Electromyogram
| Introduction |
| Downloads |
| Logbook |
[06-MAY-26] This page presents electromyogram (EMG) recordings made in mice and rats with our implantable telemetry sensors. Our Subcutaneous Transmitters (SCT) and our Intraperitoneal Transmitters (IPT) provide bipolar inputs sufficiently sensitive to measure biopotentials as small as 5 μV rms with bandwidth anywhere up to 640 Hz and sample rates up to 4096 SPS. These devices are equipped with flexible, silicone-insulated leads to carry biopotentials to their inputs, and we can prepare the ends of the leads in any number of ways so as to obtain a secure connection to our recording target. When recording EMG, our target is always a muscle, so we use lead terminations suitable for fastening directly to soft tissue, in contrast with terminations that we might secure to the skull to record from the brain.
[06-MAY-26] A repository of recordings and characteristics files. We provide example hour-long raw data files in NDF format, and larger sets of processed data in text files.
[07-MAY-26] We have one hundred hours of EMG recorded from four mice with A3048S2-SS SCTs. The recordings were performed at a local contract research organisation (CRO) whose name we are not mentioning at their request. All surgeries and housing performed in compliance with local regulations. All four SCTs are equipped with C-Leads terminated with S-Coils. Devices No117 and No121 were implanted first, with 50-mm leads. Devices No11 and No12 were implanted several days later, observed by us, and have 35-mm leads. Telemetry reception in the CRO laboratory is perfect without a Faraday enclosure. We place the cage with four mice on an A3038C Animal Location Tracker (ALT).
We process the recordings in eight-second intervals, using the ALT position measurements to obtain the average speed of the animals in centimeters per second for each interval, and using the telemetry signals to obtain the root mean square amplitude of each signal in the band 30-120 Hz. We select this band after observing that the waking EMG power starts to rise at about 40 Hz and starts to roll off at 80 Hz. The 80-Hz roll-off is due to the 80-Hz low-pass filter in the A3048S2. The 40-Hz rise is a feature of the EMG in all four animals.
No117 electrodes in left and right trapezius. This signal has amplitude 6-60 μV including intermittent ECG pulses of amplitude 200 μVpp, often far larger than other variations in the signal. Because these pulses are intermittent, we cannot remove them to expose what we hope to be underlying EMG. We note, however, that the intermittent ECG seems often to vanish when the animal is stationary, allowing the signal amplitude to drop as low as 6 μV. No121 electrodes are implanted side by side in the left trapezius. This signal sometimes contains traces of ECG, but the pulses are never more than 20 μVpp and never contribute significantly to the rms amplitude, which varies from 4-40 μV rms. The amplitude is almost always above 10 μV rms when the animal moves more than 10 cm/s. When the animal activity is 0-2 cm/s the signal amplitude is either 12 μV or 5 μV. We believe this is EMG, showing activity, waking, and sleeping.
Of the No11 and No12 transmitters, one has both leads implanted in the left trapezius, the other has one in the left and one in the right, but we don't yet know which. Neither signal contains significant ECG in the beginning, but as the recording progresses, we see more and more large spikes on the No11 recording. Signal amplitude varies 4-40 μV, with No12 showing some peaks of 60 μVpp. As with No121, amplitude is greater than 10 μV when the animal is moving around, and is 4-12 μV when it is stationary. We have strong, intermittent ECG is to our EMG recording from SCT No117. In the figure below we show an example of No117 with ECG compared to No121 without ECG when both animals are active at 4 cm/s in a one-second interval.
Looking at longer periods of ECG and activity measured in eight-second intervals, we see in the recording from No117 spikes during inactive periods that we don't see in the signal from No121. These spikes correlate with periods in which ECG enters the signal and increases its amplitude. They occur more often when the animal is inactive and awake, but sometimes when the animal is inactive and asleep. We conclude that intermittent ECG undermines our sleep scoring, but does not render the scoring completely inaccurate, because the apparent waking of the animal due to the arrival of ECG in otherwise low-amplitude EMG is rare.
[12-MAY-26] One of the four mice died last week, the one with the No11 SCT. The other three appear healthy and active, see video below. We transfer seven days of EMG recordings. Initial examination shows no loss of amplitude in the active EMG, and no invation of ECG.
[13-MAY-26] One of the four mice died in the past week, No11. This signal showed larger and more frequent non-EMG spikes as the days went by, until in hour 122 of our study period, the EMG and activity drop down to below sleep levels. This is the moment the animal died. Thereafter, EMG power is 2-4 μV rms and activity is 1-3 cm/s. These two values obtained from a dead animal give us our noise floor for both EMG and activity recording. Even if the body generates not signal, the transmitter and small local movements combine to generate a singal of up to 4 μV amplitude, and movement of the other animals around and about the dead animal generate apparent activity of up to 3 cm/s.
We have been band-pass filtering our EMG to 30-120 Hz before calculating amplitude. The spectrum of our waking EMG starts to rise at about 50 Hz and then starts to fall at 80 Hz, which is the corner frequency of the A3048S2's low-pass filter. We imaging that if we were to remove this filter, and instead let the amplifier deliver gain up to the limit of its operational amplifiers, which is about 600 Hz, we would see significantly more EMG power, although we are not sure how much more: perhaps double the amplitude. We would not be sampling fast enough to see 300-Hz oscillations, but we would be sampling fast enough to measure the amplitude of 40-600 Hz oscillations. We note that in the sleeping EMG, the amplitude drops so low that we can see, below 30 Hz, the fundamental and harmonic of ECG, as well as the fundamental of respiration.
