Semi-supervised Vein Segmentation of Ultrasound Images for Autonomous Venipuncture

Chen, Yu; Wang, Yuxuan; Lai, Bolin; Chen, Zijie; Cao, Xu; Ye, Nanyang; Ren, Zhongyuan; Zhao, Junbo; Zhou, Xiao-Yun; Qi, Peng

doi:10.1109/iros51168.2021.9636149

Cited by 7 publications

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several studies have examined the use of machine learning approaches to enhance vein visualization by detecting and highlighting peripheral and deep veins in US images [17], [18]. VeniBot [19]is a portable robotic system for autonomous venipuncture that uses a semi-supervised vein segmentation scheme from ultrasound images for autonomous navigation and guidance. VeeBot [20] is a robotic system based on an EPSON robot, that has been customized for intravenous access and uses NIR imaging as the primary vein-detection method, and ultrasound to confirm the target vein has adequate blood flow.…”

Section: A Perception: Vein Visualization and Localizationmentioning

confidence: 99%

Robotic Devices for Assisted and Autonomous Intravenous Access

Koskinopoulou

Cheng

Acemoğlu

et al. 2023

IEEE Trans. Med. Robot. Bionics

View full text Add to dashboard Cite

Intravenous Access (IVA) is the most common invasive medical procedure. Globally, it is estimated that over one billion IVA devices (needles/catheters) are used annually. However, the overall failure rate in this procedure is unacceptably high, reaching values between 35 and 50%. This has driven a great deal of research and technological development over recent decades, including the integration of different levels of autonomy in IVA medical devices to greatly improve this process. This paper will review these recent technical developments, including methods and systems for vein imaging and localization, needle insertion, venipuncture detection, catheter placement, and complete robotic IVA platforms. Furthermore, this paper explores emerging technical aspects, current limitations, and new research directions that may enable wider clinical translation and better acceptance of robotic IVA technologies.

show abstract

Section: A Perception: Vein Visualization and Localizationmentioning

confidence: 99%

Robotic Devices for Assisted and Autonomous Intravenous Access

Koskinopoulou

Cheng

Acemoğlu

et al. 2023

IEEE Trans. Med. Robot. Bionics

View full text Add to dashboard Cite

show abstract

“…This approach employs shadow enhancement and shadow removal mechanisms to encourage the segmentation network to extract features from shadow-free regions at both the image and feature levels. Chen et al [54] developed a semi-supervised segmentation network based on the Mean Teacher method to assist robots in vascular puncture. While Mean Teacher has achieved some success in SSL tasks, its consistency loss is more sensitive to noise.…”

Section: Consistency Regularization Suvosmentioning

confidence: 99%

Semi-supervised Learning for Real-time Segmentation of Ultrasound Video Objects: A Review

Jin,

Zhaojun,

Yanping

2023

Advanced Ultrasound in Diagnosis and Therapy

View full text Add to dashboard Cite

Real-time intelligent segmentation of ultrasound video object is a demanding task in the field of medical image processing and serves as an essential and critical step in image-guided clinical procedures. However, obtaining reliable and accurate medical image annotations often necessitates expert guidance, making the acquisition of large-scale annotated datasets challenging and costly. This presents obstacles for traditional supervised learning methods. Consequently, semi-supervised learning (SSL) has emerged as a promising solution, capable of utilizing unlabeled data to enhance model performance and has been widely adopted in medical image segmentation tasks. However, striking a balance between segmentation accuracy and inference speed remains a challenge for real-time segmentation. This paper provides a comprehensive review of research progress in real-time intelligent semi-supervised ultrasound video object segmentation (SUVOS) and offers insights into future developments in this area.

show abstract