Ultrasound-Guided Assistive Robots for Scoliosis Assessment With Optimization-Based Control and Variable Impedance

• Published in: IEEE robotics and automation letters Vol. 7; no. 3; pp. 1 - 8
• Main Authors: Duan, Anqing, Victorova, Maria, Zhao, Jingyuan, Sun, Yuxiang, Zheng, Yongping, Navarro-Alarcon, David
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.07.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Contact force; End effectors; Human engineering; Impedance; Learning from Demonstration; Medical Robots and Systems; Optimization; Optimization and Optimal Control; Physical Human-Robot Interaction; Probes; Quadratic programming; Robot kinematics; Robots; Scoliosis; Service robots; Task analysis; Task and Motion Planning; Ultrasonic imaging
• ISSN: 2377-3766; 2377-3766
• DOI: 10.1109/LRA.2022.3186504

Assistive robots for healthcare have witnessed a growing demand over the past decades. In this paper, we investigate the development of an optimization-based control framework with variable impedance for an assistive robot to perform ultrasound-guided scoliosis assessment. The conventional procedure for scoliosis assessment using ultrasound imaging typically requires a medical practitioner to slide an ultrasound probe along a patient's back while maintaining a certain magnitude of the contact force. To automate such a procedure, we need to consider multiple objectives, such as contact force, position, orientation, energy, posture, etc. To coordinate different objectives, we propose to formulate the control framework as a quadratic programming problem with each objective weighted by a tunable task priority, subject to a set of equality and inequality constraints. As the procedure requires the robot to establish a constant contact force with the patient during scanning, we incorporate variable impedance regulation of the end-effector to enhance safety and stability during the physical human-robot interaction; The variable impedance gains are then retrieved by learning from medical expert's demonstrations. The proposed methodology is evaluated with a robotic system performing autonomous scoliosis assessment with multiple human subjects involved. The effectiveness of our approach is verified by the coronal spinal images obtained with the robot.