As soon as a file is opened it is displayed as a raw EMG file and may then be subject to any or all of the features listed above. The data may be left as a whole or truncated into smaller files of specific data for further analysis. Of course not only EMG data may be displayed at any one time. The Y axis may be labelled with any units governed by what sensors where connected at the time of recording, e.g. Kg or lbs for force or degrees for limb movement. These units are then saved with the file and automatically compute every time the file is reopened.
The digital inputs are displayed on a separate graph to the analog inputs. Using this facility toe contact and heel off may be obtained using the contact switch assembly type no. FS4.
Once an EMG waveform has been obtained, it is often necessary to process the waveform to extract useful information. The type of analysis depends upon the application but the following guidelines may be helpful:
- EMG waveforms are fundamentally bursts of random signals with certain characteristics that indicate timing, force and fatigue.
- After acquiring a waveform, remove unwanted sections of the trace. There is often a period before the activity of interest and a period after - removing these reduces the data and speeds up subsequent processing.
- Display the waveform as an RMS signal to obtain meaningful amplitude values. See RMS Filter for further information.
- As an alternative to an RMS display, rectify and then average the signal to produce a mean absolute waveform.
- Use the median frequency filter to estimate muscle fatigue. See the section Median Frequency Filter for further information.
- Frequency analysis only has a meaning for muscle activity where the length of the contracting muscle remains constant.
- Several filters may be cascaded. For example, the RMS value can be produced followed by a median frequency filter. See the Filters Overview section for further information.
- To measure muscle timing, be careful not to apply too much filtering. Taking an average over many samples will effectively delay the onset of muscle activity as seen on the filtered waveform.