Muscle knot state measurement using multi-axis kneading reaction force in physical massage robot system

In this study, we propose a method for the rapid and accurate measurement of muscle knot states. The state of muscle knots is defined by four factors: three-dimensional location within the inspection area, radii, radius ratio, and orientation around the vertical axis. The proposed method utilizes circular motions under a constant applied force to continuously acquire information about the surrounding area, enabling the assessment of muscle stiffness trends. This method achieves a 2.2 mm average location estimation accuracy, outperforming comparative methods by 1.0 mm and reduced measurement time by 79.5%. For ellipsoidal muscle knots, the proposed method achieved orientation estimation with an angular error of 8.4 degrees. In cases where muscle knots were located under a 15mm-thick simulated skin layer and encapsulated by simulated muscle, measurement became more difficult. The comparison method showed an estimated radius that was 47.0 mm increase, while the proposed method showed only a 5 mm increase. Future studies will aim to enhance adaptability to complex anatomical structures, where muscle layers overlap and muscle stiffness varies.