A Kinematic Information Acquisition Model That Uses Digital Signals from an Inertial and Magnetic Motion Capture System

Abstract

This paper presents a model that enables the transformation of digital signals generated by an inertial and magnetic motion capture system into kinematic information. First, the operation and data generated by the used inertial and magnetic system are described. Subsequently, the five stages of the proposed model are described, concluding with its implementation in a virtual environment to display the kinematic information. Finally, the applied tests are presented to evaluate the performance of the model through the execution of four exercises on the upper limb: flexion and extension of the elbow, and pronation and supination of the forearm. The results show a mean squared error of 3.82° in elbow flexion-extension movements and 3.46° in forearm pronation-supination movements. The results were obtained by comparing the inertial and magnetic system versus an optical motion capture system, allowing for the identification of the usability and functionality of the proposed model.

Conversion of Upper-Limb Inertial Measurement Unit Data to Joint Angles: A Systematic Review

“…Sensor placement on the body is often chosen to minimize soft tissue or skin motion artifacts and includes the flat portion of the sternum, the lateral distal aspect of the upper arm, and the dorsal distal aspect of the forearm. For the simultaneous collection of IMU and optoelectronic data, retro-reflective markers and IMUs have been placed independently on anatomical landmarks [ 36 , 52 , 60 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 , 81 , 82 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 ], retro-reflective markers placed directly on IMUs [ 30 , 41 , 91 , 92 , 93 , 94 , 95 , 96 , 97 , 98 , 99 , 100 , 101 , 102 , 103 , 104 , 105 , 106 ], or a combination of both approaches [ 61 , 107 , 108 ,…”

“…To establish a relationship between IMU orientation and that of an underlying anatomical coordinate system, a sensor-to-segment calibration process is typically employed. Common calibration methods include predefined sensor alignment with a segment [ 66 , 94 , 103 , 104 ], static pose alignment [ 67 , 69 , 70 , 71 , 72 , 73 , 74 , 83 , 84 , 88 , 98 , 99 , 100 , 102 , 105 , 109 ], functional joint movements [ 92 , 107 ], or their combinations [ 30 , 36 , 41 , 61 , 68 , 79 , 81 , 91 , 96 , 97 , 106 , 108 , 110 , 112 , 113 , 114 ], as well as use of IMU palpation calipers [ 90 , 93 ] ( Figure 4 ). Predefined sensor alignment involves the placement of an IMU on the body to align with a bone-fixed reference frame.…”

“…Predefined sensor alignment involves the placement of an IMU on the body to align with a bone-fixed reference frame. For static pose calibration, a subject may perform one or more pre-determined static postures, for example, standing with the palms facing to the front [ 107 , 112 ], an N-pose with arms neutrally placed alongside the body [ 41 , 83 , 100 , 106 , 111 ], or a T-pose with arms abducted to [ 69 , 86 , 97 ]. Static calibration then aims to define the sensor coordinate system using gravity as a reference by averaging resultant accelerometer data for a given pose while also establishing a neutral or reference alignment of a joint.…”

“…To convert IMU orientation data to joint angles, definitions of anatomical joint coordinate systems have been established using the Denavit–Hartenberg representation [ 61 , 67 , 89 ], orthonormal [ 30 , 41 , 66 , 68 , 70 , 71 , 73 , 79 , 83 , 86 , 91 , 92 , 93 , 94 , 95 , 96 , 97 , 98 , 101 , 106 , 107 , 108 , 109 , 110 ], and non-orthonormal segment coordinate systems [ 36 , 52 , 75 , 113 ] ( Figure 5 ). The orientation of one segment relative to another is computed using 3D Euler angle decomposition [ 30 , 36 , 41 , 52 , 66 , 68 , 77 , 78 , 79 , 80 , 84 , 85 , 86 , 88 , 90 , 91 , 92 , 93 , 95 ,…”

“…A total of 39 studies of high ( n = 9) [ 36 , 41 , 52 , 84 , 88 , 92 , 96 , 97 , 113 ], moderate ( n = 25) [ 30 , 61 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 74 , 80 , 81 , 83 , 93 , 94 , 98 , 99 , 101 , 106 , 108 , 110 , 111 , 112 , 115 ] and low quality ( n = 5) [ 73 , 77 , 78 , 82 , 89 ] compared elbow joint angles derived from IMUs with those from optoelectronic systems. Twenty-six of these studies measured motion about the two primary degrees of freedom of the elbow joint, flexion–extension and pronation–supination.…”