1. IMUs

 Sensors that use a combination of MEMS technoi, to determine the change in relative orientation ot body or body segment over time are called inertial measurment units (IMUs). IMUs usually include gyroscope and accelerometer but may also include magnetometers or a global positioning GPS) to provide attitude, heading amd postioning Triaxial sensors are used in IMUs, so that three-dimensional motion of segments relative to other segments can be calculated.

 MEMS gyroscopes usually take the form of a tuning-fork, an oscillating wheel, a Foucault pendu-lum or a wine glass resonator, and take advantage of what is known as the Coriolis effect. In principle, each gyroscope has a vibrating mass which is suspended on a support and, as the vibrating mass rotates, a force is produced (Coriolis force), which is proportional to the rate of rotation of the mass. So, the amount of rotation is calculated by inte-grating the angular velocity of the plane of vibration. MEMS accelerometers  are positioned in an orthogonal configuration to the gyroscope on a printed circuit board that makes up the IMU, together with any other sensors that are used by that system. Accelerometers are used in IMUs to calculate the change in position of segments relative to other segments, which is determined by double integration of the acceleration information.

The gyroscopes provide direct measures of the angular velocity in the local coordinate system of the sensor in the three planes. The accelerometers measure the acceleration, which can also be used to determine the inclination of the sensor with respect to gravity. The magnetometers measure the magnetic field, which can be used to determine the sensor incli-nation with respect to the earth's magnetic field.


Although IMUs are increasingly miniaturized and often combined with wireless data acquisition capa-bilities, which makes them extremely attractive for biomechanical analysis, they do suffer from inherent measurement issues. To measure the position and ori-entation of a segment, the IMU must integrate the raw data outputs from the accelerometer and the gyro-scope. The position of the segment in space is calcu-lated from the accelerometer, and then the orientation is calculated once the position is known, so the accu-racy of the gyroscope depends on the accuracy of the accelerometer to determine the correct position. Accelerometers tend to be sensitive to pressure, temperature and height changes and if these are not corrected, this can introduce a bias in the data which can cause drift.

Finally, the capture frequency of the IMU is important because the sensor can only estimate what happens in relation to position and orientation between each data point, which may mean that some I of the real movement is lost if the sensor only i captures at low frequency. Inertial sensors can be used individually, but for 1 biomechanical applications these have been incorpo-I rated into Mocap suits (e.g. the XSENS MVM suit) and systems that can use two or more sensors to measure specific joints (e.g. Delsys). These systems use an inertial sensor to represent each body segment. They are an exciting development for clinicians, as they are able to measure three-dimensional movement outside of a laboratory setting. Their accuracy is, however, dependent upon the quality of the combined  sensor system and the algorithms used to calculate the three-dimensional position and orientation of each segment. Inertial sensors have been used in sports  such as downhill skiing, speed skating, horse riding and baseball pitching, and in studies ivolving clinical gait assesment and clinical task, such as sit-to-stand.