Abstract:In trauma and critical care settings, rapid and precise intravascular access is key to patients' survival. Our research aims at ensuring this access, even when skilled medical personnel are not readily available. Vessel bifurcations are anatomical landmarks that can guide the safe placement of catheters or needles during medical procedures. Although ultrasound is advantageous in navigating anatomical landmarks in emergency scenarios due to its portability and safety, to our knowledge no existing algorithm can autonomously extract vessel bifurcations using ultrasound images. This is primarily due to the limited availability of ground truth data, in particular, data from live subjects, needed for training and validating reliable models. Researchers often resort to using data from anatomical phantoms or simulations. We introduce BIFURC, Bifurcation Identification for Ultrasound-driven Robot Cannulation, a novel algorithm that identifies vessel bifurcations and provides optimal needle insertion sites for an autonomous robotic cannulation system. BIFURC integrates expert knowledge with deep learning techniques to efficiently detect vessel bifurcations within the femoral region and can be trained on a limited amount of in-vivo data. We evaluated our algorithm using a medical phantom as well as real-world experiments involving live pigs. In all cases, BIFURC consistently identified bifurcation points and needle insertion locations in alignment with those identified by expert clinicians.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to quickly and accurately achieve vascular puncture through an autonomous robotic ultrasound system in trauma and intensive care settings. Specifically, the research aims to develop an algorithm that can automatically identify vascular bifurcation points and provide the optimal needle insertion position in the absence of skilled medical personnel, ensuring that patients can receive necessary treatment in a timely manner. This challenge is mainly due to the inability of existing technologies to autonomously extract vascular bifurcation points in ultrasound images, especially the limited availability of in - vivo data, which restricts the ability to train and validate reliable models. ### Main Problems 1. **Rapid and Accurate Vascular Puncture**: In trauma and intensive care settings, rapid and accurate vascular puncture is crucial for patient survival. 2. **Lack of Skilled Medical Personnel**: In many cases, especially in mass - casualty incidents or areas with insufficient resources, skilled medical personnel may not be available. 3. **Identification of Vascular Bifurcation Points in Ultrasound Images**: Existing algorithms are unable to autonomously extract vascular bifurcation points from ultrasound images, especially when in - vivo data is limited. ### Solutions The paper proposes a new algorithm named BIFURC (Bifurcation Identification For Ultrasound - driven Robotic Cannulation), which combines expert knowledge and deep - learning techniques, can efficiently detect vascular bifurcation points on limited in - vivo data, and provide the optimal needle insertion position for the autonomous robotic puncture system. ### Key Contributions 1. **3D Vascular Skeletonization**: For the first time, a method of creating 3D vascular skeletonization using a linear ultrasound probe is proposed to accurately and autonomously identify vascular bifurcation points. 2. **Optimal Needle Insertion Point Detection**: An algorithm for automatically detecting the optimal needle insertion point in the femoral region is proposed for the first time. 3. **Practical Application Evaluation**: Evaluation is carried out in more real - life examples than other machine - learning studies on ultrasound - guided needle insertion in the femoral region. 4. **Expert - level Performance**: The algorithm can achieve expert - level performance in both simulated data and live - pig experiments, autonomously identifying bifurcation points and accurately determining the needle insertion position. ### Method Overview 1. **Vascular Segmentation**: The RESUS algorithm is used to segment blood vessels, although there are some limitations, such as adjacent blood vessels being merged into a single object. 2. **Centerline Prediction**: - **Erosion Algorithm**: Remove noise and small blood vessels in the output of the segmentation model and separate adjacent blood vessels. - **Blood Vessel Detection**: Use the center of the minimum enclosing circle as the position of each segment to filter out noise. - **Blood Vessel Tracking**: Group the center points into different trajectories to represent their spatial configurations. - **Trajectory Merging**: Merge trajectories based on medical heuristic rules. 3. **Bifurcation Point Identification**: Iteratively search for pairs of points in each trajectory and check whether the conditions of time, distance, and origin trajectory are met. 4. **Needle Insertion Point Identification**: Select a point 2 centimeters away from the bifurcation point as the needle insertion position. ### Experimental Results - **Segmentation Accuracy**: The average IoU score is 0.659. - **Bifurcation Point Error**: The average Euclidean distance is 7.66 millimeters. - **Processing Time**: The average recognition time is 2.61 seconds. - **The success rate is 85.7%**: In tests on six pigs and one phantom model, BIFURC performs excellently in identifying and locating bifurcation points. ### Conclusion The BIFURC algorithm shows high efficiency and accuracy in identifying vascular bifurcation points and the optimal needle insertion position, significantly improves time efficiency, and is expected to assist practitioners of different experience levels to achieve near - expert performance.

Bifurcation Identification for Ultrasound-driven Robotic Cannulation

Vessel Bifurcation Localization Based on Intraoperative Three-Dimensional Ultrasound and Catheter Path for Image-Guided Catheter Intervention of Oral Cancers

Towards Autonomous Robotic Minimally Invasive Ultrasound Scanning and Vessel Reconstruction on Non-Planar Surfaces

3D reconstruction of coronary artery bifurcations from intravascular ultrasound and angiography

Vessel Pose Estimation for Obstacle Avoidance in Needle Steering Surgery Using Multiple Forward Looking Sensors

Bifurcation detection in intravascular optical coherence tomography using vision transformer based deep learning

3D reconstruction of coronary artery bifurcations from coronary angiography and optical coherence tomography: feasibility, validation, and reproducibility

AI-Enabled, Ultrasound-Guided Handheld Robotic Device for Femoral Vascular Access

Semi-supervised Vein Segmentation of Ultrasound Images for Autonomous Venipuncture.

Reconstruction and Quantification of the Carotid Artery Bifurcation From 3-D Ultrasound Images

Experimental validation and clinical feasibility of 3D reconstruction of coronary artery bifurcation stents using intravascular ultrasound

Toward Smart, Automated Junctional Tourniquets-AI Models to Interpret Vessel Occlusion at Physiological Pressure Points

Image-based Intraluminal Contact Force Monitoring in Robotic Vascular Navigation

Real-time Visualized Vessel and Catheter Path Reconstruction for Intravascular Intervention

Autonomous Robotic Screening of Tubular Structures based only on Real-Time Ultrasound Imaging Feedback

Toward Robotically Automated Femoral Vascular Access

DopUS-Net: Quality-Aware Robotic Ultrasound Imaging based on Doppler Signal

Adaptive Kinematic Control of a Robotic Venipuncture Device Based on Stereo Vision, Ultrasound, and Force Guidance

Preliminary Evaluation of an Ultrasound-Guided Robotic System for Autonomous Percutaneous Intervention

Computer Aided Intracranial Aneurysm Treatment Based on 2D/3D Mapping, Virtual Deployment and Online Distal Marker Detection

In-Vivo Validation of a Novel Robotic Platform for Endovascular Intervention