographer in single follicular cycles and 0.979 (0.971, 0.985) and 0.920 (0.892, 0.943) in multiple follicular cycles, respectively, while CR-Unet had no intra-group variation. BlandÀAltman plot analysis indicated that the 95% limits of agreement between senior sonographer and CR-Unet (À2.1 to 1.1 mm, À2.02 to 0.75 mm) were smaller than those between senior sonographer and junior sonographer (À1.51 to 1.15 mm, À2.1 to 1.56 mm) in single and multiple follicular cycles. The average operating times of diameter measurement taken by the junior sonographer, senior sonographer and CR-Unet were 7.54 § 1.8, 4.87 § 0.84 and 1.66 § 0.76 s, respectively (p < 0.001). Correlation analysis indicated that both manual and automated follicular area correlated better with follicular volume than diameter. The deep learning algorithm and the new biomarker of follicular area hold potential for clinical application of ultrasonic follicular monitoring.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Modeling Proceduresmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CR-Unet-Based Ultrasonic Follicle Monitoring to Reduce Diameter Variability and Generate Area Automatically as a Novel Biomarker for Follicular Maturity

Liang

Fang

et al. 2020

Ultrasound in Medicine & Biology

Self Cite

View full text Add to dashboard Cite

show abstract

“…The remaining layers of the new network are initialized randomly and trained according to the new task (Yosinski et al 2014). In ultrasound, transfer learning is used in fetal ultrasound , pelvic ultrasound (Li et al 2020) including prostate cancer detection (Azizi et al 2017), breast cancer classification (Yap et al 2018), thyroid nodules classification (Liu et al 2017b;Chi et al 2017), liver fibrosis classification (Meng et al 2017), cardiac ultrasound ), bone detection (Baka et al 2017) and abdominal classification (Cheng and Malhi 2017).…”

Section: Deep Learning Approachesmentioning

confidence: 99%

Ultrasound tissue classification: a review

et al. 2020

View full text Add to dashboard Cite

Ultrasound imaging is the most widespread medical imaging modality for creating images of the human body in clinical practice. Tissue classification in ultrasound has been established as one of the most active research areas, driven by many important clinical applications. In this paper, we present a survey on ultrasound tissue classification, focusing on recent advances in this area. We start with a brief review on the main clinical applications. We then introduce the traditional approaches, where the existing research on feature extraction and classifier design are reviewed. As deep learning approaches becoming popular for medical image analysis, the recent deep learning methods for tissue classification are also introduced. We briefly discuss the FDA-cleared techniques being used clinically. We conclude with the discussion on the challenges and research focus in future.

show abstract

“…US image segmentation witnessed its significant improvement in this deep learning era [6]. However, due to the low contrast, low resolution and speckle noise in US images, the accuracy of these methods is hampered in ambiguous boundary regions, such as the blurred boundary or shadow-occluded parts [5]. Taking the 3D ovarian US volume as an example, as shown in Fig.…”

Section: Introductionmentioning

confidence: 99%

Contrastive Rendering for Ultrasound Image Segmentation

Liang

Shi

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Ultrasound (US) image segmentation embraced its significant improvement in deep learning era. However, the lack of sharp boundaries in US images still remains an inherent challenge for segmentation. Previous methods often resort to global context, multi-scale cues or auxiliary guidance to estimate the boundaries. It is hard for these methods to approach pixel-level learning for fine-grained boundary generating. In this paper, we propose a novel and effective framework to improve boundary estimation in US images. Our work has three highlights. First, we propose to formulate the boundary estimation as a rendering task, which can recognize ambiguous points (pixels/voxels) and calibrate the boundary prediction via enriched feature representation learning. Second, we introduce point-wise contrastive learning to enhance the similarity of points from the same class and contrastively decrease the similarity of points from different classes. Boundary ambiguities are therefore further addressed. Third, both rendering and contrastive learning tasks contribute to consistent improvement while reducing network parameters. As a proof-of-concept, we performed validation experiments on a challenging dataset of 86 ovarian US volumes. Results show that our proposed method outperforms state-of-the-art methods and has the potential to be used in clinical practice.

show abstract

CR-Unet: A Composite Network for Ovary and Follicle Segmentation in Ultrasound Images

Cited by 84 publications

References 35 publications

CR-Unet-Based Ultrasonic Follicle Monitoring to Reduce Diameter Variability and Generate Area Automatically as a Novel Biomarker for Follicular Maturity

CR-Unet-Based Ultrasonic Follicle Monitoring to Reduce Diameter Variability and Generate Area Automatically as a Novel Biomarker for Follicular Maturity

Ultrasound tissue classification: a review

Contrastive Rendering for Ultrasound Image Segmentation

Contact Info

Product

Resources

About