Design and Evaluation of a Handheld Robotic Device for Peripheral Catheterization

Area of Science:

• Medical robotics
• Medical imaging
• Machine learning

Background:

• Medical robots enhance surgical procedures with improved dexterity, vision, and safety.
• Peripheral catheterization is crucial for various medical interventions but can be challenging.

Purpose of the Study:

• To present the mechanical design and experimental validation of a handheld robotic device for peripheral catheter insertions.
• To evaluate the performance of an ultrasound deep learning algorithm for vessel segmentation.
• To assess the device's accuracy and repeatability in simulated challenging conditions.

Main Methods:

• Development of a handheld robotic device (VeniBot) integrating ultrasound imaging, miniaturized robotics, and machine learning.
• Implementation of a deep learning algorithm for ultrasound-based vessel segmentation.
• Experimental validation using tissue-mimicking phantom models simulating difficult peripheral catheter placement.

Main Results:

• VeniBot achieved high success rates: 97±4% for vessel punctures and 89±7% for sheath cannulations (n=240).
• The ultrasound deep learning algorithm demonstrated effective vessel segmentation capabilities.
• The device showed high accuracy and repeatability in simulated challenging peripheral catheterization scenarios.

Conclusions:

• The handheld robotic device (VeniBot) shows viability for semi-automated peripheral catheterization.
• This technology has the potential to enhance clinical workflow and improve patient comfort.
• Further development could lead to safer and more efficient catheterization procedures.