Home › Advanced Brain Monitoring, Inc.

Advanced Brain Monitoring, Inc.

File info: application/pdf · 59 pages · 2.68MB

Advanced Brain Monitoring, Inc. - BIOPAC

59 pages · 2 MB — Introduction. Welcome to the B-Alert Software Manual. The purpose of this guide is to familiarize AcqKnowledge users with. Advanced Brain Monitoring Inc.'s ...

https://www.biopac.com/wp-content/uploads/B-Alert-Software-Manual

Introduction. Welcome to the B-Alert Software Manual. The purpose of this guide is to familiarize AcqKnowledge users with. Advanced Brain Monitoring Inc.'s ...

Full PDF Document

If the inline viewer fails, it will open the original document in compatibility mode automatically. You can also open the file directly.

Extracted Text

Rev: 04.02.2021
B-Alert Software Manual
v. 1.4

42 Aero Camino, Goleta, CA 93117 Tel (805) 685-0066 | Fax (805) 685-0067
info@biopac.com | www.biopac.com

Distributed by Biopac Systems, Inc.

Manufactured by Advanced Brain Monitoring, Inc.

B-Alert Software Manual
Table of Contents
Introduction................................................................................................................................................... 4 Minimum System Requirements................................................................................................................... 4
Chapter 1: B-Alert Graphical User Interface (GUI) .......................... 5
1.1 ABM data acquisition devices and peripherals................................................................................. 5 1.2 Overview of Online Data Acquisition Actions ................................................................................. 5 1.3 Acquisition ........................................................................................................................................ 6 1.4 Acquire & Retransmit ....................................................................................................................... 7 1.5 View Retransmitted .......................................................................................................................... 7 1.6 Test Impedance ................................................................................................................................. 8 1.7 Acquire Metric Benchmarks ............................................................................................................. 9
Chapter 2: B-Alert Gauges GUI.......................................................... 10
2.1 B-Alert GUI Overview ................................................................................................................... 10 2.2 Gauges Window.............................................................................................................................. 10 2.3 Raw Signals Window...................................................................................................................... 12 2.4 Decon. Signals Window (Decontaminated Signals) ....................................................................... 12 2.5 Artifacts Window............................................................................................................................ 13 2.6 Impedances Window....................................................................................................................... 14 2.7 Markers ........................................................................................................................................... 14
Chapter 3: Acquiring Metric Benchmarks ........................................ 15
3.1 Metric Benchmark Tasks ................................................................................................................ 15 3.2 Steps for Acquiring Metric Benchmarks ........................................................................................ 15 3.4 Exiting Benchmarking Session ....................................................................................................... 20
Chapter 4: Advanced Settings and Operations ................................. 23
4.1 Taskbar............................................................................................................................................ 23 4.2 Settings............................................................................................................................................ 24 4.3 Operations ....................................................................................................................................... 26
Chapter 5: Data Analysis and Review ................................................ 31
5.1 Inspect EEG Record........................................................................................................................ 31 5.2 Offline playback using EVA........................................................................................................... 32 5.3 Create Definition File ..................................................................................................................... 35 5.4 Generate Summary File .................................................................................................................. 36 5.5 Offline Analysis .............................................................................................................................. 37
Chapter 6: External Sync Unit (ESU)................................................. 38

www.biopac.com

B-Alert Software Manual 6.1 General ESU Information ............................................................................................................... 38 6.2 ESU Block Diagram ....................................................................................................................... 41 6.3 Configuring the ESU Settings......................................................................................................... 42 6.4 Third Party Protocols and Packet Structure .................................................................................... 43 6.5 Visualizing Third Party data during Data acquisition..................................................................... 45
Appendix A: Data Outputs Guide....................................................... 46
A.1 Overview......................................................................................................................................... 46 A.1 EEG-based Metrics including Cognitive State & Workload Outputs ............................................ 51 A.3 Power Spectral Densities Outputs................................................................................................... 51 A.4 Heart Rate Outputs.......................................................................................................................... 56 A.5 Z-Score Outputs .............................................................................................................................. 57 A.6 Actigraphy Outputs ......................................................................................................................... 58 A.7 Signal Quality Outputs.................................................................................................................... 58

www.biopac.com

B-Alert Software Manual
Introduction
Welcome to the B-Alert� Software Manual. The purpose of this guide is to familiarize AcqKnowledge users with Advanced Brain Monitoring Inc.'s optionally licensed B-Alert� Cognitive States Analysis software. In addition to the AcqKnowledge with B-Alert installation, this licensing level includes stand-alone B-Alert "Live" software, accessible via a separate B-Alert Desktop shortcut.
Access to B-Alert features integrated with AcqKnowledge software (as well as basic B-Alert wireless hardware setup) is explained in the Biopac AcqKnowledge Software Guide and the B-Alert with AcqKnowledge Quick Guide. Please note that in order to use B-Alert in conjunction with AcqKnowledge, the integrated B-Alert software controls must be accessed within the AcqKnowledge application.
The explanations found in this manual demonstrate Advanced Brain Monitoring Inc.'s proprietary B-Alert software as functioning independently from the AcqKnowledge application. Therefore some of the dialogs and functionalities will vary from those generated by using B-Alert within the AcqKnowledge user interface.
For a full explanation of the B-Alert wireless EEG hardware and hardware setup, please refer to the B-Alert� X24/X10 User Manual.
� For step-by-step direction, request B-Alert User Training Videos from support@biopac.com.
Minimum System Requirements
� Personal computer (PC) with minimum PentiumTM 2.4 GHz processor � Minimum of 2 GB of installed RAM memory and 4 MB virtual memory � Windows 7 or higher operating system � .NET framework version 3.5 installed � Minimum of 50 MB hard disk space per 5-hour session � VGA or higher resolution video adapter � One available USB port (two for validation) � Monitor size between 15" and 21" required for Metric Benchmarking Sessions

www.biopac.com

B-Alert Software Manual
Chapter 1: B-Alert Graphical User Interface (GUI)
1.1 ABM data acquisition devices and peripherals

B-Alert Headset x24

X-Series USB Receiver B-Alert Dongle or ESU

1.2 Overview of Online Data Acquisition Actions
The B-Alert Live acquisition Software has five Online Action icons located in the upper left corner of the home screen. The Online Action icons are functions related to live physiological recording: signal quality checks, data acquisition, and monitoring of data acquisition. The Offline Action icons are grouped to the right and enable functions related to data handling for previously recorded signals. See Chapter 5: Data Analysis and Review for detail on Offline Actions.

Online Actions

Offline Actions

The Acquisition function allows users to start/stop data acquisition with their selected mobile EEG device. The communication port is automatically detected, and the Headset Configuration is automatically informed.
The Acquire & Retransmit function allows users to begin data acquisition with their selected mobile EEG device on a primary computer and automatically stream live data across a TCP/IP connection. The transmitted signals can then be viewed on a second computer using the View Retransmitted function. Both computers must be connected on a local area network. Note: when using this function, data is only stored on the primary computer (not on the secondary machine).
The View Retransmitted function allows users to view live signals (raw EEG signals and real time metrics). This function is intended for use on a monitoring unit (secondary computer) when the primary acquisition computer is set to retransmit using the Acquire & Retransmit function. This function does NOT store any data on the secondary computer viewing data. Data will only be stored on the computer running the 'Acquire & Retransmit' function.

www.biopac.com

B-Alert Software Manual
The Acquire Metric Benchmarks function allows the experimenter to run the three tasks required for creating an individual subject's unique definition file (.def file). In order to ensure the validity and accuracy of the Cognitive State and Workload Metrics on a subject, that individual must run the Benchmarks one time when fully rested. Completing them will automatically generate their specific definition file. The three tasks include:
� 3-Choice Vigilance Task for High Engagement and Workload
� Auditory Psychomotor Vigilance Task for Low Engagement � Visual Psychomotor Vigilance Task for Distraction
The Test Impedance function allows the user to test the impedance levels at each electrode site. Low impedance (<40 k) is recommend to ensure good quality of data; the lower the impedance values, the better the conductivity between the scalp and electrodes and thus the better the quality of the signal.
1.3 Acquisition
Before starting an Acquisition, plug the B-Alert USB Receiver into the PC and switch ON the headset. Wait for the solid green LED pattern on the B-Alert headset (and, if you are using an ESU, the solid light on the ESU) to confirm that the B-Alert USB Receiver and Headset have established a connection. Click on the "Acquistion" button to automatically find the communication port and open the Acquisition Settings dialog box. A 9-digit number will be assigned to each session. As shown in the Acquisition Settings dialog box below, the first four digits are required for the subject number, two digits are provided for the task type, and the other three categories only allow for a single digit to be entered. If any fields or digits are left blank, zeros will be automatically inserted in front of the entered values. The categories are designed to provide flexibility when used in any protocol, especially those with cross over or repeated measure designs. This dialog box will open each time with the previously entered session number. If you wish to compute the Cognitive State and Workload Metrics in real-time, select the appropriate check boxes. Artifact Decontamination is essential for Cognitive State and Workload metric classifications and will be automatically selected. Select the definition file (.def) that corresponds to the individual being recorded. The subject number of the definition file must match the subject number entered for the current acquisition session. If your computer appears to be computationally over stressed, you may wish to deselect real-time Artifact Decontamination, Cognitive State Metric, and Workload Metric processing (but please note that only raw data will be displayed and stored).

To obtain Cognitive State Metrics and Workload Metric, the Benchmarks create a unique definition file for
each subject previously created.

The four digit subject numbers must match for the
session and definition file.

After the session number is entered, click OK. The Gauges GUI will open with the Gauges tab active by default. Click the blue play button to begin acquisition. Refer to Chapter 6: B-Alert Gauges GUI for additional detail during recordings.
Warning: If the combination of Subject/Session/Group/Task numbers is identical to a previous acquisition, all files will be overwritten. A warning window will appear to prevent accidental overwrite.

www.biopac.com

B-Alert Software Manual Note: Some device configurations may not support all processed parameters.
1.4 Acquire & Retransmit
The Acquire & Retransmit function allows technicians to monitor real time data acquisition from a secondary computer, as long as primary and secondary computers are connected via TCP/IP.

Primary computer: (Server)
(Using Acquire & Retransmit function)
-Data Storage
Wired/wireless transmission

TCP/IP

Secondary computer: (Passive Client)
(Using View Retransmitted function)
-Signal Monitoring -NO data Storage

Subject
Before starting Acquire & Retransmit, plug the B-Alert USB Receiver into the primary computer and switch ON the headset. Wait for the solid green LED pattern on the headset and solid LED pattern on B-Alert USB Receiver to confirm the Receiver and Headset have connected. On the primary computer, click the "Acquire & Retransmit" icon. The Acquire and Retransmit Settings dialog will appear. Enter appropriate session information (i.e., subject number, group, etc.) and click OK to begin acquiring and retransmitting data. Select the Brain State options and corresponding definition file to generate EEG-based Metrics in real time during data acquisition.

1.5 View Retransmitted
After starting Acquire & Retransmit (on the primary computer), click the "View Retransmitted" icon on the secondary computer (connected to primary computer via TCP/IP). The View Retransmitted Settings dialog will be presented. Enter the IP address of the primary computer into the IP Address field within this dialog (see the steps below to determine the IP Address for the primary computer). Once OK is clicked, the Gauges GUI program will open and the signals being acquired will be presented. Note: acquisition via the Acquire & Retransmit function must be started on the primary computer before the View Retransmitted function is started on secondary computer. The View Retransmitted function will not work on the same computer running the 'Acquisition' function.

www.biopac.com

B-Alert Software Manual
IP address of Primary Computer

IP Port (Default to 4505)

Tips for finding the IP address for the primary computer:
o Open the Start Menu and select the Run option. o In the Run window, type "cmd". o When the command prompt window opens, type "ipconfig" and click enter. o The IP Address will be presented. o Enter this IP Address into the IP Address field on the secondary computer.
If the connection is unsuccessful:
1) Check firewall settings of both primary and secondary computer. Turn the firewall off (not recommended), or provide access to B-Alert Live software
2) Change the IP Port Number on the primary computer, and update the number accordingly on the secondary computer (using Settings/Configure Data Streaming menu).
1.6 Test Impedance
The Test Impedance function allows the user to test the impedance levels at each electrode site. Impedances measure resistance between the scalp and electrode in k, where lower values reflect lower resistance and thus better conductivity between scalp and electrodes. ABM strongly recommends using an impedance check before starting any data collection to optimize data quality. Before starting Test Impedance, plug the B-Alert USB Receiver into the PC and switch ON the headset. Wait for the solid green LED pattern on the headset and the solid LED pattern on the B-Alert USB Receiver to confirm the Receiver and Headset have connected. When the "Test Impedance" button is clicked, the Test Impedance Settings dialog will open. Enter the session number and click OK.

www.biopac.com

B-Alert Software Manual Click the large blue "Start" button to begin the impedance test.

Select All channels for impedance check

De-Select All channels for Impedance check

Impedance values < 40k will be highlighted in green, values 40k - 80k will be yellow, and impedance values > 80k will be red (indicating that the sensor is outside the acceptable range). Impedance values will typically decrease for the first 30-45 min after headset application. For more tips on lowering impedances, refer to the Hardware Manual. ABM recommends sensor impedances below 40k for optimal data quality, but all B-Alert users should determine a standard impedance threshold for data collections to maintain data quality across subjects. Impedances higher than 40k (yellow) can still collect good quality EEG, and thus may not be reason to exclude a participant from continuing to collect. If high impedances are seen across all EEG channels, try replacing the mastoid sensors before troubleshooting each EEG site -- high Reference impedances will impact negatively all EEG impedance measurements. To reduce time during troubleshooting, users can down select and customize which channel(s) will be checked by using the check boxes in the "Get" column, or the "Select All" and "De-Select All" buttons in the ribbon.
The following output files are generated during an impedance test; these files are useful for technical support, but not for data analysis:
� xxxxxxxxx.xxxxxx.xxxxxx.Impedance.Results.csv: This CSV file will contain the impedance values from each impedance test performed. If repeated impedance tests are performed for the subject, a unique CSV file will be saved each time, organized by system time.
� xxxxxxxxx.xxxxxx.xxxxxx.Impedance.ebs: An impedance EBS file is saved in the subject folder for each impedance test performed. The system time is included in the file name to allow simple file tracking.
1.7 Acquire Metric Benchmarks
The Acquire Metric Benchmarks function allows the experimenter to run the three tasks required for definition file (.def file) creation: 3-Choice Vigilance task (3CVT), Auditory Psychomotor Vigilance Task (APVT), and Visual Psychomotor Vigilance Task (VPVT). See

www.biopac.com

B-Alert Software Manual Chapter 3: Acquiring Metric Benchmarks for more details.
Chapter 2: B-Alert Gauges GUI
2.1 B-Alert GUI Overview
After initiating a data acquisition using the "Acquisition" or "Acquire & Retransmit" action button, the B-Alert GUI will be loaded. The B-Alert GUI visualizes raw and processed physiological signals in a range of formats to provide a comprehensive visualization of physiology in real time and offline.

Play Button

Stop Button

Elapsed/ Gauge Time

Display Window Tabs

2.2 Gauges Window
The gauges are customizable so that experimenters can view data and Metrics in a way most meaningful to the nature of their testbed. The default gauges include the High Engagement and Distraction classifications from the Cognitive State Metrics, as well as the Workload Metric, and Heart Rate (when acquired). Heat maps display EEG power spectral densities (PSDs) in both spatial and temporal map4s for the traditional Hz bands (i.e., Beta, Alpha, Theta, and Gamma).

www.biopac.com

B-Alert Software Manual
1. Meter Gauges: These include the High Engagement and Distraction classifications from the Cognitive State 1 Metrics, as well as the Workload Metric, and Heart Rate (when acquired). For the Metrics, the default 1 presentation setting is second-by-second changes averaged on a trailing 3-second window. For users with 2 longer duration testbeds, it may be helpful to reduce the sensitivity of the Metrics with the "Avg" button. 1 Here, users can display more meaningful outputs by varying the sensitivity from two to thirty seconds. 2 2. Time Series: Shows z-scored values (updated every second) in a 15-second window. Three views are 3 selectable: (1) Workload, High Engagement, and Distraction (default); (2) Drowsy, Low Engagement, and
High Engagement; and (3) Overall Heart Rate, Overall Alpha, and Overall Theta.
3. Surface Map: Shows z-scored time history (5-sec) of the PSDs of individual bins calculated across all the channels.
4. Heat Map: Shows 4-second averaged distribution of PSDs for the specified EEG band across the scalp. The PSDs are computed from decontaminated data (if available), smoothed using Kaiser-window, and averaged across 3-epoch overlays. Default bands are: Theta (3-7 Hz), Alpha (8-13 Hz), Beta (13-29 Hz), and Gamma (25-40 Hz).
5. Movement: Shows movement scale computed from the change in the dominant angles of the 3-axis accelerometer.
6. Cog State: Shows the relative percentage of each Cognitive State Metric for the cumulative session. Please note the inverse relationship between each of the classifications.
7. Z-Score: Shows z-scored second-by-second values of EEG Bands for specified electrodes in histogram format. Three default views are available: (1) POz, Fz, Cz for Theta band; (2) POz, Fz, Cz for Alpha band; and (3) Overall Alpha and Theta across all channels and HR.
Note: The availability of the gauges depends on the type of the device and brain state processing selected. For example: 1) Workload Metric classification is not available for the B-Alert X4, as it does not include sufficient electrode site coverage. 2) If no brain state classifications are selected in Acquisition Settings, the corresponding gauges will not be available. 3) If Artifact decontamination is not selected in Acquisition Settings, PSDs will be computed from raw data.

www.biopac.com

B-Alert Software Manual
2.3 Raw Signals Window
The Raw Signals window displays raw data signals, as shown below.

2.4 Decon. Signals Window (Decontaminated Signals)
The Decon Signals window visualizes EEG signals using ABM's validated Artifact identification and decontamination algorithms which identify and remove 5 artifact types: EMG, EOG, excursions, saturations, and spikes. The use of overlaid windowing and data computation results in an approximately 2.5 - 3 second delay in the decontaminated signals. Depending on the amount of noise in the signal and the number of channels in the device, the decontaminated signal could lag up to several additional seconds. For additional details regarding the artifact identification and decontamination procedures, refer to Appendix A: Data Outputs Guide. Use of color coding and displayed running percentages provide users with a real time visualization of data quality. The time scale, monitor gain, and other viewing settings in the Decon Signals window are the same as those on the Raw Signals tab. B-Alert Live does not employ any artifact decontamination procedures for the ECG channel. The data quality indicators (color and percentages) are built for the EEG-based Metrics--- these values may not be appropriate data quality measures for some EEG applications.

www.biopac.com

B-Alert Software Manual

Total % of seconds (epochs) with Good quality
data across all EEG channels (includes all
artifact types) &
Total % of seconds with Good quality data across
Cognitive Metric Sites Green > 95%
95% > Yellow > 85% 85% < Red

% artifact by channel EMG: % of seconds contaminated with EMG
artifact Other: % of other artifacts
(blinks, saturation, excursion, spikes)
Green > 5% EMG/Other 5% >Yellow > 15% 15%> Red

Tip: If EMG > 5% in multiple channels, subject
is likely frowning, clenching teeth, chewing
gum, or moving excessively

Tip: Click on the "Cognitive Metric Sites % Good Data" Flag on the toolbar to change to "Workload Metric Sites % Good Data"
2.5 Artifacts Window
The artifacts window displays the color-coded artifact percentages for each sensor site.

Artifacts Window -- Displays artifact % in color by sensor site
Green > 5% EMG/Other 5% >Yellow > 15% 15%> Red

Note: This artifact window displays sensors for an x10 model.

www.biopac.com

B-Alert Software Manual
2.6 Impedances Window
The Impedances Window allows users to check data quality prior to initiating a data acquisition.
Imp. Check � This feature is not available while
acquisition is running. If acquisition has not been started or is stopped, user can initiate an impedance check by clicking the "Imp. Check" button, then the
"Play" button.
2.7 Markers
The following keyboard shortcuts allow the user to register `Start' and `Stop' markers in the data, indicating regions of interest:
� Shift + S � Start marker (Green marker appears on screen across channels) � Shift + E � Stop marker (Red marker appears on screen across channels) In the Gauges tab, the markers can be used in the Time Series window. In Raw and Decon data tabs, the EEG data at the beginning of the closest epoch will be marked. The markers are saved both in a readable `.mrk' file, as well as in the EDF file. The markers can be visualized in EVA (Refer Section 9) during offline analysis. Note: 1) The markers area available only on the "Raw Signals" tab. 2) The markers are synchronized, i.e. Shift+E will be active only after Shift+S.

www.biopac.com

B-Alert Software Manual
Chapter 3: Acquiring Metric Benchmarks
3.1 Metric Benchmark Tasks
The acquisition of baseline data (Metric Benchmark) is used to create the individualized EEG profiles required for the Cognitive State and Workload Metrics to be valid and accurate across individuals. Each complete Benchmarking Session includes three distinct tasks:
� 3-choice vigilance task (3CVT) � Visual Psychomotor Vigilance Task (VPVT) � Auditory Psychomotor Vigilance Task (APVT) Three durations are supported that vary in increasing order of classification accuracy: � Short_Benchmark: Takes around 9-10 minutes to complete � Medium_Benchmark: Takes around 15 minutes to complete � Long_Benchmark: Takes around 30 minutes to complete Typically, benchmark data only needs to be obtained one time for each individual if performed on a healthy, rested subject. Additional session iterations are recommended, however, when pre- and post-conditional changes are made. Note: Standard lengths of the session may be appended with additional time if the user fails to respond consistently
(suggesting they are fatigued or fell asleep, and thereby distorting the baseline measures).
3.2 Steps for Acquiring Metric Benchmarks
1. Acquire Metric Benchmark
Click the "Acquire Metric Benchmark" icon to select the desired baseline configuration. The Acquire Baseline Data Settings dialog will open. Select the desired baseline configuration from the drop down menu and enter the subject number (XXXX), group number (G), and session iteration number (Y). The session task type (WW) and task iteration (N) values will be generated automatically for all Benchmarking sessions. All EEG data for benchmarking is captured in *Signals.Raw.edf file. All event data is captured individually for each task. Each task will output an *Events.edf Example:
Notice: The first 9 digits are determined by the configuration described above. The next 6 digits are determined by the date of acquisition. The next 6 digits are determined by the time of acquisition. The next descriptor preceding .Events.edf explains the corresponding benchmark task.
Baseline Configuration Selection Menu
Check 'Baseline Technical Monitoring' box to enable monitoring of real time data acquisition from a secondary computer
using 'Remote AMP Monitor' function (see step 2 below). Note of the IP address of this computer.
After entering the session information, BLS will automatically perform a 60-second signal quality/artifact check (press the blue PLAY icon to view data during 60 second check). During this time, the signals presented on the screen should be visually inspected to see if EEG, EMG, and ECG signals are properly presented. If there are signals that appear abnormal, make any necessary adjustments.

www.biopac.com

B-Alert Software Manual
After a quality check is completed, the task will start. The following window will be shown; press the left arrow button on the keyboard to begin the task.
Note: It is highly recommended to perform an impedance check and to make sure that all channels are in the Green before starting the Metric Benchmarks task.

www.biopac.com

B-Alert Software Manual
2. Monitoring Signal During Benchmarking Tasks (Available only for X24 & X10)
When possible, ABM strongly encourages technicians to monitor the data quality of signals throughout the Benchmarking Tasks using the 'Remote AMP Monitor' function. This function acts identical to the 'View Retransmitted' function, where technicians can configure the software to view the live physiological data on the computer running the Benchmarking Tasks over TCP/IP. Note: Data on secondary computer will be only visualized (no data storage).

To launch the Remote AMP Monitor, go to the Operations menu at the top of the main B-Alert Control GUI and select Remote AMP Monitor. Remote AMP monitor function must be launched after the testing has started (during training of 3-CVT), following the 60 second data quality check and impedance check.
A 'Remote AMP Monitor' pop-up window will appear with a text box to allow users to enter in the IP address of the primary computer (i.e., the PC running Acquire Metric Benchmarks). Enter the IP, and then click 'OK' to launch the monitor. The Remote AMP monitor display utilizes functions from an older version of the B-Alert Live software, so the signal presentation will look different than the Gauges/Heat map window, but it displays the same information.

www.biopac.com

B-Alert Software Manual
3. Completing the Benchmarking Tasks The subject will be prompted to complete at least three neurocognitive tests. Each task itself is an abbreviated version of similar tasks from ABM's Alertness & Memory Profiler comprehensive assessment software. The standard session length for the tasks may be appended with additional time if the user fails to respond consistently (suggesting they are fatigued or fell asleep, and thereby distorting the baseline measures).
a. 3CVT: After the instructions are presented, the 3-Choice Vigilance Task begins with a Practice Session:

During the practice session, the session instructions will alert the subject when an incorrect response is made to insure that he/she understands the task.
As soon as the user demonstrates that they understand each of the responses, the practice session will terminate and the testing session will begin.

The length of the 3CVT (after the practice period) will depend on the baseline type selected. During the first 5-minute period, stimuli appear frequently and require a high state of alertness. The inter-stimulus intervals are extended in the remaining 15 minutes of the 3CVT in order to better identify individuals who are unable to remain engaged (i.e., excessive daytime drowsiness or other sleep related disorders). Subjects who are unable to sustain performance within a normal range across the 3CVT will be flagged as having an invalid baseline session.

www.biopac.com

B-Alert Software Manual b. VPVT: The Visual Psychomotor Vigilance Task begins with the following instructions.

c. APVT: The Auditory Psychomotor Vigilance Task begins with the following instructions.

After completing the third and final Benchmarking Task, a completion dialogue will appear informing the participant to wait for the technician. The technician must press the F11 key at this time to complete the testbed and initiate the final quality check. See Section 3.4 for more information.

Once acquired, the Benchmarking session files will be stored in the subject's folder. The definition file needed for EEG-based Metrics will also be automatically generated and placed in the subject's folder. The .def file is required to ensure validity and accuracy across each individual subject.

www.biopac.com

B-Alert Software Manual
3.3 Acquisition Troubleshooting
During any of the three tasks in the Benchmarking sessions, the user may use the following shortcuts � Ctrl+Alt+Q (Quit): Quit the session. For use by the technician. � Ctrl+Alt+R (Restart): Restart the current task. For use by the technician. � Ctrl+Alt+P (Pause): Pause the current task. It is not recommended to pause the task. � Ctrl+Alt+C (Continue): Resume after Pause. � Ctrl+Alt+J (Jump): Skip the current task. � Ctrl+Alt+V (Review): Review previous instructions before the commencement of the task.
Note: If the session is terminated prematurely before the completion of all the tasks, use Ctrl+Alt+Q shortcut key to terminate the session in order to save all the outputs to the files. Failure to do so will result in all output data being lost.
3.4 Exiting Benchmarking Session
Function Key F11 - is used to get out of a "Technician Requested" window, when subject is kicked out of a task due to a lack of response, or to exit out of the last screen of the Benchmark.
The following are examples of when to use this button:

www.biopac.com

B-Alert Software Manual Upon completing the final task, this screen will appear:
Once the technician exits out of this screen using F11, the main screen of the B-Alert Control GUI will appear on the subject's computer and completion status will be informed by a dialog. Note: After completing the Benchmarking session, BLS will automatically run a data quality and performance check.
Following these checks, the software will create a definition file. Creation of the baseline file may take up to 10 min, during which time the software will appear unresponsive. A "Definition File is created" pop up window will notify the user when processing is complete (5 - 10 min), and benchmark file and summary report are generated.
If the definition file and benchmark report generation are not automatically triggered, or fail for some reason, you can create them offline by selecting the "Create Definition File" button and following the instructions in Chapter 5. The dialogue summary (xxxxx_BenchmarkReport.csv, where xxxxx indicates subject/session ID) will also be created when invoking the action offline.

www.biopac.com

B-Alert Software Manual
If any of the data is determined to be 'bad', the reasons for classifying the definition file as such and suggested action will be presented in this box. The status of each task is based on a three part check: user performance, EEG data quality, and EEG classification breakdown. If any task(s) are identified as being marginal or bad, re-running the flagged tasks is recommended if time permits. After completing the re-run tasks, users can place all three valid tasks into the same folder, and then create a .definition file (.def) on those tasks. These performance and classification thresholds are designed to notify users when BLS identifies an EEG benchmark that does not behave within normal ranges, indicating that the .def file may not provide valid outputs. Users are encouraged to determine whether or not to use a .def file and classification.
The definition file, if created successfully, will be labeled as follows:
xxxx_by_MMDDYY_HHMMSS.def
Where,
� xxxx = subject number � b9, b15, b30 for short, medium, long session � MMDDYY = date � HHMMSS = time
A summary report called "BenchmarkReport.csv" will be created for each session and will provide additional information regarding the quality of the definition file that was created. The sections included in the report are as follows:
� Session Quality o Based on Artifacts o Based on subjects performance o Based on Cognitive State classifications
� Cognitive State Metric classifications for the session o Thresholds for three tasks o Quality based on the thresholds
� Individual/Population Metrics for each task o Quality o % classifications (based on epochs) o Valid/Total epochs used o % artifacts o Average of classifications for the entire task o Performance parameters (Correct responses, Reaction time, lapses)

www.biopac.com

B-Alert Software Manual
Chapter 4: Advanced Settings and Operations
4.1 Taskbar
The settings bar can be hidden by deselecting the settings bar option in the view menu. If the settings bar option is selected, the settings taskbar appears as shown.

1. Find Port: Selecting "Find Port" will locate and assign the serial port for communication between the headset and the B-Alert Live Control GUI. Before clicking 'Find Port,' make sure the headset is connected to its corresponding dongle or ESU by checking the headset for a solid LED light pattern. Note: All operations invoke "Find Port" automatically, thus clicking the Find Port button is optional
2. Headset Config: Select type of headset (in most cases this will be automatically selected; when a wrong configuration is selected the user will be notified).

GUI Display Name: X24 10-20
X24 Visual
X10 Standard X4 Standard X4 APPT X4 Sleep Profiler

Includes:
B-Alert X24 with 10-20 sensor strip or Stat X24 with 10-20 sensor strip
B-Alert X24 with Visual sensor strip or Stat X24 with Visual sensor strip
B-Alert X10 or Stat X10
B-Alert X4 or Stat X4
Reserved for ABM internal testing
Reserved for specialized use

www.biopac.com

B-Alert Software Manual 3. Channel Map Ellipsis Button: Select channel visibility.
Decon. Channel Visibility
Raw signals Visibility. Can also used to select 1) Third-party 2) X, Y, Z tilt 3) Optical (X4)
Additional channels that can be selected are as follows: � Third-party data � events send from third-party software to headset's corresponding USB dongle or ESU � X, Y, Z tilt � output of 3-axis accelerometer in angles � Optical � IRED signal used to derive pulse (available only in the X4)
4.2 Settings
The Settings menu can be used to configure: 1) Data Streaming, 2) Data Storage, or 3) APPT:

www.biopac.com

B-Alert Software Manual
1. Configure Data Streaming Selecting the Data Streaming option opens the Data Streaming Info window:

a. No Data Stream b. TCP and UDP � This protocol allows the user to select from the options listed on the left side of the Data
Streaming Info window to be passed via TCP. When this option is selected, the time stamp information will be passed via UDP, while the selected data is transmitted via TCP: c. Time Stamp: Elapsed time for the current session (seconds and milliseconds). d. Raw EEG: Raw EEG signal. e. Decontaminated EEG: EEG signal after removal of artifacts as identified by ABM decon algorithms. f. Artifact: The start and stop times of artifacts identified in the EEG. g. Heart Rate: the beat to beat heart rate and heart rate variability. h. Classification PSD: Power Spectral Densities from the 1-40 Hz bins of each channel, smoothed using a Kaiser Window (See Appendix A: Data Outputs Guide for more details). The PSDs are also averaged across 3-epoch overlays. i. Raw PSD: PSDs computed without decontaminating artifacts and without applying Kaiser Windows. The PSDs are still averaged across 3-epoch overlays. j. B-Alert Classification: Probabilities of High Engagement, Low Engagement, Distraction, and Sleep Onset (i.e. the Cognitive State Metrics). k. Workload Classification: Probability of High Workload. l. Head Movement: Accelerometer data. m. Third Party Data: Third-party events sent to ABM receiver (ESU). n. IP Port: Designates the IP Port that the server opens for all clients to connect. o. UDP Only � passes only the Time Stamp data via UDP.

www.biopac.com

B-Alert Software Manual
2. Configure Data Storage
This option allows users to configure which BLS output files are created during Data Acquisition (in both 'Acquisition' and 'Acquire & Retransmit'). The default setting is to save all output types, but users have the option to deselect outputs as desired.

3. Configure APPT and Sleep Profiler Reserved for ABM internal R&D.
4.3 Operations
The operations menu can be used to sync ABM devices to an ABM receiver, configure an ABM receiver, view hardware details, upload firmware, and invoke remote monitoring of a Benchmarking Session.

www.biopac.com

B-Alert Software Manual
1. Sync Devices The B-Alert headsets are shipped paired to either a B-Alert Dongle or an ESU (ESU). If the pairing is lost, the devices must be synced again using the "Sync Device" function in the Operations Menu.
a. Plug the ESU receiver into your PC. Note: Ensure that all nearby B-Alert products, including headsets and ESUs, are turned off and unplugged.

b. Switch ON the B-Alert headset. Wait until the headset LED pattern is flashing Green before proceeding to the next step:

Flashing GREEN
c. In the Operations menu, click on "Sync Devices". d. Click the "Search Devices" button and wait. It will take 45-60 seconds for the USB device to search for
other Bluetooth devices in range. e. After the search has completed, all Bluetooth devices identified in range will appear under the "Select
Device" drop down menu. f. Look for the B-Alert headset. The device should have a 12 digit BDA address followed by: (ABM EEG
XXXXXXXXXXX). Example: 000BCE042CCA (ABM EEG 300103056):

g. Click "Connect" and wait for the pop-up window indicating that the headset has connected: "Device connection succeeded. Please restart device."

www.biopac.com

B-Alert Software Manual
h. Click 'OK,' then switch OFF the Headset. i. Unplug the ESU receiver, and then re-plug it into the laptop/computer. j. Switch the headset back ON; wait 5-10 seconds for the Green LED on headset to turn Solid Green,
signaling the headset has synced. 2. Configure ESU Users with the ESU have the option to configure the serial and parallel port settings. Please refer to Chapter 6 for additional details.

3. Hardware and Software ID
The 'Hardware and Software ID' option allows users to check the HW settings of the B-Alert Headset as well as any corresponding USB dongles or ESU receiver(s).

Note: The information in this window is valuable for debugging purposes. If possible, please send a snapshot of this window during all hardware-related correspondence with support representatives.

www.biopac.com

B-Alert Software Manual Upload Firmware to Headset The "Upload Firmware to Headset" option allows users to update the headset firmware. Please note that a corresponding USB dongle or ESU receiver is required. The procedure is as follows: a. Connect the headset to the dongle or ESU (the solid Green LED confirms connection). Select "Upload Firmware to Headset" from the Operations menu. b. Click the 'Browse' button to select path to the firmware .txt file:
c. After selecting the Headset firmware .txt file, click the 'Upload' button:
d. Wait 20-30 seconds. A pop-up window will appear to notify you that the firmware upload was successful:

e. Finalize the firmware upload by restarting the Headset (i.e., switching off/on).
Note: For X4, firmware cannot be uploaded wirelessly (hence "Find port" is disabled). Please connect the device to the computer using a USB cable and follow the same procedure.
4. Upload Firmware to ESU The "Upload Firmware to ESU" option allows users to update firmware on the headset's USB dongle and/or ESU receiver. The procedure is as follows:
a. Plug in the USB dongle or ESU receiver. Select "Upload Firmware to ESU" from the Operations menu.
b. Click the 'Browse' button to select the path to the firmware .hex file:

www.biopac.com

B-Alert Software Manual c. After selecting the ESU receiver firmware .hex file, click the 'Upload' button. d. Wait 15-30 seconds. A pop-up window will appear to notify you that the firmware upload was successful:
e. Finalize the firmware upload by restarting the ESU receiver (unplugging/re-plugging).

www.biopac.com

B-Alert Software Manual
Chapter 5: Data Analysis and Review
The B-Alert Live Software has four offline action icons which are used for off-line data analysis, after data acquisition:

5.1 Inspect EEG Record
To begin Playback of a previously recorded file, click the "Inspect EEG Record" icon. This will open a browse window. From the browse window, select the desired .EBS (or .EDF) file and click the Open button. The Play EBS Settings dialog will then be presented. Select the playback mode by either checking or un-checking (for off-line playback) the Real-Time Playback check box.

1. Playback: Previously acquired sessions can be reviewed in either Real-Time or Offline Playback.
a. Real-Time Playback � The file will be opened and presented in real-time, exactly the same as when the recording was initially acquired. Real-Time Playback mode is ideal for matching events of interest up to video recordings.
b. Offline Playback � The file will be opened with the EVA offline analysis software. This allows the user to scroll forward and backward within the session to view signals and classifications. Offline mode is ideal for review of artifact problems. See Section 9.2 Offline Playback using EVA for more details.
2. Brain State: For Real-Time Playback session, it is optional to select whether the Cognitive State and Workload Metrics classifications are included. Artifact decontamination is also optional.
3. Definition File: If the subject has a Benchmark definition file created and the classification options are both checked, then the definition file must be selected. Note: the definition file selected must have the same 4-digit subject number as the session subject number. This option is only applicable for Real-Time Playback and not for Offline Playback (EVA).

www.biopac.com

B-Alert Software Manual
5.2 Offline playback using EVA
If 'Real-Time Playback' is left unchecked, the entire data file will be opened using the EVA software (the offline data viewing platform). After clicking 'OK' with the 'Real-Time Playback' box unchecked, EVA will launch:
Scale Bar

Channel Labels

+/ - Click the + or symbols to change display range for
EEG or ECG channels

Slider Bar 1. Toolbar:

Status Bar

The toolbar controls the behavior of one channel when the channel is selected. To select a channel, hold the Ctrl key and left-click on the channel. The selected channel will be highlighted in blue. Once selected, the buttons on the toolbar can be used:
Expand Selection: Expand the selected channels and hide all other channels from view.
Remove Selection: Remove a selected channel from the view.
Undo Selection: Undo the last edit made.
Zoom Up: Change the scaling of the selected channel by zooming in.
Zoom Down: Change the scaling of the selected channel by zooming out.
Zoom Default: Change the scaling of the selected channel back to the default values. 2. Status Bar:

www.biopac.com

B-Alert Software Manual A number of time measures are available to identify periods of interest in a file. These time measures are based on the location of the mouse as you move it horizontally across the screen: X dist: Not currently supported. Y dist: Not currently supported. Slope: Not currently supported. CMT (Clock Mouse Time): Allows the user to see a point in time based on the inserted clock time. EMT (Elapsed Mouse Time): The Elapsed time since the start of the record. CST (Clock Start Time): The time at the start (left edge) of the screen. This value will not change as the cursor is moved across the screen. EST (Elapsed Start Time): The record elapsed time at the start (left edge) of the screen. This value will not change as the cursor is moved across the screen.
3. Slider Bar:

A number of methods are available to scroll through the file. The icons in the bottom left and right corners of the monitor shift the file:

Backward to start of the file Backward one screen length Backward one second

Forward to end of the file Forward one screen length Forward one second

Pg Up: Backwards one screen length

Pg Dn: Forward one screen length

Left-click & drag the Pointer at the bottom of the screen to move anywhere in the file. 4. Scale Bar:

The options on the scale bar are used to modify the presentation of the signals as they appear with the EVA program in offline playback:
Time Scale: Adjusts the time scale for the signals presented on the screen.
Back: After the time scale is adjusted, the Back button will revert to the previous time scale. Set 1, Set 2, Set 3: The default channel configurations recommended by ABM are presented when EVA opens. Set 1 is defaulted to show the raw EEG and ECG signals; users can select different channels to open in sets (similar to Tabs). Use the buttons to toggle between Sets 1, 2, and 3.
Save as Default: If any adjustments are made to the presentation of signals on Sets 1, 2, or 3 and it is desired to keep these settings as the default, use the Save as Default button. The next time EVA is opened, the new default settings will be presented.

www.biopac.com

B-Alert Software Manual Close View: Closes the default signal view and opens the channel configuration view. To return the default signal view, click the Open View button located on the channel configuration view. Use the 'InSet1', 'InSet2', and 'InSet3' check boxes to default channels to open in different sets in EVA.
5. Right Click Menu: Right click on the channels in the Main Window to bring up the Right-Click Menu:
a. Signal properties: � Phys. Min/Phys. Max and Digital Minimum/ Digital Maximum: adjusts the Y axis maximum and minimum ranges (in �V)

� Zoom Increment: Changes channel view by a zoom factor � Size of Signal (%): Resizes channel range based on % � ISI average and Average (mean): Not supported � Sample Rate: A fixed value and based on the samples/second in the .ebs � Total Sample: Total # of data samples within a given .ebs � Smoothing: Obsolete (not currently supported). � In Set: Obsolete (not currently supported).

www.biopac.com

B-Alert Software Manual
b. Channels: To adjust which channels are presented on the screen, right-click on the screen and select the channels option from the pop-up menu. Select or de-select the desired channels you wish to be visible in the current Set.

c. Monitor Properties: allows users to customize the display settings: fonts, colors, and axis properties. d. Auto Scale: Adjusts the Channel Y axis range based on Max/Min values for that channel across the entire
file. e. Auto Scale All: Adjusts the Channel Y axis range based on Max/Min values for each channel in Set 1
across the entire file.
5.3 Create Definition File
After completing a Benchmarking session, the definition (.def) file is automatically generated and saved to the subject folder. This .def file is used to generate the Benchmark Report file. If the .def file needs to be recreated, the "Create Definition File" icon can be used. Note: Signal .edf and all baseline Events .edf files must be located within the same folder, and have the same 9-digit # for a .def file to generate properly. Follow the steps below to create a .def file.
1. Acquire necessary Benchmarking data for that individual subject from each of the three tasks (3CVT; VPVT; APVT).
2. Click the "Create Definition File" icon. The Definition File Settings window will open.

3. Browse Signal Files to locate the *Signals.Raw.edf file in the session sub-folder for the desired session 4. Browse Event Files to locate the 3file needed: the 3CVT, VPVT, and APVT *Events.edf files that correspond
to the Signal.Raw.edf file previously selected 5. Click OK.
6. After several minutes, the .def file creation will finish. 7. When the definition file creation is complete, the "Definition File Report" window will be presented. This
window summarizes the results of .def file creation. If either the performance or EEG artifacts are outside of normal acceptable ranges, then you may see recommendations in this report to re-run Benchmark task(s). Once created, the individualized .def file can be reused across sessions (for Cognitive State and/or Workload Metric classifications) for a given individual.

www.biopac.com

B-Alert Software Manual

8. Check for the Definition File Report in the folder path to ensure the .def file was created. 9. Note: In case of repeated benchmark, user must input all applicable (up to 3) Signal .edf files, but *only* the
three correct Events .edf files. 10. For additional consultation regarding definition file validity, please contact your technical support
representative.
After completing a Benchmarking session or running the Create Definition file function, the Benchmark Report is automatically generated and saved to the subject folder alongside the .def. The Benchmark Report summarizes performance, classification data, and PSDs (along with heart rate, if available) across the three benchmarking tasks: the 3-Choice Vigilance Task (3-CVT; *.Events.edf), Visual Psychomotor Vigilance Task (VPVT; *.Events.edf), and Auditory Psychomotor Vigilance Task (APVT; *.Events.edf). A Benchmark Report will also be generated automatically during the definition file creation (see section below). For comparison purposes, population norms from a database of healthy, fully rested participants are included in the Benchmark Report, for each of the three tasks.
5.4 Generate Summary File
To create data outputs with previously recorded data sessions, users can use the "Generate Reports" function. Consult Appendix A: Data Outputs Guide for additional information on the data output procedures and file descriptions.
1. Click on the "Generate Reports" icon.
2. The Generate Reports window will appear:

3. Browse to and select the desired .ebs file. The .def file must be placed alongside the .ebs file to generate reports.
4. Click OK.
� Wait for the Pop-Up window to appear, indicating summary file has completed:

www.biopac.com

B-Alert Software Manual
5. Summary.csv will be created with the following contents: a. 30 second averages of HR b. 30 second averages of bandwidths c. 30 second averages of classifications (if definition file is selected)
5.5 Offline Analysis
1. Click on the "Offline Analysis" icon. The Generate Outputs window will open:

2. Browse to and select the desired .edf or .ebs file. The .def file can be placed alongside the .ebs/.edf file to generate outputs.
3. Wait for the pop-up window to appear, indicating offline analysis has completed:

4. The following files can be generated using offline analysis
a. HR_beat.csv : Beat by Beat detection and computation of Heart Rate b. HR_epoch.csv: Epoch by epoch detection of heart rate c. Diff_class.csv : Decontaminated differential channel PSD d. Diff_raw.csv: raw differential channel PSD e. Ref_class.csv : Decontaminated referential channel PSD f. Ref_raw.csv : raw referential channel PSD
Note: For comprehensive processing of data, ABM offers standalone offline processing software called BAlert Lab. Please contact customer support for more information.

www.biopac.com

B-Alert Software Manual
Chapter 6: External Sync Unit (ESU)
6.1 General ESU Information
ABM B-Alert Device

BluWeIintroeteolertfhsasce

ABM MC-ESU

USB

B-Alert SDK

API

Third

Party

Application

bin file

WindowsPC

COM2 COM3 COM4

Third Party
1

Third Party
2

Third Party
3

Parallel Port

Third-party 4

www.biopac.com

B-Alert Software Manual

COCMOM1 1 (S(eSreiarila1l 1
PoProtr)t):
COM1 not currWenirteldy supHpeoardtseedts
ONLY

COM2 (Serial 2
Port)

COM3 (Serial 3
Port)

COM4 (Serial 4
Port)

Parallel Input

USB Mini (to PC)

The External Syncing Unit (ESU) can be used to timestamp data from both ABM devices and events from third party applications at millisecond level precision through a dedicated robust hardware timer. This addresses the limitations of time stamping using Windows timers, which introduces an average variable latency of ~30-50 msec in most circumstances due to the Windows scheduler. This variable latency can cause havoc in EEG signal analysis, especially in studies that rely on synchronization with external stimuli in order to extract event-relatedsynchronization/desynchronization as it reduces the Signal-To-Noise-Ratio (SNR) when averaging EEG samples to extract features such as evoked potentials. The ESU thus eliminates Windows-related variable delay by time stamping data packets externally.
The ESU also functions as a Bluetooth receiver for ABM devices, and it has the ability to decode ABM protocol and stream the data to the SDK via the USB port (i.e., the USB port registers a Virtual COM port in the PC). B-Alert X10 and X4 devices bundle two samples from each acquired signal into a single packet that gets time stamped every 8 msec, while the X24 sends only one sample and is time stamped every 4 msec. COM1 is currently not supported for connections, while COM 2, 3, and 4 can be used to acquire third-party events via RS232 serial-port protocol. Applications can also send events via the parallel port, either with or without the STROBE signal. All third party events will be time stamped without delay in the ESU, however, in the event of conflict, EEG packets will have priority over other data. Proprietary cables supplied with the ESU (see Table 1) must be used to send data. All thirdparty applications must follow ABM protocol (specified in this chapter) while sending data to the ESU. The SDK will unpack the third-party events and store them in a bin file or an Events.EDF file along with ebs/edf data files for offline analysis. The third party events can also be acquired in real-time using SDK APIs.

www.biopac.com

B-Alert Software Manual Table 1: ESU Cables Cable Name USB Cord (connect ESU to PC)
Serial Port to ESU cable
Parallel to ESU Cable

Photo

www.biopac.com

B-Alert Software Manual
6.2 ESU Block Diagram

X10: EEG data is read in packets of 2 samples/channel. As EEG is sampled at 256Hz (=3.91ms), the read operation takes place ~ every 8ms. Each packet is time stamped (but samples within the packet are not � they are equidistant).
X24: Each data packet contains 1 sample/channel, thus each packet will be time-stamped ~ every 4ms.
Note: Bluetooth transmission will add more delays.

EEG Headset

UART

transceiver

Third-party device

Events Events

Parallel port Serial port

ESU

Serial Port: Configure serial port as: 57600, 8n1. Send data in ABM Universal third party data format.

Parallel Port: There are two ways to send data via parallel port: 1) Structured, i.e., using STROBE to signal that valid data is present on 8-bit data lines; or 2) Un-structured, i.e., signalizing valid data using changes in 8-bit data lines. The lines are sampled every 0.1ms (=precision for detection of a state change). The new state (1 byte) is only stored in a buffer when the state changes. Time-stamping is done with ms precision.

www.biopac.com

B-Alert Software Manual
6.3 Configuring the ESU Settings
1. Plug in the ESU into USB port on a computer. 2. Open B-Alert Live Control GUI:
3. Under the 'Operations' Menu, select ESU Configuration. The ESU Configuration window will appear:

4. Use the drop down menus to configure ESU Third Party Data (3 Serial Ports and 1 Parallel Port) as desired. See Section 6.4 Third Party Data Format (below) for the specifications for the different T-Party settings.
Warning: Do NOT use the port labeled COM1 in the ESU for sending events from third-party interfaces. COM1 is currently not supported.
5. Click the 'Write' Button to write settings changes to the ESU. 6. Wait. A pop-up window will appear to confirm the configuration was saved. Restart the ESU by unplugging it
and then re-plugging it into the computer's USB.

www.biopac.com

B-Alert Software Manual

6.4 Third Party Protocols and Packet Structure

The ESU supports multiple protocols for acquiring third party data. The specifications of the protocols are listed below:

Serial communication protocols

Type

Baud Rate (kb/s)

PNNL (recommended) 57600

SMI

9600

ASL

19200

AMP

Reserved

Data bits 8 8 8 8

Stop bit 1 1 1 1

Parity None None None None

Parallel communication protocols

Type

STROBE (YES/NO)

PNNL

Yes

ANITA

Yes

DAIMLER

E-PRIME (recommended)

Packet Structure (YES/NO) Yes Yes No* No*

*For protocols without a packet structure, the changes in the 8-bit parallel lines are time-stamped. Idle values (0) must be used in between valid data and the values must be held for at-least 100 uses for registration.

www.biopac.com

B-Alert Software Manual
1. Sending data to the ESU The third party applications must send packets according to the ABM Universal Protocol to the serial and parallel ports of the ESU. The packet follows BIG ENDIAN format and has the following fields:

Flag 0x56, 0x5A
(2 bytes)

Packet Length (2 bytes)

Packet Type (1 byte)

User Data (Number of bytes =
Packet Length)

Check Sum (1 byte)

� Flag: fixed (0x56,0x5A) � Packet Length: packet length is equal to the number of bytes in the User Data field � Packet Type: depends on the type of protocol
o PNNL (Serial = 1, Parallel = 2) o SMI (Serial = 3) o ASL (Parallel = 4) o AMP (Parallel = 8) o ANITA (Parallel = 6) o DAIMLER (Parallel = 7) o EPRIME (Parallel = 10) � User Data: � Checksum = 255 - Mod(SumOfAllBytes(Packet Length, PTY, DATA), 256) Note: Sample programs for reference that mimic sending data through the serial port and the parallel port can be found within the installation folder.
2. Decoding data from the ESU B-Alert Live Software unpacks the data received from the ESU and stores it in a bin file, or an events edf+ file. The events packet has the following structure for each event:

Flag 0x56, 0x56
(2 bytes)

Message Counter (1 byte)

ESU Timestamp
(4 bytes)

Packet Length (2 bytes)

Packet Type (1 byte)

User Data (Number of bytes =
Packet Length)

Check Sum (1 byte)

� Flag: fixed (0x56,0x56) � Message Counter: (reserved ) � ESU Timestamp: 4-bytes ESU timestamp in BIG ENDIAN format (the timestamp can be reconstructed
using multiples of 2 (for example, timestamp in millisec = BYTE1*2^24 + BYTE2*2^16 + BYTE3*2^8 + BYTE4*2^0 ) � Packet Length: equal to the number of bytes in User Data field � Packet Type: protocol used � User Data � Checksum: Mod(SumOfAllBytes(Packet Length, PTY, DATA), 256)

www.biopac.com

B-Alert Software Manual
6.5 Visualizing Third Party data during Data acquisition
To monitor the stream of incoming data to the ESU (Serial and Parallel), users can select the additional channels for display from the Channel Map Window (see Chapter 4: Advanced Settings and Operations). After selecting the 'Third Party' option, a 'Third-Party' channel will appear at the bottom of the raw signals window. All incoming packets to the ESU (if formatted in proper structure) will appear as strobes in this line:
Serial/Parallel Packet received by ESU
Note: These strobes will only reflect the time when the packet was received by the ESU, and will not contain any information about the content of the packet.

www.biopac.com

B-Alert Software Manual

Appendix A: Data Outputs Guide
A.1 Overview
The B-Alert Live Software (BLS) allows users to utilize patented and peer-review validated EEG-based Metrics, Automatic Decontamination, PSD, and ECG algorithms if desired. These data outputs are either created in real time (See Real-Time Data Outputs Overview Table below) or offline. Below is an overview of some of the methods and techniques used by BLS to compute its various measures.
1. Computing Power Spectral Densities: Power spectral density (PSD) is computed by performing Fast Fourier Transform (FFT) on a segment of data that is of interest, and calculating the amplitudes of the sinusoidal components for designated frequency bins. Input variables to this transformation are an EEG segment for which PSD is to be computed, and its length; output variables include PSD amplitudes. Frequency domain variables are based on the power spectral density derived after application of a 50% overlapping window, and a FFT with ('_raw') and without ('_class') application of a Kaiser window. The B-Alert software provides two sets of PSD values (Ref_Raw and Ref_Class) from 1 to 40 Hz for each EEG channel that are logged to obtain a Gaussian distribution. Selected 1-Hz bins are averaged, then logged to create conventional EEG bands (e.g., theta = 3 � 7 Hz, alpha = 8 � 13 Hz, etc.). 'Diff_' files contain data for Differential EEG channels used for the Cognitive State Metric classifications (FzPOz, CzPO, FzC3, C3C4, and F3Cz). 'Ref_' files contain PSD values for 9 referential EEG channels (POz, Fz, Cz, C3, C4, F3, F4, P3, and P4).
Both sets of PSD output files apply a 50% overlapping window which averages the PSD across three x onesecond overlays to smooth the data. The illustration below shows that overlays 0, 1, and 2 are averaged (each overlay containing 256 data points with 128 data points being shared for each overlay) to provide the PSD values for epoch n:

50% Overlap � Ref_Raw, Diff_Raw

Epoch n

Overlay 1

Overlay 0

Overlay 2

50% Overlap � Ref_Class, Diff_Class (Kaiser windowing applied)
Epoch n

For the 'Ref_Class.csv' and 'Diff_Class.csv' PSD calculation, a Kaiser window is applied to each overlay in order to accentuate the contribution of power from the signal in the middle third of the overlay, and minimize the impact of signal near each edge of the overlay. Windowing reduces the likelihood of extreme PSD values resulting from edge-effects when an EEG wave shape does not begin or end at the exact edge of an overlay. No Kaiser windowing is used for Ref_Raw and Diff_Raw analysis. ABM recommends using the _Raw data for users wishing to use PSD-based computations using this PSD overlaying procedure.
2. Decontaminating Signals: Prior to computing the 1-Hz PSD bins, the raw signals are processed to eliminate known artifacts. Spikes, excursions and amplifier saturations, which occur when ambulatory EEG is acquired, can impact both low and high frequencies. EMG will contaminate the beta and gamma frequency ranges. Eye blinks (EOG) occur in the same frequency range as theta activity.

www.biopac.com

B-Alert Software Manual

Before Decontamination:

After Decontamination:

� Excursions and amplifier saturation � contaminated periods are replaced with zero values, starting and ending at zero crossing before and after each event.
� Spikes caused by artifact are identified and signal value is interpolated.
� Invalid Epochs � If more than 128 zero values are inserted for an overlay, the overlay is excluded from the epoch average; if 2 of the 3 overlays are rejected, the epoch is classified `invalid' (-99999 inserted for PSD value) and should be excluded from analysis.
� EMG � a combination of High Frequency EMG (based off 70 - 128 Hz bins for each overlay) and Low Frequency EMG (based off 35 - 40 Hz) is used to identify periods with excessive EMG. If only one overlay has EMG, the PSD for the epoch is based on the average of the remaining two overlays. If excessive EMG is detected in two overlays, the second is classified as `EMG' and should be excluded from analysis.
� Eye Blinks (EOG) � wavelet transforms deconstruct the signal and a regression equation is used to identify the EEG regions contaminated with eye blinks. Representative EEG preceding the eye blink is inserted in the contaminated region.
The artifacts are colored as follows:

Artifact Type EMG Eye-blinks Saturation Excursion Spike

Color Orange Blue Light Blue Dark Green Dark Brown

www.biopac.com

B-Alert Software Manual 3. Classifying the Cognitive State And Workload Models:

The EEG-based Metrics are normalized to an individual subject using 5-7 min of baseline data from three distinct tasks (3CVT, VPVT, and APVT), with the sleep onset classification predicted from the baseline PSD values, for a total of 15-17 min of data across tasks. Based on this identification data, a probability-of-fit is then generated for each of the four classes for each epoch with the sum of the probabilities across the four classes equaling 1.0 (e.g., 0.45 high engagement, 0.30 low engagement, 0.20 distraction and 0.05 sleep onset). Cognitive State for a given second represents the class with the greatest probability using numeric labels (.1 = sleep onset, .3 = distraction, .6 = low engagement, and .9 = high engagement). The Cognitive State Metrics are derived for each one-second epoch using 1 Hz PSDs (from the bins from differential sites FzPOz, CzPO, FzC3, C3C4, F3Cz) in a four-class quadratic discriminant function analysis (DFA) that is fitted to the individual's unique EEG patterns. The table below identifies and briefly describes each baseline task, and associates the task with the B-Alert classification.

Benchmark Task
3CVT
VPVT APVT None

Action
Discriminate between primary vs. secondary or tertiary stimulus every 1.5 to 3 seconds Respond to visual probe every 2 seconds Respond to audio tone every 2 seconds Derived by regression from other 3 tasks

B-Alert Class probabilities High Engagement
Low Engagement Distraction Sleep Onset

Important Note: Failure to collect these 3 specific baseline tasks will result in an inability to utilize the B-Alert Cognitive State Classifications.

www.biopac.com

B-Alert Software Manual
The Cognitive State Metric probabilities for each individual should be interpreted in a relative, rather than absolute, manner. Three standardized benchmarking tasks normalize the metrics to each individual. High population variability for EEG activity requires individualized model fitting, which is done for each 1Hz bin (from 1-40Hz), and is not fit to classic summed bandwidths/rhythms (i.e., theta, alpha, beta, etc.) to optimize classification measures.
Two individuals will generate somewhat different probabilities for the same task due to a) their innate capability, and b) their state during the benchmarking tasksIf a participant is mentally balancing their checkbook during the APVT task, for example, they will not generate as much alpha activity as they would in a relaxed state. This may increase the occurrence of Distraction probabilities when applied to a different task in which they do mentally relax. Participants are more aroused the first time they complete a benchmarking session due to the novelty, so it is preferable to reuse the individual's DFA to classify new data/sessions rather than re-run the identification tasks repeatedly. Participants should avoid consuming excessive caffeine or nicotine immediately prior to conducting their Benchmarking Tasks data; and the session should occur in the morning (8am-10am) after a full night of sleep to collect an optimal session.
The Cognitive Workload Metric is a generalized model (i.e., it is not individually fit), thus it should also be interpreted in a relative manner. For the linear 2-class workload DFA, probabilities closer to 1 reflect higher workload. EEG workload is correlated with increased working memory load and difficulty level in mental arithmetic and other complex problem solving tasks. ABM has 2 workload models -- one model was built on a Forward digit span (FBDS) task (recommended to use, as it fits for ~85% of population) and the other built on a backward digit span (BDS) task (fits ~15% of population). ABM's data outputs also contain the mean probability between the FBDS and BDS model.
Z-scoring is a useful transformation to convert the relative Cognitive State Metrics into values that can be compared across participants, or for a repeated-measures within-subject experimental design.
4. Output files: EEG and other raw data are stored in both EBS and EDF files. The EDF file is compatible with most of the third-party EDF viewers. The csv output files generated with the B-Alert Live software share common formatting features. For example, all file names begin with the nine digit subject/session number (XXXXXXXXX), followed by a 6 digit date (YYYYYY), followed by a 6 digit timestamp(ZZZZZZ), followed by the label which describes the data. For generated files, one row of data is provided per second of recording time. The first column lists the subject/session number, the second column the elapsed time (since the start of recording) in hour:minute:second:millisecond (HH:MM:SS:MS), and the third column the system clock time associated with the start of the primary (middle) overlay for the epoch. Output files use a comma separated value (CSV) format for easy import into statistical/analytical software applications.

Real-Time Data Outputs Overview Table

Real-Time Output file name

Description

Data file

Xxxxxxxxx.yyyyyy.zzzzzz.ebs Xxxxxxxxx.yyyyyy.zzzzzz.Signals.Raw.edf
Xxxxxxxxx.yyyyyy.zzzzzz.Decon.edf Xxxxxxxxx.yyyyyy.zzzzzz_Events.edf Xxxxxxxxx.yyyyyy.zzzzzz.Impedance.Results.csv

Proprietary Extensible bio-signal file format. Stores raw data with identical sampling such as EEG and EKG. European data format containing raw data with variable sampling such as EEG, EKG, Accelerometer data, Marker etc. The EDF file is compatible with most third party viewers. Some channels in EDF file are selectable (see Chapter 4: Advanced Settings and Operations). Decontaminated data from Signas.Raw.edf Events generated by the ESU stored in EDF+ annotation format. Lists the values obtained for each channel each time impedance was measured.

Automatically Generated during Acquisition � for all EEG Channels

Xxxxxxxxx.yyyyyy.zzzzzz.Ref_Raw.csv

Absolute PSD from 1 to 40 Hz, relative PSD from 1 to 40 Hz, and EEG bands labeled by channel (no edge-effect window)

www.biopac.com

B-Alert Software Manual Xxxxxxxxx.yyyyyy.zzzzzzRef_Class.csv

Absolute PSD from 1 to 40 Hz, relative PSD from 1 to 40 Hz, and EEG bands labeled by channel (with Kaiser window)

Automatically Generated during Acquisition � Derived Signals

Xxxxxxxxx.yyyyyy.zzzzzz.HR_beat.csv Xxxxxxxxx.yyyyyy.zzzzzz.HR_epoch.csv

Presentation of heart rate based on beat-to-beat interval Beat-to-beat heart rate interpolated to sec-by-sec value

Optionally Generated with Cognitive State Metrics

Xxxxxxxxx.yyyyyy.zzzzzz.Classification.csv Xxxxxxxxx.yyyyyy.zzzzzz.Diff_Raw.csv Xxxxxxxxx.yyyyyy.zzzzzz.Diff_Class.csv Xxxxxxxxx.yyyyyy.zzzzzz.Zscore.csv Xxxxxxxxx.yyyyyy.zzzzzz.Zscore_psd.csv

Probabilities for sleep, distraction, low and high engagement, cognitive state from DFA with greatest probability, probability of high workload based on forward and backward digit span (FBDS), backward digit span (BDS), and average of FBDS and BDS. Absolute PSD from 1 to 40 Hz, relative PSD from 1 to 40 Hz, and EEG bands for differential channels: FzPO,CzPO,FzC3,C3C4, and F3Cz (no edge-effect window). Absolute PSD from 1 to 40 Hz, relative PSD from 1 to 40 Hz, and EEG bands for 5 differential channels FzPOz, CzPO, FzC3, C3C4, F3Cz (with Kaiser window). Updates and applies mean and standard deviation with each new second to provide z-scores for both the Cognitive State (sleep onset, distraction, low and high engagement) and Workload Metrics ( three workload measures). Updates and applies mean and standard deviation with each new second to provide z-scores for PSD for all channels requested in initialization process.

Other files

Xxxxxxxxx.yyyyyy.zzzzzz.Actigraphy.csv Xxxxxxxxx.yyyyyy.zzzzzz.Artifact.csv Xxxxxxxxx.yyyyyy.zzzzzz.ArtifactInfo.csv
Xxxxxxxxx.yyyyyy.zzzzzz.thirdparty.bin Xxxxxxxxx.yyyyyy.zzzzzz.optical.bin Xxxxxxxxx.yyyyyy.zzzzzz.accelerometer.csv

Contains raw tilt values from accelerometer and derived movement vales/scales. Contains detected artifacts and their start/end markers. Contains epoch-by-epoch details of artifacts detected in each channel such as number of datapoints affected, number of datapoints with inserted zero values, etc. Stores third-party packets sent to the ESU (only for ESU). Stores IRED data from the headset (only for X4). Stores raw accelerometer data in bin format.

Note: The files created during acquisition depend on the selection in "Configure Data Storage" setting (see Chapter 4: Advanced Settings and Operations). It also depends on the type of processing selected (Artifact Decontamination), EEG-based Metrics (Cognitive State and Workload classifications), and type of the device used (X4, X10, X24).

www.biopac.com

B-Alert Software Manual

A.1 EEG-based Metrics including Cognitive State & Workload Outputs

1. _Classification.csv

Classification.csv shows the second by second (epoch by epoch) data outputs for ABM's two EEG models: (1) 4-Class Cognitive State Metric model of drowsiness: (Sleep Onset, Distraction, Low Engagement, High Engagement); and 2-Class model of Workload (High Workload, Low Workload). These outputs will only be generated if the user has checked either of the 'Brain State' check boxes (Cognitive State Metrics and/or Workload Classification), and selected the appropriate .def file prior to initializing a data Acquisition.

Column Column Name

Description

SessionNum

Session Name (.ebs File Name)

Elapsed Time

Total Elapsed Time for session in seconds (hh:mm:ss:ms)

Clock Time

Local Computer Time(hh:mm:ss:ms)

ProbSleepOnset Sleep Onset classification probability (0-1)

ProbDistraction

Distraction classification Probability (0-1)

ProbLowEng

Low Engagement classification probability (0-1)

ProbHighEng

High Engagement classification probability (0-1)

CogState

The highest Probability in columns D, E, F, G will determine what

the epoch is classified. Classifications are: 0.1: Sleep onset, 0.3:

Distraction, 0.6: Low Engagement, 0.9: High Engagement.

Seconds with excessive artifact where classification data could

not be computed are identified with 0.05, 1, and/or 2.

ProbFBDSWorkload Raw Workload probability (FBDS model), where higher

probability reflects higher WL (FBDS is the best model for 85% of

population).

ProbBDSWorkload Alternate WL model (BDS model): Not recommended for use,

higher probability reflects higher WL (Best WL model for other

15% of population).

ProbAveWorkload Average Workload between 2 models (in Columns I and J).

A.3 Power Spectral Densities Outputs

_Diff_Class.csv

_Diff_Raw.csv

_Ref_Class.csv

_Ref_Raw.csv

Power spectral density (PSD) is computed by performing Fast Fourier Transform (FFT) on a segment of data that is of interest, and calculating the amplitudes of the sinusoidal components for designated frequency bins. Input variables to this transformation are an EEG segment for which PSD is to be computed, and its length; output variables include PSD amplitudes. Frequency domain variables are based on the power spectral density derived after application of a 50% overlapping window, and a FFT with (_raw) and without (_class) application of a Kaiser window. Refer to Outputs Overview above for additional information regarding the PSD analysis procedures.

1. Diff_Class.csv

PSDs (1-40Hz) for the differential channels (FzPOz, CzPO, FzC3, C3C4, F3Cz) are computed for generating ABM's classifications for each second of a given .ebs file. PSDs in this file are computed for each second of a given session with the Kaiser Windowing procedure described above. Relative power values are derived by subtracting the logged power of the individual Hz bin from the summed logged power for the EEG band (1-40 Hz) for that channel.

Column Column Name

Description

SessionNum

Session Name (.ebs File Name)

Elapsed Time

Total Elapsed Time for session in seconds (hh:mm:ss:ms)

www.biopac.com

B-Alert Software Manual

Clock Time

fzpoz_1

fzpoz_2

F - AQ fzpoz_3-40

fzpoz_rel1

AS - CE fzpoz_rel2-rel40

fzpoz_Delta_1_2

fzpoz_ThetaSlow_3_5

fzpoz_ThetaFast_5_7

fzpoz_ThetaTotal_3_7

fzpoz_AlphaSlow_8_10

fzpoz_AlphaFast_10_13

fzpoz_AlphaTotal_8_13

fzpoz_Gammma_25_40

fzpoz_Beta_13_29

czpoz_1

CP - QF

ThetaOverall_3_7

AlphaOverall_8_13

BetaOverall_13_29

GammaOverall_25_40

Local Computer Time or ESU TimeStamp (if configured) (hh:mm:ss:ms) PSD power at channel FzPoz (with Kaiser windowing) for the 1 Hz bin PSD power at channel FzPoz (with Kaiser windowing) for the 2 Hz bin PSD power at channel FzPoz (with Kaiser windowing) for the 3-40 Hz bins Relative PSD power at channel FzPoz (with Kaiser windowing) for 1 Hz Bin Relative PSD power at channel FzPoz (with Kaiser windowing) for 2-40 Hz Bin PSD for Delta Bandwidth at channel FzPoz (not relative PSD) summed from Hz bins 1-2 PSD for Theta-Slow Bandwidth (not relative PSD) summed from Hz bins 3-5 PSD for Theta-Fast Bandwidth (not relative PSD) summed from Hz bins 5-7 PSD for Theta-Total Bandwidth (not relative PSD) summed from Hz bins 3-7 PSD for Alpha-Slow Bandwidth (not relative PSD) summed from Hz bins 8-10 PSD for Alpha-Fast Bandwidth (not relative PSD) summed from Hz bins 10-13 PSD for Alpha-Fast Bandwidth (not relative PSD) summed from Hz bins 8-13 PSD for Gamma Bandwidth (not relative PSD) summed from Hz bins 12-15 PSD for Beta Bandwidth (not relative PSD) summed from Hz bins 13-29 PSD power at channel CzPOz (with Kaiser windowing) for the 1 Hz bin PSD/Relative PSD information for all differential channels (CzPOz, FzC3, C3C4, F3Cz uses the same naming convention as FzPOz, described above) PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4, F3Cz) for Theta-Total Bandwidth (not relative PSD) summed from Hz bins 3-7 PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4, F3Cz) for Alpha Bandwidth (not relative PSD) summed from Hz bins 8-13 PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4, F3Cz) for Beta Bandwidth (not relative PSD) summed from Hz bins 13-29 PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4, F3Cz) for Beta Bandwidth (not relative PSD) summed from Hz bins 30-40

www.biopac.com

B-Alert Software Manual

2. Diff_Raw.csv

PSDs (1-40Hz) for the differential channels (FzPOz, CzPO, FzC3, C3C4, F3Cz) are computed for generating ABM's classifications for each second of a given .ebs file. PSDs in this file are computed for each second of a given session without the Kaiser Windowing procedure. Relative power values are derived by subtracting the logged power of the individual Hz bin from the summed logged power for the EEG band (1-40Hz) for that channel.

Column Column Name Description

SessionNum

Session Name (.ebs File Name)

Elapsed Time

Total Elapsed Time for session in seconds (hh:mm:ss:ms)

Clock Time

Local Computer Time or ESU TimeStamp (if configured)

(hh:mm:ss:ms)

fzpoz_1

PSD power at channel FzPoz (without Kaiser windowing) for the 1 Hz bin

fzpoz_2

PSD power at channel FzPoz (without Kaiser windowing) for the 2 Hz bin

F - AQ fzpoz_3-40

PSD power at channel FzPoz (without Kaiser windowing) for the 3-40 Hz bins

fzpoz_rel1

Relative PSD power at channel FzPoz (without Kaiser windowing) for 1 Hz Bin

AS - CE fzpoz_rel2-rel40 Relative PSD power at channel FzPoz (without Kaiser windowing) for 2-40 Hz Bin

fzpoz_Delta_1_2 PSD for Delta Bandwidth at channel FzPoz (not relative PSD)

summed from Hz bins 1-2 (without Kaiser windowing)

fzpoz_ThetaSlow PSD for Theta-Slow Bandwidth (not relative PSD) summed from

_3_5

Hz bins 3-5 (without Kaiser windowing)

fzpoz_ThetaFast_ PSD for Theta-Fast Bandwidth (not relative PSD) summed from Hz

5_7

bins 5-7 (without Kaiser windowing)

fzpoz_ThetaTotal PSD for Theta-Total Bandwidth (not relative PSD) summed from

_3_7

Hz bins 3-7(without Kaiser windowing)

fzpoz_AlphaSlow PSD for Alpha-Slow Bandwidth (not relative PSD) summed from

_8_10

Hz bins 8-10 (without Kaiser windowing)

fzpoz_AlphaFast_ PSD for Alpha-Fast Bandwidth (not relative PSD) summed from Hz

10_13

bins 10-13 (without Kaiser windowing)

fzpoz_AlphaTotal PSD for Alpha-Fast Bandwidth (not relative PSD) summed from Hz

_8_13

bins 8-13 (without Kaiser windowing)

fzpoz_Gamma_2 PSD for Gamma Bandwidth (not relative PSD) summed from Hz

5_40

bins 25-40 (without Kaiser windowing)

fzpoz_Beta_13_2 PSD for Beta Bandwidth (not relative PSD) summed from Hz bins

13-29 (without Kaiser windowing)

czpoz_1

PSD power at channel CzPOz (with Kaiser windowing) for the 1 Hz bin (without Kaiser windowing)

CP - QF

PSD/Relative PSD information for all differential channels (CzPOz, FzC3, C3C4, F3Cz uses the same naming convention as FzPOz (described above)

ThetaOverall_3_ PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4,

F3Cz) for Theta-Total Bandwidth (not relative PSD) summed from

Hz bins 3-7 (without Kaiser windowing)

AlphaOverall_8_ PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4,

F3Cz) for Alpha Bandwidth (not relative PSD) summed from Hz

bins 8-13 (without Kaiser windowing)

www.biopac.com

B-Alert Software Manual

GammaOverall_2 PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4,

5_40

F3Cz) for Beta Bandwidth (not relative PSD) summed from Hz bins

25-40 (without Kaiser windowing)

BetaOverall_13_ PSD Across ALL 5 differential channels (FzPoz, CzPoz, FzC3, C3C4,

F3Cz) for Beta Bandwidth (not relative PSD) summed from Hz bins

13-29 (without Kaiser windowing)

3. Ref_Class.csv

PSDs (1-40Hz) for the referential channels are computed for generating ABM's classifications for each second of a given .ebs file. PSDs in this file are computed for each second of a given session with the Kaiser Windowing procedure described above. Relative power values (_rel) are derived by subtracting the logged power of the individual Hz bin from the summed logged power for the EEG band (1-40Hz) for that channel.

Column Column Name

Description

SessionNum

Elapsed Time

Session Name (.ebs File Name) Total Elapsed Time for session in seconds (hh:mm:ss:ms)

C D E F - AQ AR AS - CE CF CG CH CI CJ

Clock Time POz_1 POz_2 POz_3-40 POz_rel1 POz_rel2-rel40 POz_Delta_1_2 POz _ThetaSlow_3_5 POz _ThetaFast_5_7 POz _ThetaTotal_3_7 POz _AlphaSlow_8_10

Local Computer Time or ESU TimeStamp (if configured) (hh:mm:ss:ms) PSD power at channel POz (with Kaiser windowing) for the 1 Hz bin PSD power at channel POz (with Kaiser windowing) for the 2 Hz bin PSD power at channel POz (with Kaiser windowing) for the 3-40 Hz bins Relative PSD power at channel POz (with Kaiser windowing) for 1 Hz Bin Relative PSD power at channel POz (with Kaiser windowing) for 2-40 Hz Bin PSD for Delta Bandwidth at channel POz (not relative PSD) summed from Hz bins 1-2 (with Kaiser windowing) PSD for Theta-Slow Bandwidth at channel POz (not relative PSD) summed from Hz bins 3-5 (with Kaiser windowing) PSD for Theta-Fast Bandwidth at channel POz (not relative PSD) summed from Hz bins 5-7 (with Kaiser windowing) PSD for Theta-Total Bandwidth at channel POz (not relative PSD) summed from Hz bins 3-7 (with Kaiser windowing) PSD for Alpha-Slow Bandwidth at channel POz (not relative PSD) summed from Hz bins 8-10 (with Kaiser windowing)

POz

PSD for Alpha-Fast Bandwidth at channel POz (not relative PSD)

_AlphaFast_10_13 summed from Hz bins 10-13 (with Kaiser windowing)

POz

PSD for Alpha-Fast Bandwidth at channel POz (not relative PSD)

_AlphaTotal_8_13

summed from Hz bins 8-13 (with Kaiser windowing)

POz _Gamma_25_40 PSD for Gamma Bandwidth at channel POz (not relative PSD)

summed from Hz bins 25-40 (with Kaiser windowing)

POz _Beta_13_29

PSD for Beta Bandwidth at channel POz (not relative PSD)

summed from Hz bins 13-29 (with Kaiser windowing)

Fz_1

CP ADX

PSD power at channel Fz (with Kaiser windowing) for the 1 Hz bin (with Kaiser windowing)
PSD information for all referential channels (POz, Fz, Cz, C3, C4, F3, F4, P3, P4); uses the same naming convention and analysis as Fz (described above-- all with Kaiser windowing)

www.biopac.com

B-Alert Software Manual

ADY ThetaOverall_3_7

Mean PSD across ALL 9 referential channels (POz, Fz, Cz, C3, C4,

F3, F4, P3, P4) for Theta-Total Bandwidth (not relative PSD)

summed from Hz bins 3-7 (with Kaiser windowing)

ADZ AlphaOverall_8_13 Mean PSD across ALL 9 referential channels (POz, Fz, Cz, C3, C4,

F3, F4, P3, P4) for Alpha Bandwidth (not relative PSD) summed

from Hz bins 8-13 (with Kaiser windowing)

AEA GammaOverall_25_40 Mean PSD across ALL 9 referential channels (POz, Fz, Cz, C3, C4,

F3, F4, P3, P4) for Gamma Bandwidth (not relative PSD)

summed from Hz bins 25-40 (with Kaiser windowing)

AEB BetaOverall_13_29 Mean PSD across ALL 9 referential channels (POz, Fz, Cz, C3, C4,

F3, F4, P3, P4) for Beta Bandwidth (not relative PSD) summed

from Hz bins 13-29 (with Kaiser windowing)

4. Ref_Raw.csv

PSDs (1-40Hz) for the referential channels are computed for generating ABM's classifications for each second of a given .ebs file. PSDs in this file are computed for each second of a given session without the Kaiser windowing procedure. Relative power values (_rel) are derived by subtracting the logged power of the individual Hz bin from the summed logged power for the EEG band (1-40 Hz) for that channel.

Column Column Name

Description

SessionNum

Session Name (.ebs File Name)

Elapsed Time

Total Elapsed Time for session in seconds (hh:mm:ss:ms)

Clock Time

Local Computer Time or ESU TimeStamp (if configured) (hh:mm:ss:ms)

POz_1

PSD power at channel POz (without Kaiser windowing) for the 1 Hz bin

POz_2

PSD power at channel POz (without Kaiser windowing) for the 2 Hz bin

F - AQ POz_3-40

PSD power at channel POz (without Kaiser windowing) for the 3-40 Hz bins

POz_rel1

Relative PSD power at channel POz (without Kaiser windowing) for 1 Hz Bin

AS - CE POz_rel2-rel40

Relative PSD power at channel POz (with Kaiser windowing) for 2-40 Hz Bin

POz_Delta_1_2

PSD for Delta Bandwidth at channel POz (not relative PSD) summed from Hz bins 1-2 (without Kaiser windowing)

POz _ThetaSlow_3_5 PSD for Theta-Slow Bandwidth at channel POz (not relative

PSD) summed from Hz bins 3-5 (without Kaiser windowing)

POz _ThetaFast_5_7 PSD for Theta-Fast Bandwidth at channel POz (not relative

PSD) summed from Hz bins 5-7 (without Kaiser windowing)

POz _ThetaTotal_3_7 PSD for Theta-Total Bandwidth at channel POz (not relative

PSD) summed from Hz bins 3-7(without Kaiser windowing)

POz _AlphaSlow_8_10 PSD for Alpha-Slow Bandwidth at channel POz (not relative

PSD) summed from Hz bins 8-10 (without Kaiser windowing)

POz

PSD for Alpha-Fast Bandwidth at channel POz (not relative

_AlphaFast_10_13

PSD) summed from Hz bins 10-13 (without Kaiser windowing)

POz

PSD for Alpha-Fast Bandwidth at channel POz (not relative

_AlphaTotal_8_13

PSD) summed from Hz bins 8-13 (without Kaiser windowing)

POz _Gamma_25_40 PSD for Gamma Bandwidth at channel POz (not relative PSD)

summed from Hz bins 25-40 (without Kaiser windowing)

POz _Beta_13_29

PSD for Beta Bandwidth at channel POz (not relative PSD)

www.biopac.com

B-Alert Software Manual

summed from Hz bins 13-29 (without Kaiser windowing)

Fz_1

PSD power at channel Fz (without Kaiser windowing) for the 1

Hz bin

CP -

PSD/Relative PSD information for all referential channels

ADX

(POz, Fz, Cz, C3, C4, F3, F4, P3, P4) uses same naming

convention and analysis as Fz (described above -- all without

Kaiser windowing)

ADY ThetaOverall_3_7

Mean PSD across ALL referential channels (POz, Fz, Cz, C3, C4,

F3, F4, P3, P4) for Theta-Total Bandwidth (not relative PSD)

summed from Hz bins 3-7 (without Kaiser windowing)

ADZ AlphaOverall_8_13 Mean PSD across ALL referential channels (POz, Fz, Cz, C3, C4,

F3, F4, P3, P4) for Alpha Bandwidth (not relative PSD)

summed from Hz bins 8-13 (without Kaiser windowing)

AEA GammaOverall_25_40 Mean PSD across ALL 9 referential channels (POz, Fz, Cz, C3,

C4, F3, F4, P3, P4) for Gamma Bandwidth (not relative PSD)

summed from Hz bins 25-40 (without Kaiser windowing)

AEB BetaOverall_13_29 Mean PSD across ALL 9 referential channels (POz, Fz, Cz, C3,

C4, F3, F4, P3, P4) for Beta Bandwidth (not relative PSD)

summed from Hz bins 13-29 (without Kaiser windowing)

A.4 Heart Rate Outputs
_HR_beat.csv
_HR_epoch.csv
Signal processing performs 5 operations on the raw ECG signal, the output being a signal with QRS complexes enhanced, and other components attenuated. The r to r interval is calculated to determine each heart rate. Heart rate variability will be determined based on detected heart rate values. ABM's Heart Rate (HR) algorithm computes beat to beat (_HR_beat.csv) and second by second (_HR_Epoch.csv).

1. _HR_beat.csv

This file contains the beat to beat heart rate data based on ABM's Heart Rate algorithm.

Column Column Name

Description

SessionNum

Session Name (.ebs File Name)

Elapsed Time

Elapsed time (hh:mm:ss:ms)

Clock Time

Local Computer Time or ESU TimeStamp (if configured)

(hh:mm:ss:ms)

Beat Quality

1 or 0 value:

0 = beat quality good,

1 = beat quality poor based on artifact in ECG channel.

Inter-Beat Interval

Inter-beat Interval.

Heart Rate

Beat to Beat Heart Rate (beats per minute).

2. _HR_epoch.csv

This file contains the second by second (epoch by epoch) heart rate based on B-Alert Live's Heart Rate algorithm.

Column Column Name

Description

SessionNum

Elapsed Time

Session Name (.ebs File Name) Elapsed time for the detected beat (hh:mm:ss:ms)

Clock Time

Local Computer Time or ESU TimeStamp (if configured) (hh:mm:ss:ms)

www.biopac.com

B-Alert Software Manual

Beat Quality

1 or 0 value:

0 = beat quality good,

1 = beat quality poor based on artifact in ECG channel.

Inter-Beat Interval

Inter-beat Interval: Interpreted to the second

Heart Rate

Second to Second Heart Rate (beats per minute)

A.5 Z-Score Outputs
To remove individual variability from the various metrics, a z-score output is also provided. The Z-score for this purpose is calculated on the mean and standard deviation (SD) for at least the first 5 seconds, with additional epochs (seconds) added. The initial period to begin calculation is defaulted to 5 sec, but can be set by the user. A new mean and SD are computed if the raw data differs more than 2.5 SD from the mean. Epochs that are invalid will not be included in Z-score calculation. Instead the mean, SD, and z-score from a previous valid second are used.
After acquiring the configured period of valid seconds to determine the mean and SD, the ZScore algorithms begins computing the z-score for each epoch for input signal by:

ZScore = (Epoch value - mean)/ standard deviation

1. ZScore.csv

The ZScore.csv output file contains classification, workload, and heart rate information.

Column Column Name

Description

SessionNum

Session Name (.ebs File Name)

Elapsed Time

Elapsed time for the detected beat (hh:mm:ss:ms)

Clock Time

Local Computer Time or ESU TimeStamp (if configured)

(hh:mm:ss:ms)

class_higheng

Z-Score of High Engagement Probability (-99999 if it falls within Z-

score computation window)

class_loweng

Z-Score of Low Engagement Probability (-99999 if it falls within Z-

score computation window)

class_distraction Z-Score of Class Engagement Probability (-99999 if it falls within

Z-score computation window)

class_drowsy

Z-Score of Class Drowsy Probability (-99999 if it falls within Z-

score computation window)

wl_fbds

Z-Score of Workload probability (FBDS model), where higher

probability reflects higher WL (FBDS is best model for 85% of

population)

(-99999 if it falls within Z-score computation window)

wl_bds

Z-Score of Alternate Workload probability (BDS model): Not

recommended for use, higher probability reflects higher WL (BDS

Best WL model for other 15% of model)

(-99999 if it falls within Z-score computation window)

wl_ave

Z-score of mean workload probability (mean of BDS and FBDS

models), higher probability reflects higher workload (-99999 if it

falls within Z-score computation window)

HeartRate

Z-Score of the second by second Heart Rate (-99999 if during Z

score computation window)

2. _ZScore_PSD.csv
Z-scored info (same Z-scoring procedure as used for _ZScore output) for Heart Rate, and PSD info for 9 referential channels (POz, Fz, Cz, C3, C4, F3, F4, P3, P4) and 5 differential channels (FzPOz, CzPOz, FzC3, C3C4, F3Cz). The 5 differential channels available are fixed, based on the channels required for ABM's classification models.

www.biopac.com

B-Alert Software Manual

Column A B C

Column Name SessionNum Elapsed Time Clock Time

HeartRate

OvRawRef_0

OvRawRef_1

poz_3

poz_6

poz_7

poz_8

K-BJ

Description
Session Name (.ebs File Name) Elapsed time for the detected beat (hh:mm:ss:ms) Local Computer Time or ESU TimeStamp (if configured) (hh:mm:ss:ms) Z-Score of Heart Rate (-99999 if it falls within Z-score computation window) Interpreted to the second Z-score PSD (No Kaiser Window) for Theta Total (3-7Hz bins) Z-score PSD of (No Kaiser Window) for Alpha Total (8-13Hz bins) Z-scored PSD (No Kaiser Window) for Theta Total (3-7 Hz bin) at Channel POz Z-scored PSD (No Kaiser Window) for Alpha Total (8-13 Hz bin) at Channel POz Z-scored PSD (No Kaiser Window) for Gamma (25-40 Hz bin at Channel POz Z-scored PSD (No Kaiser Window) for Beta (13-30 Hz bin) at Channel POz Z-Scored PSD for all 9 Referential channels (POz, Fz, Cz, C3, C4, F3, F4, P3, P4) and 5 Differential channels using same naming conventions described above (FzPOz, CzPOz, FzC3, C3C4, F3Cz)

A.6 Actigraphy Outputs

1. Actigraphy.csv

The Actigraphy.csv output file contains both raw tilt data as well as the derived/processed data from the accelerometer.

Column Column Name

Description

SessionNum

Elapsed Time

Clock Time

X_Raw

X_Angle

Session Name (.ebs File Name) Elapsed time for the detected beat (hh:mm:ss:ms) Local Computer Time or ESU TimeStamp (if configured) (hh:mm:ss:ms) Raw x-axis tilt value X-axis Angle derived from tilt value

Y_Raw

Raw Y-axis tilt value

Y_Angle

Y-axis Angle derived from tilt value

Z_Raw

Raw Z-axis tilt value

Z_Angle

Z-axis Angle derived from tilt value

Movement_value Sum of change in two dominant angles

Movement_Scale

Value between 0-5 derived from the change in two dominant

angle using proprietary algorithm

A.7 Signal Quality Outputs

1. Artifact.csv

Artifacts (i.e., EMG, Eye blinks, Saturation, Spike, Excursion) detected in the signal.

Column Column Name

Description

StartEpoch

StartDP

EndEpoch

Start Epoch of the detected artifact Start Datapoint of the detected artifact End Epoch of the detected artifact

www.biopac.com

B-Alert Software Manual

End DP

End Datapoint of the detected artifact

Artifact Type

Artifact type (0 - Spike, 1 - Excursion, 2 - Saturation, 3 - EMG, 5 -

Eye blink)

Channel

Channel in which the artifact was detected (Eye blinks are reported in all channels, hence channel field is empty)

Rule

Artifact rule (reserved for internal debugging)

2. _missed_blocks.csv

This output file can be used to determine whether BT packets (or blocks) were dropped or missed during a data collection.

Column A B

Column Name sessionID Blocks

Start Counter

End Counter

Epoch Start

Offset Start

Epoch End

OffsetEnd

Hour

Minute

Second

Millisecond

Description .ebs file Name Blocks: this increments for each blocks of lost packets, for example if packets 5,6,7 were missed and then 25,26,27,28 were missed, There will be two row entries in the csv file. Blocks will be 1 for former and 2 for late Count when missed block started (6-bit counter in the headset (debugging) Count when missed block stopped (6-bit counter in the headset (debugging) Starting second (Epoch) of missed block Starting Sample (datapoint or Offest) of missed block (in 1/256sec) Ending second (Epoch) of missed block Ending Sample (datapoint or Offest) of missed block (in 1/256sec)
Time of storing to file (or time of first packet AFTER missed blocks)