IoT-based Remote Control Study of a Robotic Trans-esophageal Ultrasound Probe via LAN and 5G

TL;DR

This study demonstrates the feasibility of remotely controlling a robotic trans-esophageal ultrasound probe via IoT over LAN and 5G, analyzing hysteresis effects and comparing control devices for accuracy and performance.

Contribution

It introduces an IoT-based remote control system for a TEE robot using LAN and 5G, and evaluates control device performance and hysteresis effects.

Findings

Hysteresis effects differ between axes with deadbands of 15 degrees.

Positioning errors are approximately 0.5 degrees for manual and robotic control.

Button-based gamepad performs better than joystick in task completion time.

Abstract

A robotic trans-esophageal echocardiography (TEE) probe has been recently developed to address the problems with manual control in the X-ray envi-ronment when a conventional probe is used for interventional procedure guidance. However, the robot was exclusively to be used in local areas and the effectiveness of remote control has not been scientifically tested. In this study, we implemented an Internet-of-things (IoT)-based configuration to the TEE robot so the system can set up a local area network (LAN) or be configured to connect to an internet cloud over 5G. To investigate the re-mote control, backlash hysteresis effects were measured and analysed. A joy-stick-based device and a button-based gamepad were then employed and compared with the manual control in a target reaching experiment for the two steering axes. The results indicated different hysteresis curves for the left-right and up-down steering axes with the input wheel's deadbands found to be 15 deg and deg, respectively. Similar magnitudes of positioning errors at approximately 0.5 deg and maximum overshoots at around 2.5 deg were found when manually and robotically controlling the TEE probe. The amount of time to finish the task indicated a better performance using the button-based gamepad over joystick-based device, although both were worse than the manual control. It is concluded that the IoT-based remote control of the TEE probe is feasible and a trained user can accurately manipulate the probe. The main identified problem was the backlash hysteresis in the steering axes, which can result in continuous oscillations and overshoots.