EMG Analysis Software
The EMG Analysis software is a research quality analysis program that implements a wide range of powerful analysis methods using Fast Fourier Transform (FFT) techniques as well as many traditional EMG analysis methods making it especially suitable for educational as well as research uses. This application, like the basic EMG Graphing program, fully supports all C3D formats as well as several older file formats and raw data from Dataq data acquisition systems for stand-alone functionality. Data can be processed and exported to third-party applications for additional analysis via C3D files or standard ASCII formatted files compatible with Excel, SAS and MATLAB etc.
This is the one of the most powerful, yet easy to use, software packages using FFT analysis that is available to both the clinician and researcher. The program reads EMG data directly from C3D files as well as the native file formats from B|T|S, Motion Analysis Corporation, Vicon Motion Systems. Written for the Kinesiologist driven environment, directly to clinical specifications, this software effortlessly delivers instant viewing and full color reports using sophisticated Frequency Spectrum, Power Spectrum and Muscle Correlation techniques. The EMG Analysis program includes all of the basic features of the EMG Graphing version of this program and added many analysis and data export options:
EMG Graphing Features included
- Calibrate your EMG data using gain switches or calibration signals to display EMG amplitude in uV at skin surface.
- Includes adult and age-matched child normal activity databases for EMG activity comparisons.
- Display EMG normalization by Gait Cycle, signal amplitude and Manual Muscle Test (MMT) signal amplitude.
- Data processing includes High Pass and Low Pass filters, DC offset removal, noise reduction, and automatic event determination.
- Reads raw EMG data from Dataq, Motion Analysis, Vicon and C3D files
- Saves all processed EMG data in C3D files for easy retrieval and exchange.
- Templates automatically identify muscle and calibration information when EMG files are opened.
- Print multiple and individual EMG Gait Cycle reports displaying the signal as raw or rectified EMG data.
- Automatic detection of gait cycles from event switch data, force plate and marker data, or C3D event data.
- Export processed EMG gait cycles to ASCII files to create Raw EMG and Rectified EMG files (Excel compatible).
- Direct interface to the MLS Report Generator for Presentation Standard Graphics suitable for posters and slides.
EMG Analysis Features
- Moving Average Envelope EMG reports with adjustable window size.
- Linear Envelope EMG reports with adjustable filter frequencies via dual pass digital FIR filters for zero delay processing.
- RMS data Analysis EMG reports with adjustable window size, preprocessing high pass filter
- Intensity Filtered Averaging EMG reports with adjustable window size and signal suppression levels.
- EMG threshold detection EMG reports with 10 adjustable thresholds using moving average and Linear envelopes.
- Zero Crossing detection EMG reports with adjustable hysteresis levels.
- Integrate over Time EMG reports with adjustable reset time interval.
- Integrate and Reset EMG reports with adjustable reset levels.
- EMG Power Spectrum reports with with FFT windows determined by the EMG data.
- EMG Amplitude Distribution reports.
- Cocontraction correlation EMG reports with user determined low pass filter and optional high pass artifact filters.
- Signal aliasing reports for data collection diagnosis and Quality Assurance.
- Process and export rectified, window and linear envelope data to C3D files for third-party analysis.
- Export processed data as ASCII files using user specified number of data points, optionally normalized or scaled.
- Multimedia support - listen to any EMG channel and export EMG data as audio files.
Buy it once, use it everywhere - the EMG Analysis application, like all Motion Lab Systems applications, is a site licensed application. The purchase of an EMG Analysis license allows multiple copies of the software to be used within the licensed environment, permitting its use on multiple computers, laptops etc., making it very easy to use in academic and research environments without any hardware access keys or restrictive licensing requirements.
EMG Trial Reports
Raw EMG Data - opening an EMG data file with EMG Analysis will display the raw, unprocessed data for each recorded EMG channel. The channels can be labeled with the individual muscle names, side and scaled at surface potential (microvolts). Heel contact and Toe-off is optionally indicated by vertical lines on each plot with user selected color coded of the EMG data according to limb side. Once this information is entered, the EMG Analysis will store it in the C3D file as part of the data record as that it will always be available whenever anyone opens the file.
Rectified EMG Data - the EMG data can also be plotted and printed as rectified EMG. The EMG Analysis program offers the user complete control over the size of both the displayed image at all times. Data displayed on the screen can be previewed before sending it to the printer to allow the user to optimize the printed output - EMG trial reports can be compressed onto a single page or printed out in full over several pages.
Envelope EMG Data - the EMG Analysis program supports the display of enveloped EMG data for trial reports, displaying the complete enveloped gait cycle as enveloped data across all EMG channels. The individual gait cycle envelopes can be averaged within the trial and displayed in the EMG Analysis reports and of course, the entire trial can be printed as a single display showing the relationships between the individual muscle contractions.
All three trial display options offer the optional display of normal EMG activity bars underneath the subjects EMG records together with gait cycle events shown as heel strike (solid lines) and toe-off (dotted lines), indicating the stance and swing phases of gait. Each EMG channel can be labeled with side, a label and muscle name together with either the calibrated EMG levels (skin surface or intramusclular), or the actual recorded signal levels in volts.
EMG Analysis Reports and Methods
The EMG Analysis program can generate a gait cycle graph of raw EMG for each recorded EMG channel. Graphs can be scaled by %, or surface potential (microvolts) and display full heel contract to hell contact data with swing and stance phases of gait indicated. "Normal" EMG activity bars can be displayed and printed and each EMG graph labeled with a the muscle name. The muscle names and activity bars are fully configurable by the user and can be edited or translated into any language. The EMG Analysis program allows the user to define individual normal activity and create a number of individually named activity profiles for different age groups and EMG protocols. EMG Analysis is supplied with sample normal activity profiles and can be downloaded and run as an evaluation version on any Windows PC.
In addition to the raw gait cycle EMG Graphs, EMG Analysis can also display the EMG data as a rectified signal. The graph below illustrates this and, in addition, is scaled in microvolts at skin surface as a result of a calibration operation. EMG Analysis supports EMG level calibration via built-in sources (available on most Motion Lab Systems EMG systems), or via an external source if your EMG system does not include calibration facilities.
"Normal" EMG activity bars may be displayed and printed and each EMG graph assigned a muscle name. The muscle names and activity bars are fully configurable by the user and can be edited or translated into any language. The EMG Analysis program allows the user to define individual normal activity and create a number of individually named activity profiles for different age groups and EMG protocols. EMG Analysis is supplied with normal adult and age matched children's activity profiles from documented sources - the supplied normal database is fully editable.
Each EMG graph can be displayed and printed in a user-selected color to indicate the limb (Left/Right) that is assigned to a specific EMG channel. These colors are used in both the screen displays and the printed output. The results of the analysis (both raw gait cycle and rectified gait cycle data) can be exported to Excel compatible ASCII text files, scaled by maximal effort, manual muscle test or skin surface levels.
EMG Analysis can generate gait cycle reports using two envelope methods. Using the first method the user can create reports using a standard user defined window filter with an adjustable window period specified in milliseconds- the graph illustrated to the right was generated using a moving window of 150ms.
Trials that contain multiple cycles of EMG data can average the defined cycles to generate standard deviation displays. Each individual EMG graph is labeled with the muscle name, scaling method, and displays the gait cycle phasic activity together with normal activity bars for adult or age matched normal children. Each printed reports will display the window method, together with the selected gait cycle period and stand/swing percentage. This analysis option is not available in the basic EMG Graphing version.
The second envelope analysis method provided by EMG Analysis applies a user-controlled FIR (Finite Impulse Response) filter to the gait cycle. The cutoff frequency is user-selected - a single or dual pass filter is applied to the envelope to generate the individual plots allowing the graphs to be generated with typical data processing delays for comparison with older data, or with delay free results for current research.
The Averaged data can be plotted as a scaled value or as a percentage of maximum. The graph on the left uses a 6Hz, dual pass filter to smooth the EMG activity. The vertical dotted lines on the graph indicate the transition from stance to swing while the dual pass method eliminates the delay that is inherent in standard windowing methods.
Each of the enveloping and averaging methods has its own set of user controlled preset options that control the windowing period (specified in milliseconds) and the linear envelope filter frequency (specified in Hertz). The EMG analysis application allows the results of the analyses to be exported as ASCII files for further processing by third-party applications including Excel, SAS, MATLAB, Statistica etc - options are provided to normalize the exported data to a set numbers of points so that the exported data can be averaged without further processing. In addition, options are provided that allow the amplitude exported data can be scaled or calibrated. This analysis method is only available in the EMG Analysis application.
The RMS (Root Mean Square) Analysis differs from the Moving Average and Linear Envelope analysis methods in that it starts with raw EMG data instead of rectified EMG data. The analysis option was added to the EMG Analysis software at the request of a user who was able to specify the series of mathematical operations required, and the order in which they were to be performed. Motion Lab Systems is happy to add additional EMG Analysis functionality to the EMG Analysis program when users can provide precise descriptions of the required function.
The RMS analysis performs the following operations on the raw EMG data. A high pass filter with a user selectable cutoff frequency (usually in the range of 1 to 20Hz) is applied to the raw EMG data to remove any DC offset in the data stream. A root mean square value for the EMG data stream is then processed (the raw EMG data values are squared and then the square root calculated) and the resulting output is then averaged over a selectable period, usually in the range of 20 to 499 milliseconds. This analysis method is only available in the EMG Analysis application.
The Intensity Filtered Average (IFA) analysis option implements the algorithm described in the paper Computer Algorithms to Characterize Individual Subject EMG Profiles During Gait published in 1992 by Ross Bogey, Lee Barnes, and Jacquelin Perry, MD. This paper describes a processing method based on both the timing and the relative amplitude of the EMG signal that generates better results than the standard Moving Average or Linear Envelope methods. It is proposed that the IFA method generates an envelope output that more closely represents the onset, duration, and cessation of the subjects muscle activity.
The Intensity Filtered Average method differs from the various enveloping analysis methods because it excludes brief EMG activity bursts (less than 5% of the gait cycle or EMG period), and activity below a preset threshold, from the averaging process. Averaging methods used by the Window, Linear Envelope and RMS analysis methods all generate activity envelopes that include all of the available EMG data. The rational for the IFA method is that by excluding relatively low level EMG activity from the envelope and averaging process, the results better reflect the essential muscle activity during gait as the low intensity and short duration bursts have very little effect on the joint motion.
We developed an intuitive user interface for the Motion Lab Systems EMG Analysis program that places all of the commonly used options and commands on the mouse menu. Extensive field testing allows the program to make optimal use of customized menus that automatically display the correct defaults for the user. EMG Analysis is an exceptionally easy program to use.
The Threshold analysis option in EMG Analysis can produce muscle activity reports of EMG activity within any gait cycle. The activity can be quantisized over a number of individually adjustable activity levels using EMG envelope data derived from either FIR filtered data or a moving window. Although most reports will divide the EMG activity into two or three levels, thye Threshold analysis method supports up to ten individual activity levels. Unlike many programs that simply report the EMG activity as simply on or off, EMG Analysis allows the user to display and print reports that overlay the raw EMG data on the EMG activity display to confirm that the selected activity thresholds are correct. As with other EMG Analysis reports, normal EMG activity bars can be displayed from the included age-matched child and adult normal database. This analysis method is only available in the EMG Analysis application.
The Zero crossing analysis reports the number of times that the raw EMG signal crosses zero. This is reported both as a numeric value (shown in brackets by the muscle name and title, above the graph), and graphically.
This EMG Analysis function implements a function that controls the amount of hysteresis in the comparator function used by the zero crossing analysis. Hysteresis is the difference between the EMG signal level input and the DC zero level and setting this parameter controls the point at which the comparator turns off and turns on. The total number of Zero Crossings for the analyzed period is displayed after the muscle name. High values for the comparator hysteresis value reduce the number of zero crossings that are detected, thus reducing the functions sensitivity to baseline noise in the EMG data. This analysis method is only available in the EMG Analysis application.
The Integrate over Time analysis is performed on the rectified raw EMG signal. The process of integration over time produces an output that is proportional to the level of the EMG signal over a given period that the user can select. The integration process sums the rectified EMG values for the selected time period – at the end of the time period the signal output is reset to zero and the integration process restarted.
The EMG Analysis program implements this function with a single option that defines the integration period in milliseconds. This can bebe set to a value between 1ms and 2000ms (2 seconds). Typical values are between 10 and 200ms depending on the activity period under investigation. This analysis method is only available in the EMG Analysis application.
Like its companion function, the Integrate and Reset Analysis is a mathematical function that is performed on the rectified raw EMG signal by the EMG Analysis program. This process of integration and reset produces an output that is proportional to the level of the EMG signal.
The integration process sums the rectified EMG values until the EMG reaches a set percentage of the average level – once this level is reached the signal output is reset to zero and the integration process restarted. This results in a series of saw-tooth waveforms whos repetion rate is directly proportional to the amplitute of the original EMG signal.
Both the Integrate over Time and the Integrate and Reset analysis functions were popular when only limited computation power was avaialbe to the researcher as both of these functions can be easily implemented in hardware. They are included in the EMG Analysis program as reference methods for the researcher and classroom demonstration.
In addition to the FIR filtering, included as part of the EMG data pre-processing features, EMG Analysis includes a number of advanced analysis features that use fast Fourier Transform (FFT) techniques. The FFT is an efficient algorithm, well suited to computation, that computes the discrete Fourier transform (DFT) and its inverse in a rapid and efficient manner. FFTs are used in a wide variety of applications and are particularly useful in performing frequency analysis functions within this application. The EMG Analysis program uses FFTs to perform EMG Power Spectrum Analysis, EMG Amplitude distribution Analysis, Correlation Analysis and Signal Aliasing analysis in addition to EMG signal filtering.
The EMG Power Spectrum analysis displays the distribution of the various frequency components in the EMG signal. The user has complete control of all the Power Spectrum variables and can select the start frequency and end frequency points for the spectra. Power Spectrum analysis can be performed over the entire EMG period or any EMG burst can be manually selected.
The default method uses the processed EMG envelope to automatically determine and analyze the longest period of activity for each individual muscle thus restricting the Power Spectrum analysis to the major burst of activity and eliminating the baseline from the FFT calculations. This method produces a spectral report that displays the EMG activity without biasing the resuts with baseline noise. In addition, a further option allows the user to restrict the Power Spectrum analysis to ignore EMG signals below a fixed percentage of the EMG signal.
The Amplitude Distribution Analysis measures the amplitude density function and displays the amplitude distribution of the EMG signal representing the relative percentage of time that the EMG signal is at a given amplitude. In this analysis the range of EMG amplitudes is plotted across the horizontal axis while the total time at each level is plotted along the side.
This analysis is useful in looking at activity over longer periods then individual gait cycles. High levels at the left side of the graph (lower EMG levels) indicate that EMG activity is relatively light whereas the further the activity extends to the right side of the graph suggests a greater degree of muscle activity.
Most amplitude distribution graphs will display either resting activity (a single peak to the left) or a double peak indicating both resting and work periods within the analyzed EMG recording. Amplitude Distribution analysis is a useful tool in biofeedback and relaxation analysis as well as an investigation into muscle activity and ergonomics. Some studies have indicated that workers whose amplitude distributions are strongly skewed to the right tend to report more muscle pain than others.
The Cocontraction Analysis allows you to view the degree of correlation between the simultaneous activation of antagonistic muscles. The EMG Analysis program includes a muscle cocontraction analysis function that allows the user to chose pairs of antagonistic muscle to compare. This function processes the rectified and smoothed EMG data using high pass of low pass filters that can be set by the user to provide a consistent cocontraction measurement environment.
The concontraction results can be plotted as processed envelope graphs with each muscle identified by color and also exported as an ASCII text file for additional processing. A concontraction index is calculated and displayed for each muscle pair.
If an analog data signal is not sampled at a high enough rate then signal aliasing can corrupt the sampled data by folding back all analog signals that have a frequency greater than half the sampling frequency (also known as the Nyquist point). This is the phenomenon that is known as aliasing and can cause problems when recording EMG signals. It is very important to avoid aliasing artifact in an EMG signal as is can not be removed from the EMG data. The EMG Analysis aliasing function can be used on any recorded EMG data to help select a suitable EMG sampling rate. In this example the selected EMG channel shows that aliasing signals are present in the data indicating that there are problems with the recording sampling rates which were too low for the frequencies present in the EMG data.
The EMG Analysis application, like all Motion Lab Systems applications, is a site licensed application. The purchase of an EMG Analysis license allows multiple copies of the software to be used within the licensed environment, permitting its use on multiple computers, laptops etc., making it particularly useful in academic and research environments as it does not require hardware access keys or impose restrictive licensing requirements.