Final Report Summary - IMOVE (Extraction of information on muscle control during movements)
Electrical properties of muscles have been under intense scientific and clinical investigation for decades. From the neurophysiologic point of view, it is essential to achieve a deeper understanding of neuromuscular alterations and of their relation to work condition, immobilisation, overtraining and microgravity. Measurable indicators of incipient degeneration, effectiveness of treatment, and preventive actions are required to practice evidence based medicine, rehabilitation and training of athletes.
Technical difficulties associated with recording and analysis of electromyograms have limited the accuracy with which the characteristics of individual motor units can be established during movements. The existing information extraction techniques have mainly been applied to the isometric muscle contractions, with the muscle geometry kept constant during the measurement session. On the other hand, the contractions of human muscles are almost always dynamic, with the muscle moving with respect to the skin.
The main objectives of the IMOVE project (see http://storm.uni-mb.si/iMOVE(opens in new window) for details) were to design and implement automated signal processing techniques capable of extracting the information about the individual motor units from high-density surface electromyograms (hd-sEMG) recorded during controlled dynamic contractions of skeletal muscles, to study the feasibility, efficiency and repeatability of information extraction during movements, and to define the recommendations for sensors, sensor placement and signal processing in dynamic conditions.
In the framework of the project, two different decomposition techniques have been developed and thoroughly tested. The first method operates offline, requires recordings over several repetitions of dynamic muscle contractions and enables identification of individual motor unit discharge patterns during dynamic contractions at relatively high contraction velocities (e.g. elbow radial velocities of up to 90 degrees per second in the case of biceps brachii muscle). The second method identifies individual motor units in real time, does not rely on cyclostationarity of hd-sEMG measurements, but enables motor unit identifications in slow muscle contractions only (e.g. elbow radial velocities of up to 20 degrees per second in the case of biceps brachii muscle). Both methods have been tested on experimental hd-sEMG signals, recorded during different dynamic contractions of biceps brachii and tibialis anterior muscles. In addition, tests on synthetic dynamic hd-sEMG signals have been performed in order to systematically assess their performance on signal epochs of different lengths, different velocities of muscle contractions and with different configurations of surface electrode arrays. In these preliminary tests, up to 15 motor units has been reliably identified per contraction.
The developed decomposition techniques have been utilised in physiological studies of muscle cramps (in collaboration with Laboratory for Engineering of the Neuromuscular System from Politecnico di Torino, Italy and Georg-August University in Göttingen, Germany), exercise induced changes in muscle fibres (in collaboration with the Department of Biology of Physical Activity, University of Jyväskylä, Finland) and pathological tremor (in collaboration with the Department of Neurorehabilitation Engineering at Georg-August University in Göttingen, Germany). Other possible collateral applications of the developed decomposition approaches include objective assessment of effectiveness of rehabilitation and training of athletes, prevention and quantification of work-related neuromuscular disorders and diseases, advanced control of neuroprosthetic devices and monitoring of the musculoskeletal deterioration in microgravity environment.
Technical difficulties associated with recording and analysis of electromyograms have limited the accuracy with which the characteristics of individual motor units can be established during movements. The existing information extraction techniques have mainly been applied to the isometric muscle contractions, with the muscle geometry kept constant during the measurement session. On the other hand, the contractions of human muscles are almost always dynamic, with the muscle moving with respect to the skin.
The main objectives of the IMOVE project (see http://storm.uni-mb.si/iMOVE(opens in new window) for details) were to design and implement automated signal processing techniques capable of extracting the information about the individual motor units from high-density surface electromyograms (hd-sEMG) recorded during controlled dynamic contractions of skeletal muscles, to study the feasibility, efficiency and repeatability of information extraction during movements, and to define the recommendations for sensors, sensor placement and signal processing in dynamic conditions.
In the framework of the project, two different decomposition techniques have been developed and thoroughly tested. The first method operates offline, requires recordings over several repetitions of dynamic muscle contractions and enables identification of individual motor unit discharge patterns during dynamic contractions at relatively high contraction velocities (e.g. elbow radial velocities of up to 90 degrees per second in the case of biceps brachii muscle). The second method identifies individual motor units in real time, does not rely on cyclostationarity of hd-sEMG measurements, but enables motor unit identifications in slow muscle contractions only (e.g. elbow radial velocities of up to 20 degrees per second in the case of biceps brachii muscle). Both methods have been tested on experimental hd-sEMG signals, recorded during different dynamic contractions of biceps brachii and tibialis anterior muscles. In addition, tests on synthetic dynamic hd-sEMG signals have been performed in order to systematically assess their performance on signal epochs of different lengths, different velocities of muscle contractions and with different configurations of surface electrode arrays. In these preliminary tests, up to 15 motor units has been reliably identified per contraction.
The developed decomposition techniques have been utilised in physiological studies of muscle cramps (in collaboration with Laboratory for Engineering of the Neuromuscular System from Politecnico di Torino, Italy and Georg-August University in Göttingen, Germany), exercise induced changes in muscle fibres (in collaboration with the Department of Biology of Physical Activity, University of Jyväskylä, Finland) and pathological tremor (in collaboration with the Department of Neurorehabilitation Engineering at Georg-August University in Göttingen, Germany). Other possible collateral applications of the developed decomposition approaches include objective assessment of effectiveness of rehabilitation and training of athletes, prevention and quantification of work-related neuromuscular disorders and diseases, advanced control of neuroprosthetic devices and monitoring of the musculoskeletal deterioration in microgravity environment.