Abstract
The F-wave is a randomly occurring response to electric stimulation of the motor nerve that normally is only of clinical interest as a diagnostic tool. It’s impact on functional electrical stimulation (FES) has generally not been considered. However it has significant influence on the myoelectrical signal (MES) registered from the stimulated muscle and may subsequently have an effect on the contraction properties of the muscle, that should not be neglected. When MES is used for controlling FES of the same muscle, the F-wave is an unwanted noise component. During FES the stimulation frequency, intensity and location differs significantly from the traditional clinical investigation eliciting the F-wave and lack in literature. This paper reports some informative measurements to validate if the F-wave can be volitionally suppressed, the effect of continuous stimulation and training with visual feedback to the subject. It has been found that the F-wave in the relaxed healthy muscle is about a tenth of the MES at full contraction. Furthermore it has been found that there is no significant change in the level during short term (20min exercise) or medium term (5 sessions) of the F-wave level but that there is an increased level during ½ min of constant stimulation. It is concluded that the F-wave therefore has to be considered as a noise that cannot be volitionally repressed.
1. Introduction
Possibilities to use the residual volitional myoelectric signal (MES) from the paretic muscles to control FES (MeCFES) of the same muscle has a great potential and is still being investigated [1,2,3] despite its difficulties [1,2]. After the M-wave and refractive period there is a window in the MES where the volitional control signal from the subject can be registered. An analog/digital hybrid system [2,4] can reduce the problems regarding stimulation artefacts and M-wave sufficiently to obtain some control and the system has proven the possibility to augment contraction of the paretic muscle in some stroke and spinal cord injured [5] though there is a problem obtaining smooth continuous control with a simple proportional control algorithm. The problem is caused by many factors such as the intrinsically reduced dynamic range of the MES due to the reduced number of volitionally control units, antidromic nerve blocking etc. that decreases the signal to noise ratio (S/N) of the control signal. A significant noise component (as will be shown in this paper) is the F-wave [6]. An induced action potential generated in the motor nerve will travel bi-directionally, the anti-dromic part will travel towards the cell body and where it might be reflected as a secondary action potential, the F-wave in the motor unit.
The F-wave occurs with a latency corresponding to the time it takes the action potential to travel from the site of stimulation, to the spinal cord and back to the muscle. Typically the latency is between 30-50ms for an adult person [7] and is normally only of clinical interest in diagnostics of diseases and is evoked at frequencies of 1Hz or lower[6,8]. Literature treating the properties of the F-wave during elevated frequencies, prolonged stimulation with feedback to the subject, has not been found. For the purpose of further development of the MES controlled FES study aims to test if the F-wave can be volitionally repressed.
2. Subjects and method
Biphasic charge-balanced stimulation with an inter-stimulation interval of 60ms was given to the tibial muscle of 5 healthy male subjects (Age:22-34, mean=26 years height:178-198cm mean 187). Stimulation electrodes were placed just over the head of fibula and 15-20cm below laterally to the crest of tibia modified obtain maximal dorsiflexion torque. Recording electrodes were placed perpendicularly on the midline between the stimulation electrodes to reduce stimulation artefacts. Visual feedback of the level of the root mean square of the MES in a window covering the cronodispersion (start to end) of the F-wave was given to the subject by means of a computer screen. Subjects were instructed to relax the muscle during stimulation and thereby keep the level MES activity as low as possible. Method of establishing this window is described in the accompanying paper [7]. Five sessions consisting of 20 trials has been carried out on the 5 subjects. Each trial consisted of ½min. of stimulation followed by ½min rest. MES is divided into blocks, where each block contains the MES between subsequent stimulation pulses. A third order high pass butterworth filter is applied to the signalblocks (corresponding to applying the transformed combfilter to the original signal) to remove DC-components and periodic components in the signal. The F-wave interval (tfstart –tfend) is determined by a semi-automatic method [7] and the standard deviation (SD) over the signal in this calculated for each stimulation (F-level). Signal blocks containing volitional contraction or noise are rejected if the SD over the interval from tmend-tfstart (tmend =25ms is the end of the M-wave) of a block is exceeding twice the mean value over all blocks.
Trends are calculated using the linear regression coefficients (b) on three different groups of F-levels; Trial, Session and Global. Trial regression coefficient (btrial) is calculated as F-level vs. stimulation pulses over the ½ min trial, session regression (bsession) is calculated as mean trial F-level vs. trial number (1-20) and Global regression (bglobal) is mean session F-level vs session number.
3. Results
Visual feedback of the MES level did not immediately seem to influence the performance of the subjects. In general the level remained constant but often with a slight increase towards the end. An example of the F-waves is shown in Fig1. Analysis of the linear regression (Fig 2) of F-level towards the time of each trial shows that the regression coefficient is positive (**significance level a=5%) in all but subj.D. This corresponds to 28, 45, 42, -6 and 10 % change, for the subjects A-E, over ½ min stimulation. Fig 3 show bSession as the regression of F-level vs time over the entire session as well as globally over all sessions bglobal. This is only significantly** negative for subjects A,C and E.
A full volitional contraction without stimulation give typical values in the order of 50µVRMS when calculated over the F-wave interval. Absolute levels of the F-wave and the noise floor depends on electrode position and skin condition and can therefore not be expected to be directly comparable from session to session. The mean values can be found in Fig 4. where it can be seen that F-level is significantly higher than the noise floor.
|
Fig 1. Traces of the signal blocks following stimulation. First 20ms is dominated by the M-wave. Hereafter is a window where volitional activity may be present (in this case considered as noise). In the middle is the F-wave interval where individual waves can be identified. |
|
Fig 2. Mean and standard deviation for the regression coefficients in all trials grouped by subject. |
|
Fig 3. Regression over sessions and globally over the 5 session ensemble for each subject. For the sessions means with standard deviation is shown. |
|
Fig4. Mean F-level and the corresponding noise level of all trials and their standard deviations. |
4. Discussion
Locally there is a tendency of an increased F-level during the test with ½min of constant stimulation. This increase can thus constitute a significant part of the F-level. Since there is little evidence of a decrease in F-level with the number of trials and sessions this local F-level increase must be due to the prolonged stimulation and not a habituation. The F waves constitutes a level that is about a tenth of the volitional MES and can therefore not be neglected when considering the MES for control purpose. Given the very low values of F-level change resulting from the five trials it seems that it is not possible with few training sessions to be able to volitionally suppress the F-waves significantly. Since the F-wave represents an extra contraction of the motor unit it involves increased force production. Therefore it should also be taken into account when making models for the electrically stimulated muscle since it add a stochastic component to the in-out characteristics.
5. Conclusion
We have investigated the possibility to acquire an ability to suppress the F-wave by training where visual feedback of the F-wave level was given. The findings were not positively confirming that this is possible within five sessions of 20 trials for neurological intact people. It does though not exclude the possibility that this skill can be obtained by other means. F-waves constitutes a significant proportion of the volitional MES in a healthy subject and is therefore a worse problem when the volitional MES is reduced due to a upper motor neuron lesion. Results indicates that increased F-level may be a possible result of prolonged FES with medium frequency.
Acknowledgements
This work has been sponsored by the EU-project NeuralPro with equipment and methods developed in University Twente, The Netherlands and University College London, England with support from the EU-project NEUROS.
Literature
[1] L.
Vodovnik, C. Long, J. Reswick , A. Lippay, and D. Starbuck, “Myo-electric
control of paralyzed muscles.,” IEEE Trans Biomed Eng, vol. 12, pp. 169-72,
1965.
[2] M. Popovic;, T. Keller;, I. P. I. Papas;, V.
Dietz;, and M. Morari;, “Surface-stimulation technology for grasping and
walking neuroprostheses,” IEEE Engineering in Medicine and Biology Magazine ,,
vol. 20, pp. 82 -93, 2001.
[3] R. Thorsen, R. Spadone, and M. Ferrarin, “A
pilot study of myoelectrically controlled FES of upper extremity,” IEEE Trans. Neural Syst. &
Rehab. Eng., vol. 9, pp. 161 -168, 2001.
[4] R. Thorsen, “An artefact suppressing
fast-recovery myoelectric amplifier,” IEEE Trans Biomed Eng, vol. 46, pp. 764-6,
1999.
[5] R. Thorsen, J. Burridge, N. Donaldson, M.
Ferrarin, J. Norton, and P. Veltink, “Auto-Myo-Electric Control of FES on
Tibialis Anterior. A Pilot Study With Stroke Patients.,” Proceedings of the 6
th Annual Conference of the International Functional Electrical Stimulation
Society, pp. 147-149, 2001.
[6] M. A. Fisher, “AAEM Minimonograph #13: H
reflexes and F waves: physiology and clinical indications,” Muscle Nerve, vol.
15, pp. 1223-33, 1992.
[7] M.Freschi, R.Thorsen, et.al. “An Automatic
Procedure For Determination Of F-Wave Latency And End Intervals” Proc. 7 th. Ann. Conf. of IFESS,
submitted2002
[8] E. Stålberg and B. Falck, “Clinical Motor Nerve
Conduction Studies,” Methods in Clinical Neurophysiology, vol. 4, pp. 61-80,
1993.