If more than one demultiplexer is connected to the
stimulator, the DMR signal will be set high by the
slowest demultiplexer indicating that all demultiplex-
ers are ready. The STIM signal is shared by all de-
multiplexer devices.
6 EMG MEASUREMENT
The measurement of EMG from a stimulated muscle
is of interest in order to detect residual volitional mus-
cle activity. The latter might be used to control the
stimulation. The use of array electrodes for stimula-
tion does not allow the placement of additional EMG
electrodes close to the virtual active electrode due to
the larger size of the array electrode. Therefore, EMG
measurement must be performed from virtual EMG
electrodes formed by elements of the stimulation ar-
ray. As shown in (Shalaby, 2011), EMG can be di-
rectly measured from the stimulation electrodes dur-
ing stimulation if the EMG amplifier is protected and
the electrodes are periodically discharged.
The EMG protection circuit consists of multiple
PhotoMOS relays and a high pass filter. The required
timing diagram for the PhotoMOS switches is shown
in Figure 4. During delivery of stimulation pulses two
X
last switch config. for stimulation
t
1
t
EMG switch configuration
next sequence
t
3
t
sc
t
2
t
f ilter_connected
t
4
t
f s
t
EM G_valid
PhotoMOS 1 on
t
5
PhotoMOS 2 on
1 21 2
Figure 4: EMG protection and discharging timing, t
1
. . . t
4
depend on sequence length and array setting.
PhotoMOS switches are used to mute the EMG mea-
surement (PhotoMOS 2 off). After the stimulation is
completed, another PhotoMOS relay short circuits the
electrodes over which the EMG signal is measured for
an user defined time interval to eliminate any resid-
ual charge on the electrodes (PhotoMOS 1 on). The
electrodes then are connected to the high-pass filter
(PhotoMOS 2 on) which reduces low frequency dis-
turbances. The actual EMG measurement however is
performed by an external standard EMG amplifier.
The duration of the short circuit t
sc
is between 7
and 8ms. The high-pass filter needs the time t
fs
to set-
tle. Thus it is best to activate PhotoMOS 2 as early as
possible to maximize the valid EMG recording time
t
EMG
valid
. The stimulation-induced EMG response
(M-wave) falls together with the filter transients and
is excluded from the EMG measurement. Basically,
only volitional muscle activity is captured in the time
interval t
EMG
valid
. An EMG recording of the wrist
extensor under stimulation is shown in figure 5. Parts
of the EMG recording which are disturbed by filter
transients are already blanked in the shown EMG sig-
nal.
41.0 41.5 42.0 42.5 43.0 43.5
−1000
−500
0
500
1000
1500
time [s]
EMG [uV]
EMG under stimulation
Figure 5: EMG recording of volitional muscle activity
from the stimulated wrist extensor (pulse width 100µs,
I=15mA).
7 CONCLUSIONS
The system enables researchers and health profes-
sionals to use array electrodes without large effort.
The developed software simplifies the setup of com-
plex stimulation patterns and permits a straightfor-
ward integration of array electrodes into existing
stimulation setups. Parameters like pulse widths,
stimulation currents or demultiplexer configurations
can be adjusted in real-time. The demultiplexer sup-
ports array sizes up to 60 elements for the active elec-
trode and up to 4 elements for the indifferent elec-
trode. The small switch module can be placed near
to the array electrode, avoiding extensive wiring. The
ability to measure volitional EMG signals from a pair
of virtual electrodes makes array electrodes usable for
diagnostics or control applications. In future work,
the remaining problem of filter transients must be
solved by introducing digital filters and including the
EMG amplifier into the switch module.
REFERENCES
Azevedo-Coste, C., Bijelic, G., Schwirtlich, L., and
Popovic, D. (2007). Treating drop-foot in hemi-
plegics: the role of matrix electrode. In 11th Mediter-
ranean Conf. on Medical and Biomedical Eng. and
Computing 2007, pages 654–657.
Keller, T., Lawrence, M., Kuhn, A., and Morari, M. (2006).
New multi-channel transcutaneous electrical stimula-
tion technology for rehabilitation. In Conf Proc IEEE
Eng Med Biol Soc., volume 1, pages 194–197.
BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices
348