Summary

Revised and Neuroimaging-Compatible Versions of the Dual Task Screen

Published: October 05, 2020
doi:

Summary

We developed the original Dual Task Screen (DTS) as a portable, low-cost measure that can evaluate athletes with sports-induced mild traumatic brain injury. We revised the original DTS for future clinical use and developed a neuroimaging-compatible version of the DTS to measure neural underpinnings of single and dual task performance.

Abstract

Dual task paradigms simultaneously assess motor and cognitive abilities, and they can detect subtle, residual impairments in athletes with recent mild traumatic brain injury (mTBI). However, past dual task paradigms have focused solely on lower extremity skills and have relied on cumbersome, expensive laboratory equipment – thus limiting their practicality for everyday mTBI evaluation. Subsequently, we developed the Dual Task Screen (DTS), which takes <10 minutes to administer and score, uses low-cost portable equipment, and includes lower extremity (LE) and upper extremity (UE) subtasks. The purpose of this manuscript was twofold. First, we describe the administration protocol for the revised DTS, which we revised to address the limitations of the original DTS. Specifically, the revisions included additions of smart devices to acquire more detailed gait data and inclusion of single cognitive conditions to test for disrupted cognitive performance under dual task conditions. Importantly, the revised DTS is a measure intended for future clinical use, and we present representative results from three male athletes to illustrate the type of clinical data that can be acquired from the measure. Importantly, we have yet to evaluate the sensitivity and specificity of the revised DTS in athletes with mTBI, which is the next research initiative. The second purpose of this manuscript is to describe a neuroimaging-compatible version of the DTS. We developed this version so we could evaluate the neural underpinnings of single and dual task performance, for a better empirical understanding of the behavioral deficits associated with mTBI. Thus, this manuscript also describes the steps we took to enable simultaneous functional near-infrared spectroscopy (fNIRS) measurement during the DTS, along with how we acquired and completed first-level processing of the fNIRS data.

Introduction

Each year, 42 million people worldwide sustain mild traumatic brain injuries (mTBIs)1. Although once considered benign, new research indicates that mTBIs, particularly repeat mTBIs, can elicit lasting negative consequences, such as physical, cognitive, and sleep disturbances2,3,4. Subsequently, researchers and clinicians are seeking enhanced evaluations and treatment methods to understand and address mTBI.

To date, best practice for mTBI evaluation includes self-reported symptoms and objective measurement of neurocognitive and motor function5. However, some individuals, like competitive collegiate-level athletes, are known to underreport mTBI-related symptoms6, limiting the utility of symptom reports. Objective neurocognitive and motor function measures also have limitations, including poor test-retest reliability, reliance on baseline testing, or insufficient difficulty for high-performing athletes7,8,9. However, dual task paradigms – which simultaneously assess motor and cognitive abilities – can detect subtle, residual impairments and may be particularly useful for evaluating high-performing athletes10,11,12,13,14.

Past research using dual task paradigms have often incorporated cumbersome, expensive laboratory equipment, such as motion capture systems14, to evaluate high-performing athletes. While these systems can accurately measure subtle motor impairments, they are impractical for use in everyday mTBI evaluation due to high equipment cost, limited portability, and long administration times (i.e., ≥ 45 minutes per individual). Further, many past dual task paradigm studies focused solely on lower body or lower extremity skills, such as balance or gait11,12,13,14. Arguably, upper extremity function and hand-eye coordination is also important for high-performing athletes in many sports. Thus, we developed the Dual Task Screen (DTS), which is a brief measure designed to be administered and scored in <10 minutes with portable, low-cost instruments. This original DTS included a lower extremity (LE) and upper extremity (UE) subtask, which evaluated gait speed (using a stopwatch) and hand-eye coordination under single motor and dual task conditions15.

In the first feasibility study, 32 healthy, female adolescent participants completed the original DTS. This study was designed to establish that the DTS could elicit dual task motor costs, as indicated by reduced motor performance during dual task vs. single motor conditions. We also sought to establish that the DTS could be administered and scored in fewer than 10 minutes. We found that all participants had poorer dual task motor performance on at least one subtask. Additionally, we were able to administer the DTS in an average of 5.63 minutes and score the test in 2-3 minutes15.

Although the first feasibility study was successful, a few limitations were revealed. Most notably, gait speed was measured with stopwatches, which are prone to natural human error. Therefore, in the revised DTS we used smart devices with built-in accelerometers (Table of Materials) on each ankle. This addition maintained the use of portable, low-cost instruments while still providing sophisticated measures of gait speed, total number of steps, average step length, average step duration, and step duration variability. Another limitation of the original DTS was the absence of single cognitive conditions, which prevented evaluation of dual task cognitive costs. Dual task cognitive costs are defined as poorer cognitive performance during the dual task vs. single cognitive condition. Subsequently, for both the LE and UE subtasks, we added a single cognitive condition (described in Protocol).

In addition to developing a measure for future clinical use, one of the team’s long-term goals is to evaluate the neural underpinnings of single and dual task performance in healthy athletes and contrast those findings to athletes with sports-induced mTBI. Thus, we have created a neuroimaging-compatible version of the DTS. We seek to determine if the DTS can be successfully modified for use with simultaneous functional near-infrared spectroscopy (fNIRS) measurement, and we are using a mobile fNIRS device specifically designed to accommodate gross-motor movement by reducing the influence of motion artifacts. Further, this device has the greatest amount of head coverage, to our knowledge, for mobile devices that are currently available for research purposes (Table of Materials).

In summary, the study protocol is designed to do the following:

  1. Describe the administration protocol for the revised Dual Task Screen (DTS), which is a measure we redesigned to address the limitations of the original DTS15 and a measure intended for future clinical use.
  2. Describe the research protocol for the neuroimaging-compatible Dual Task Screen (DTS), which we have designed to evaluate the neural underpinnings of single and dual task performance.

Protocol

All study procedures were approved by the Institutional Review Board (IRB) at Colorado State University, and all adult participants provided written informed consent prior to completing any study procedures. Written informed consent was provided by parents of participants under 18, and minor participants also provided written assent prior to completing any study procedures.

1. Revised Dual Task Screen (DTS)

  1. Lower Extremity (LE) Subtask
    1. Start the single motor condition.
      1. Place three yoga blocks in a horizontal position exactly 4.5 m apart along an 18 m walkway.
      2. Firmly attach smart devices to each ankle to detect heel strikes and obtain gait characteristics.
      3. Begin video recording with a camcorder on a tripod.
      4. Instruct participants to walk as quickly as possible while stepping over obstacles. Start data collection on the smartphones and sharply tap the devices simultaneously for subsequent time alignment of the two separate data streams from the left and right legs. 
      5. Measure the time-to-complete with a hand-operated stopwatch.
      6. Stop video recording.
    2. Start the single cognitive condition.
      1. Tell the participant his/her allocated time for this condition, using the time-to-complete from his/her single motor condition (rounding up to a full second).
      2. Begin video recording with a camcorder on a tripod.
      3. Instruct participants to state as many words as they can that begin with a particular letter (A or F).
        NOTE: Letters are counter-balanced between participants and between the single and dual task conditions. Numbers are counter-balanced between participants and between the single and dual task conditions.
      4. Stop video recording.
    3. Start the dual task condition.
      1. Begin video recording with a camcorder on a tripod.
      2. Instruct participants to walk as quickly as possible while stepping over obstacles while simultaneously stating as many words as they can that begin with a particular letter (A or F). Rapidly tap both accelerometers to start the condition.
      3. Measure the time-to-complete with a hand-operated stopwatch.
      4. Stop video recording.
  2. Upper Extremity (UE) subtask
    1. Start the single motor condition.
      1. Measure a distance of 1.5 m away from a wall, mark with masking tape, and instruct the participant to stand behind the tape.
      2. Place a basket of tennis balls next to the participant.
      3. Begin video recording with a camcorder on a tripod.
      4. Instruct the participant to complete a wall-toss with alternating hands for 30 s. Tell the participant that if he/she fails to catch a ball, to acquire a new ball from the basket of tennis balls. Measure time elapsed with a stopwatch.
      5. Stop video recording.
    2. Start the single cognitive condition.
      1. Begin video recording with a camcorder on a tripod.
      2. Tell the participant that he/she will be asked to sequentially subtract by 7 from a given number (100 or 150) for 30 seconds. Measure time elapsed with a stopwatch.
      3. Stop video recording.
        NOTE: Letters are counter-balanced between participants and between the single and dual task conditions. Numbers are counter-balanced between participants and between the single and dual task conditions.
    3. Start the dual task condition.
      1. Ask the participant to stand 1.5 m away from a wall.
      2. Place a basket of tennis balls next to participant.
      3. Begin video recording with a camcorder on a tripod.
      4. Instruct the participant to complete a wall-toss with alternating hands for 30 seconds. Inform the participant that, while throwing and catching the balls, he/she will be asked to sequentially subtract by 7 from a given number (100 or 150) for 30 seconds. Tell the participant that if he/she fails to catch a ball, to acquire a new ball from the basket of tennis balls. Measure time elapsed with a stopwatch.
      5. Stop video recording.
        NOTE: Letters are counter-balanced between participants and between the single and dual task conditions. Numbers are counter-balanced between participants and between the single and dual task conditions.

2. Neuroimaging-Compatible Dual Task Screen (DTS)

  1. Set up the DTS
    1. Place yoga blocks in a vertical position to mark the start and end of a 15 m walkway.
    2. Place two yoga blocks in a horizontal position exactly 5 m apart along the 15 m walkway.
    3. Measure and mark with masking tape a distance 1.5 m away from a smooth wall surface.
    4. Set up a tripod at the beginning of the 15 m walkway.
  2. Place the fNIRS device on participant’s head.
    1. Measure the participant’s head circumference and select appropriately sized fNIRS cap (Table of Materials) with pre-placed optodes and short-channel detectors.
    2. Turn on a dedicated acquisition laptop and connect to the fNIRS device’s WiFi network.
    3. Open the fNIRS acquisition software and select the fNIRS device.
    4. Perform calibration to optimize light intensity and check optode signal levels. Signal levels should be acceptable or excellent.
    5. Fix all optodes with less than acceptable signal level by removing the optode from the cap and parting the participant’s hair to ensure a direct connection of the optode to the participant’s scalp.
  3. Place accelerometers on the participant’s ankles.
    1. Firmly attach smart devices to each ankle to detect heel strikes and obtain gait characteristics.
  4. Begin LE subtask data acquisition.
    1. Open the stimulus presentation software (Table of Materials).
    2. Select the LE subtask file.
    3. Ask the participant to sit in a chair in preparation for a 60 s quiet rest period.
    4. Return to the fNIRS acquisition software and click the Start button to begin collecting fNIRS data. Enter subject ID_LE, age, and sex into the pop-up window and click Start.
    5. Return to the stimulus presentation software, inform the participant that quiet rest will begin, and press Space to start the 60 s rest period.
    6. At the end of the rest period, identify which LE Subtask condition (single motor, single cognitive, or dual task) has been selected for the 1st trial. Provide participant with instructions for that trial.
      1. Single Motor Instructions: Instruct the participant to walk as quickly as possible, while stepping over the obstacles, for 30 s. Tell participant that he/she will start when the primary researcher says “Start”. This will occur immediately after a secondary researcher taps the accelerometers. Instruct the participant that he/she should stop walking when the primary researcher says “stop.” Additionally, when the primary researcher says “stop,” the participant should put his/her feet together and remain as still as possible. At this time, the secondary researcher will tap the accelerometers a second time and place a marker (sticky note) on the floor where the participant stopped.
      2. Single Cognitive Instructions: Instruct the participant to remain standing at the start of the 15 m walkway. While standing, he/she will be asked to state as many words as possible that begin with a particular letter.
      3. Dual Task Instructions: Instruct the participant to walk as quickly as possible while stepping over the obstacles and simultaneously stating as many words as possible beginning with a particular letter. Inform him/her that he/she will also have 30 seconds for this condition. Tell participant that he/she will start when the primary researcher says “start”. This will occur immediately after a secondary researcher taps the accelerometers. Instruct the participant that he/she should stop walking when the primary researcher says “stop.” Additionally, when the primary researcher says “stop,” the participant should put his/her feet together and remain as still as possible. At this time, the secondary researcher will tap the accelerometers a second time and place a marker (sticky note) on the floor where the participant stopped.
    7. Begin video recording with a camcorder on a tripod.
    8. Press the Space bar to start the 1st trial. Monitor the 30 s timer on the stimulus presentation software; tell the participant to stop when 30 s have elapsed.
    9. Identify the 2nd trial and provide the participant with instructions. Repeat the process until the participant has completed 15 randomized trials of the LE Subtask.
    10. Stop video recording.
    11. Inform the participant that he/she will complete another 60 s seated rest period. Once the participant is seated, press Start to begin the rest period.
    12. Following the rest period, exit out of the LE subtask file in the stimulus presentation software. Stop data acquisition in fNIRS data acquisition software, but do not exit the software.
      NOTE: Letters are randomized (by the stimulus presentation software) between trials and counter-balanced between participants and between single and dual task conditions. Letters are similar in level of difficulty and include: W, D, F, T, S, H, M, A, B, and P. Numbers are randomized (by the stimulus presentation software) between trials and counter-balanced between participants and between single and dual task conditions. Numbers included: 185, 225, 220, 175, 205, 165, 170, 180, 245, and 240.
  5. Remove accelerometers from the participant’s ankles. Move to the section in the hallway designated for the UE Subtask.
  6. Begin UE subtask data acquisition.
    1. Open the stimulus presentation software.
    2. Select the UE subtask file.
    3. Ask the participant to sit in a chair in preparation for a 60 s quiet rest period.
    4. Return to the fNIRS acquisition software and click the Start button to begin collecting fNIRS data. Enter subject ID_UE, age, and sex into the pop-up window and click start.
    5. Return to the stimulus presentation software, inform the participant that the quiet rest period is about to begin, and press Space to start 60 s rest period.
    6. At the end of the rest period, identify which UE Subtask condition (single motor, single cognitive, or dual task) has been selected for the 1st trial. Provide the participant with instructions for that trial.
      1. Single Motor Instructions: Instruct the participant to stand 1.5 m away from a wall. Place a basket of tennis balls next to participant. Instruct the participant to complete a wall-toss with alternating hands for 30 s. Tell the participant that if he/she fails to catch a ball, to acquire a new ball from the basket of tennis balls.
      2. Single Cognitive Instructions: Instruct the participant to remain standing Tell the participant that he/she will be asked to sequentially subtract by 7 from a given number for 30 s.
      3. Dual Task Instructions: Instruct the participant to complete a wall-toss with alternating hands for 30 s. Inform the participant that, while throwing and catching the balls, he/she will be asked to sequentially subtract by 7 from a given number2 for 30 s. Tell the participant that if he/she fails to catch a ball, to acquire a new ball from the basket of tennis balls.
    7. Begin video recording with a camcorder on a tripod.
    8. Press the Space bar to start the 1st trial. Monitor the 30 s timer on the stimulus presentation software; tell participant to stop when 30 s have elapsed.
    9. Identify the 2nd trial and provide the participant with instructions. Repeat the process until participant has completed 15 randomized trials of the UE Subtask.
    10. Stop video recording.
    11. Inform the participant that he/she will complete another 60 s seated rest period. Once the participant is seated, press Start to begin the rest period.
    12. Following the rest period, exit out of the UE Subtask file in the stimulus presentation software. Stop the data acquisition in fNIRS data acquisition software, and then exit the software.
  7. Remove fNIRS cap from participant’s head.
    NOTE: Letters are randomized (by the stimulus presentation software) between trials and counter-balanced between participants and between single and dual task conditions. Letters are similar in level of difficulty and include: W, D, F, T, S, H, M, A, B, and P. Numbers are randomized (by the stimulus presentation software) between trials and counter-balanced between participants and between single and dual task conditions. Numbers included: 185, 225, 220, 175, 205, 165, 170, 180, 245, and 240.

Representative Results

Participants
Participants were recruited from local high school teams and university intercollegiate and club sports teams using word-of-mouth and advertising flyers. Participants were required to be between the ages of 15-22 and engage in regular participation in organized contact sports. Contact sports included all sports where physical contact with teammates or opponents is necessary during routine play. Participants also had to have normal or corrected vision and hearing, no history of neurological or psychiatric conditions, and no history of moderate or severe traumatic brain injury, per self-report.

We included data from three healthy male contact-sport athlete participants (Mean Age: 18.0 ± 2.65 yrs.) to illustrate the type of clinical data that can be acquired from the revised DTS. Data from healthy, female contact-sport athletes will be included in another publication that is not strictly Methods-focused.

Data Analysis for Revised DTS
Given the small number of participants included in the representative results, formal statistical analyses were not completed. However, for each participant, performance in the dual task condition was compared to performance in the single motor and single cognitive conditions; see below for the description of the performance metrics on both subtasks.

Performance Metrics on the LE Subtask
Single motor condition performance was quantified by gait speed (m/s), total number of steps, average step length (m), average step duration (s), and variability of step duration (SD). These data were acquired with the built-in accelerometers on the smart devices we affixed to participants’ ankles. Single cognitive condition performance was measured by the total number of words produced without repetitions, represented as words/s to account for the varied amount of time allotted for this trial. Two trained research assistants watched a video tape of the single cognitive condition and were required to reach a consensus on the total number of words produced. Finally, dual task condition performance was measured by gait speed (m/s), total number of steps, average step length (m), average step duration (s), and average step duration variability (SD), and the total number of words produced without repetitions, represented as words/second. Two trained research assistants also watched a video tape of the dual task condition and were required to reach a consensus on the total number of words produced.

Dual Task Costs on the LE Subtask
For each participant, a dual task motor cost would be represented by the following changes in gait characteristics during the dual task condition compared to the single motor condition: slower gait speed, a greater number of total steps, a smaller average step length, a longer average step duration, and a greater step duration variability. We observed that all three male participants had a dual task motor cost on the LE Subtask. Specifically, we saw slower gait speed, longer average step duration, and greater variability in step duration during dual, compared to single condition tasks; see Figure 1A. In contrast, two of three participants showed no changes in number of total steps or average step length between single motor and dual task conditions; see Figure 1A.

For each participant, a dual task cognitive cost would be represented by fewer words generated in the dual task condition compared to the number of words generated in the single cognitive task condition. We observed dual task cognitive costs in two of three participants. Specifically, these participants generated fewer words during the dual task condition compared to the single task condition; see Figure 1B.

Performance Metrics on the UE Subtask
Single motor condition performance was measured by the total number of successful catches. Two trained research assistants watched a video tape of the single motor condition and were required to reach a consensus on the total number of successful catches. Single cognitive condition performance was measured by the total number of correct subtractions. Two trained research assistants watched a video tape of the single cognitive condition and were required to reach a consensus on the total number of correct subtractions. Subtraction errors were not cumulative (i.e., “100, 92, 85…” would be recorded as one error and one correct subtraction). Finally, dual task condition performance was measured by the total number of successful catches and total number of correct subtractions. Again, two trained research assistants watched a video tape of the single cognitive condition and were required to reach a consensus on the total number of successful catches and correct subtractions.

Dual Task Cost on the UE Subtask
For each participant, a dual task motor cost would be represented by fewer successful catches during the dual task condition compared to the number of successful catches made during the single motor condition. We found that all three male participants had a dual task motor cost. Specifically, they had fewer successful catches during the dual task condition compared to the single motor condition; see Figure 2A.

A dual task cognitive cost would be represented by fewer correct subtractions the dual task condition compared to the number of correct subtractions made during the single task condition. We observed dual task cognitive costs in two of three participants. Specifically, they had fewer correct subtractions during the dual task condition compared to the single task condition; see Figure 2B.

Data Analysis for Neuroimaging-Compatible DTS
fNIRS Device Specifications
We used a mobile functional near-infrared spectroscopy (fNIRS) system (Table of Materials). The system has 32 total optodes, 16 LED sources and 16 detectors, and a wireless acquisition device that participants wear on their backs. This device is uniquely equipped to accommodate gross motor movement, and has (to our knowledge) the greatest amount of head coverage for a mobile system. Using fNIRS we evaluated brain activity via the hemodynamic response using indices of oxygenated hemoglobin (HbO) during the neuroimaging-compatible DTS.

fNIRS Head Probe
The head probe included 30 optodes (15 LED sources and 15 detectors) that were placed on the participant’s head using an fNIRS cap with built-in optode holders. We measured HbO by placing LED sources and detectors at the left and right motor cortex and two primary regions of the right-lateralized frontoparietal network16, right PFC and PPC, which we have identified with the 10-20 system17; see Figure 3. The LED sources shine near-infrared light into superficial cortical regions, and the detectors capture the refracted light, allowing us to calculate HbO values at each channel, or intersection of source and detector. Additionally, we include eight short separation detectors, which measure scalp perfusion, a nuisance variable that will be regressed out of the raw fNIRS data18,19.

Block Design for fNIRS Acquisition
Both the LE and UE subtasks were converted into a block design. Both subtasks started and ended with a 60 s seated rest period to acquire baseline hemodynamic activity. Rest was followed by 15 randomized blocks (5 single motor condition blocks, 5 single cognitive condition blocks, and 5 dual task condition blocks) that were 30 s in duration, totaling 7.5 minutes of total data collection for each subtask. Between each of the 15 condition blocks, there was a variable resting interval of approximately 6-8 s to allow the participants’ hemodynamic response to return to baseline; see Figure 4.

FNIRS Data Reduction and First-Level (Single-Subject) Analysis: Raw fNIRS data are uploaded into a proprietary programming language and numerical computing environment (Table of Materials). Channels created with short separation detectors are labeled for later regression. Default stimuli values, which were generated by the stimulus presentation software, are renamed to identify DTS blocks (e.g., single motor, single cognitive, dual motor). Next, stimulus duration parameters are set to 30 seconds for all DTS blocks and 60 s for rest periods. Basic processing is then completed using steps from a non-proprietary toolbox that is compatible with the numerical computing environment. These steps include calculating optical density and then recalculating optical density values given data from the short separation channels20. Next, optical density is converted to hemoglobin values (deoxygenated hemoglobin, oxygenated hemoglobin, and total hemoglobin) using the modified Beer Lambert Law21. Following conversion, an autoregressive model algorithm is run, which includes regression of short separation channel data. Parameters for the autoregressive algorithm are set to follow a Canonical model22. Finally, individual data can be visualized using condition contrasts (e.g. Dual vs Single); see Figure 5.

Figure 1
Figure 1: LE Subtask Performance during Single vs. Dual Task Conditions. (A) All three participants had slower gait speed, longer average step duration, and greater variability in step duration during the dual task condition compared to the single task condition, which represents a dual task motor cost on the UE subtask. Two of three participants showed no changes in number of total steps or average step length between dual and single task conditions. (B) Two of three participants generated fewer words during the dual task condition compared to the single task condition, which represents a dual task cognitive cost on the LE subtask. Please click here to view a larger version of this figure.

Figure 2
Figure 2: UE Subtask Performance during Single vs. Dual Task Conditions. (A) All three participants had fewer successful catches during the dual task condition compared to the single task condition, which represents a dual task motor cost on the UE subtask. (B) Two of three participants had fewer correct subtractions during the dual task condition compared to the single task condition, which represents a dual task cognitive cost on the UE subtask. Please click here to view a larger version of this figure.

Figure 3
Figure 3: FNIRS Head Probe. The fNIRS head probe included 15 LED sources (red circles) and 15 detectors (white circles) which were placed at the left and right motor cortex and right prefrontal cortex (PFC) and right posterior parietal cortex (PPC). This allowed us to calculate oxygenated hemoglobin (HbO) values at each channel, or intersection of source and detector. Please click here to view a larger version of this figure.

Figure 4
Figure 4: Block Design for fNIRS Acquisition. For the neuroimaging compatible version of the DTS, the LE and UE subtasks were converted into a block design. Both subtasks started and ended with a 60 second seated rest period to acquire baseline hemodynamic activity. Rest was followed by 15 randomized blocks (5 single motor condition blocks, 5 single cognitive condition blocks, and 5 dual task condition blocks) that were 30 seconds in duration. Please click here to view a larger version of this figure.

Figure 5
Figure 5: Single Subject fNIRS Data. This is a depiction of single subject fNIRS data using condition contrasts. This image contrasts oxygenated hemoglobin (HbO) during the Dual Task vs Single Motor Task from the LE subtask. Please click here to view a larger version of this figure.

Discussion

In this manuscript, we described the administration protocol for the newly revised Dual Task Screen (DTS). These revisions were completed to address limitations identified in the original DTS15 and included the addition of single cognitive conditions to test for dual task cognitive costs. It also included smart-device based accelerometry to more precisely measure gait characteristics. We included representative results which illustrate the type of clinical data that can be acquired with the DTS. We also described the research protocol for the neuroimaging-compatible Dual Task Screen (DTS), which we have designed to evaluate the neural underpinnings of single and dual task performance. The neuroimaging modality we selected was a portable functional near-infrared spectroscopy (fNIRS) device that accommodates gross motor movement by reducing the influence of motion artifacts18,19. To create a neuroimaging-compatible version, we had to convert the DTS to a block design. The block design required five repetitions, or blocks, of the single motor, single cognitive, and dual task conditions. This required use of novel cognitive stimuli (e.g., numbers and letters) of equivalent difficulty for each trial.

The addition of accelerometers was the most challenging addition to the revised DTS, as this required that we mark precisely when the obstacle walk was initiated on both of the smart devices. We simultaneously tapped the smart devices/accelerometers, prior to the participants’ first step, to create an artifact spike in the acceleration data. We also videotaped the participants’ walking, so we could match their heel strikes in the video with the heel strikes recorded by the accelerometers.

The majority of the troubleshooting, however, was completed to create a neuroimaging-compatible version of the DTS. The first obstacle we encountered was finding stimulus presentation software that could wirelessly interface with the neuroimaging acquisition software. Unlike computer-based tasks, the participant did not need to see which condition was about to occur, but the researcher had to see the conditions in order to provide instructions. Further, this stimulation software had to interface seamlessly with the acquisition software, to mark the conditions that were occurring. This is necessary for future segmentation and averaging of the neuroimaging data across all five blocks of each condition. We successfully identified a stimulus presentation software which interfaced with the fNIRS data acquisition software via a lab streaming layer. This allowed us to use both programs simultaneously. The next obstacle we encountered was modifying the DTS to a block-design, where each block was 30 seconds in duration, which is necessary for optimal fNIRS data quality. Additionally, we needed to include rest periods at the beginning and end of each subtask to measure baseline brain perfusion, due to known inter-subject variability in brain perfusion23, particularly after mTBI24. Further, we needed to add 6-10 s transition periods between blocks to allow participants’ brain activity to return to baseline. Finally, we determined that we needed to randomize the block order and counter-balance letter and number stimuli, for the cognitive tasks, to reduce practice effects and avoid neural habituation. The most challenging task to modify to a 30 s block design was the obstacle walk in the LE subtask. Prior to modification, this was an 18 m obstacle walk, and the duration was the time it took for participants to complete it. To change the 18 m walk to a 30-second block, we asked participants to repeat a 15 m walk with two obstacles (instead of three) until time was called. At the end of the 30 s block we placed a temporary marker (sticky notes) on the floor where the participant stopped. This allowed us to precisely measure the walked distance and calculate gait speed in m/s. Finally, in the stimulus presentation software, we added a video of a 30 s timer for each block, so the researcher could visualize the neuroimaging software and the duration of each block simultaneously on a laptop computer and provide verbal cues (e.g. “start” and “stop”) to the participant for the beginning and end of each block.

In the representative results, we found that the following gait characteristics showed dual task motor costs on the LE subtask: gait speed, average step duration, and variability in step duration. In contrast, total steps and average step length did not appear to show dual task motor costs, as two of three participants showed no changes on these metrics. This may represent a limitation of those metrics or the accelerometers. It could also be a result of only including representative data from three participants, although we had hoped to see dual task motor costs in 100% of participants, regardless of sample size. Even though the heel strike data from the smart devices provided detailed and precise data, a significant limitation, at present, is the amount of time and expertise it takes to process and interpret these data (up to 1.25 hours/participant). Ideally, we would like this processing and interpretation to take fewer than 10 minutes and require little-to-no prior training. We need to develop an app to streamline this processing. Additionally, although we observed consistent dual task motor costs in the representative athletes, we found that one participant did not demonstrate a dual task cognitive cost on the LE subtask and a different participant did not demonstrate a dual task cognitive cost on the UE subtask. Preferably, the method would elicit a dual task cognitive cost on both subtasks in all participants (regardless of the sample size), which may suggest a need for more challenging cognitive tasks. Alternatively, this finding may suggest that cognitive abilities are less susceptible to dual task interference and we should focus on dual task perturbations in motor performance.

The initial goal of the work was to develop a practical, sensitive tool that can improve evaluation and treatment of mTBI. In contrast to many of the dual task paradigms used in past work14, the original DTS and revised DTS use portable, inexpensive equipment, and most conditions are easy to score with no prior training. Additionally, we included a novel evaluation of upper extremity function, specifically hand-eye coordination, while previous work focused solely on lower limb or lower extremity abilities11,12,13,14. Thus, the method has significant potential to contribute to mTBI evaluation protocols, as it could be administered in a variety of environments (e.g., rehabilitation centers, doctor’s offices, gymnasiums, and athletic training rooms) for a wide-range of competitive athletes. Ultimately, we need to determine that the DTS is sensitive to effects of sports-induced mTBI, but the steps we have taken thus far suggest that the DTS administration protocol is a practical way to elicit dual task effects in high-performing athletes.

To date, mTBI evaluation is limited to self-reported symptoms and objective measures that have poor test-retest reliability, rely on baseline testing or are not challenging enough for high-performing athletes7,8,9. The DTS includes challenging tasks that evaluate both lower and upper extremity performance. Currently, we have not established that the DTS is sensitive to effects from mTBI, but we are in the process of collecting those data. Additionally, we seek to better understand the neural underpinnings of single and dual task behavior in healthy athletes and those with sports-induced mTBI by using the newly-created neuroimaging-compatible DTS. This understanding will serve to help us further refine the evaluation methods, like the DTS, and provide insight into optimal treatment paradigms.

Disclosures

The authors have nothing to disclose.

Acknowledgements

We would like to thank Ms. Isabelle Booth, a Colorado State University honors student who assisted with accelerometry data analysis. We would also like to acknowledge funding from NIH K12 HD055931 and K01 HD096047-02 issued to author J.S.

Materials

Hardware (in alphabetical order)
NIRx NIRSport2 Device: NSP2-CORE1616 NIRx Reference #: GC359 "The NIRSport 2 is a user-friendly, modular, and robust wireless functional near-infrared spectroscopy (fNIRS) platform which measures hemodynamic responses to neuroactivation via oxy-, deoxy-, and total hemoglobin changes in the cerebral cortex.The NIRSport 2 comes with a host of ready-to-implement upgrades and modules to meet the needs of a broad range of cognitive neuroscience applications." (Direct quote from nirx.net/nirsport)
NIRx NIRSCap (available in 5 difference sizes) NIRx N/A "The NIRScap consists of a measuring cap and optode holders. The optode holders fit into the slits of the measuring cap." (Direct quote from NIRx's NIRScap Getting Started Guide)
NIRx Optode Sources (x 2) NIRx Reference #: GC359 "8-source active source bundel for fiberless optical illumination with dual tip; 240 cm long." (Direct quote from NIRx Packing List Description)
NIRx Optode Detectors (x 2) NIRx Reference #: GC359 "Bundle of 8x active sensores for fiberless optical detection; dual tip; 240 cm long." (Direct quote from NIRx Packing List Description)
NIRx Short Distance Detector Probes NIRx N/A "The probes come in a bundle of eight detector clips that allows coupling of short-distance data from eight independent sources sites to one common detector channel on the instrument." (Direct quote from NIRx's Short Distance Detector Probes Getting Started Guide)
Software (in alphabetical order)
Aurora NIRx N/A "NIRSport 2 Acquistion Software. Aurora fNIRS connects to your NIRSport 2 device via Wi-Fi or USB and can set-up a complete experimental configuration in only several clicks. Thanks to the automated signal optimization algorithm, Aurora fNIRS ensures optimal signal quality before a measurement is started. Raw data, HbO and Hb concentration changtes can be visualized in real-time in several display modes. In addition, high-end whole head visualizations are immediately available. Recorded data can be exported over the integrate Lab Streaming Layer (LSL) protocol, allowing for real-time processing in Brain-Computer Interface (BCI) and Neurofeedback paradigms." (Direct quote from nirx.net/software)
Matlab Math Works N/A "MATLAB® combines a desktop environment tuned for iterative analysis and design processes with a programming language that expresses matrix and array mathematics directly. It includes the Live Editor for creating scripts that combine code, output, and formatted text in an executable notebook." (Direct quote from mathworks.com)
NIRS Toolbox Developed by Huppert Brain Imaging Lab N/A "NIRS toolbox is a Matlab based analysis program." (Direct quote from huppertlab.net/nirs-toolbox-2/)
PsychoPy Python N/A "PsychoPy is an open source software package written in the Python program,ming language primarily for us in neuroscience and experimntal psychology research." (Direct quote from psychopy.org)
Lower Tech/Cost Research Supplies* (in alphabetical order)
AmazonBasics 60-Inch Lightweight Tripod with Bag Amazon Item Model #: WT3540 This lightweight tripod is perfect for most cameras up to 6.6 pounds. Setup is quick and easy. The included bag makes storage and transport a snap.The tripod’s legs can extend from 20” to 48”. Leg locks release smoothly and glide easily to your desired height. Crank up the center post for a tripod that is 60” tall. (Direct quote from Amazon.com)
iPod Touch x 2 Apple N/A Smart device with built-in accelerometer.
Panasonic Full HD Video Camera Camcorder HC-V180K, 50X Optical Zoom, 1/5.8-Inch BSI Sensor, Touch Enabled 2.7-Inch LCD Display (Black) Amazon Item Model #: HC-V180K Compact, lightweight and easy to use, the Panasonic Full HD Camcorder HC-V180K brings a fun, worry-free experience to high-resolution video capture. Featuring a 5-axis image stabilizer for maximum handheld stability, this 1080p camera’s super-long 50X optical zoom and up to 90X intelligent zoom quickly bring distant objects in focus. A convenient 28mm wide-angle lens allows you to fit more people and scenery into settings like weddings, reunions and vacations. An advanced BSI sensor assures low-light video image quality while Panasonic’s Level Shot function automatically detects and compensates for distracting camera tilting. For added fun, the camera includes creative filter effects like 8mm Movie, Silent Movie, Miniature Effect and Time Lapse Recording, all easily accessible on the 2.7-inch LCD touch screen. A two-channel zoom microphone works in tandem with the zoom to ensure crisp, clear audio up close or at any distance." (Direct quote from Amazon.com)
Post-it Notes, 3" x 3", Canary Yellow, Pack Of 18 Pads Office Depot/Office Max Item # 1230652 "Post it® Notes stick securely and remove cleanly, featuring a unique adhesive designed for use on paper."
Scotch 232 Masking Tape, 1" x 60 Yd Office Depot/Office Max Item # 910588 "High-performance paper masking tape produces sharp paint lines in medium-temperature paint bake operations. Scotch tape provides clean removal every time, even on traditionally difficult-to-remove surfaces." (Direct quote from officedepot.com)
Stanley Tools Leverlock Tape Measure, Standard, 25' x 1" Blade Office Depot/Office Max Item #389512 "Tape rule features a power return with automatic bottom lock for easy operation. High-visibility case color makes it easy to find. Special Tru-Zero hook allows use of nail as pivot to draw circles and arcs. Tape rule offers a multiple riveted hook and polymer-coated blade for longer life, blade wear guard and comfortable rubber grip. Protected blade resists abrasion, oils, dirt and most solvents. Tape rule has Imperial ruling with consecutive feet on top and consecutive inches on bottom after the first foot. Its belt clip allows easy carrying." (Direct quote from officedepot.com)
Stopwatch Office Depot/Office Max Item # 357698 "Offers split timing, precise to 1/100 of a second. Includes 6 functions — hour, minute, second, day, month and year." (Direct quote from officedepot.com)
Tourna Ballport Deluxe Tennis Ball Hopper with Wheels – Holds 80 Balls Amazon Item Model #: BPD-80W "Balloon port 80 deluxe holds 80 balls and comes with wheels for easy Maneuverability. The handles are an extra long 33 inch for more convenient feed and pickup. Very lightweight yet durable makes this one of the most premium hoppers on the market. Loaded with patented features: legs lock in up and down position. Bars at the top slide closed so your the balls don't fall out during transport. Bars roll at the bottom so the ball slips in the hopper easily." (Direct quote from Amazon.com)
Tourna Pressureless Tennis Balls with Vinyl Tote (45 pack of balls) Amazon Item Model #: EPTB-45 "45 Pressure less tennis balls in a vinyl tote bag. Bag has a zipper for secure closure. Balls are regulation size and durable. Suitable for practice or tennis ball machines. Balls are pressure less so they never go dead. Pressure-less means they never go dead, which makes them great for tennis practice, ball machines, filling up ball baskets and hoppers, or just making sure your pet has hours of fun chasing these balls. They fit Chuck-it style dog ball launchers and automatic ball launchers. Durable rubber and a premium felt ensures their use can be universal, whether your a budding tennis player or a pet owner." (Direct quote from Amazon.com)
Velcro Velcro N/A Self-adhesive strips and wraps; used to secure smart devices.
Yoga Block 2 Pack – 2 High Density Light Weight Exercise Blocks 4 x 6 x 9 Inches Support All Poses – Lightweight Versatile Fitness and Balance Odor Free Bricks (Note: 6 blocks are needed for Dual Task Screen) Amazon N/A "These blocks are made from recycled high density EVA foam and provide firm support in a wide range of different yoga poses. This will improve your posture and you can stay in challenging poses for longer." (Direct quote from Amazon.com)
*These items or comparable items can be obtained from a number of other sources

References

  1. Gardner, R. C., Yaffe, K. Epidemiology of mild traumatic brain injury and neurodegenerative disease. Molecular and Cellular Neuroscience. 66, 75-80 (2015).
  2. Oyegbile, T. O., Dougherty, A., Tanveer, S., Zecavati, N., Delasobera, B. E. High Sleep Disturbance and Longer Concussion Duration in Repeat Concussions. Behavioral Sleep Medicine. , 1-8 (2019).
  3. Schatz, P., Moser, R. S., Covassin, T., Karpf, R. Early indicators of enduring symptoms in high school athletes with multiple previous concussions. Neurosurgery. 68 (6), 1562-1567 (2011).
  4. Yrondi, A., Brauge, D., LeMen, J., Arbus, C., Pariente, J. Depression and sports-related concussion: A systematic review. La Presse Médicale. 46 (10), 890-902 (2017).
  5. Haider, M. N., et al. A systematic review of criteria used to define recovery from sport-related concussion in youth athletes. British Journal of Sports Medicine. 52 (18), 1179-1190 (2018).
  6. Conway, F. N., et al. Concussion Symptom Underreporting Among Incoming National Collegiate Athletic Association Division I College Athletes. Clinical Journal of Sport Medicine. 30 (3), 203-209 (2020).
  7. Broglio, S. P., Guskiewicz, K. M., Norwig, J. If You’re Not Measuring, You’re Guessing: The Advent of Objective Concussion Assessments. Journal of Athletic Training. 52 (3), 160-166 (2017).
  8. Broglio, S. P., Katz, B. P., Zhao, S., McCrea, M., McAllister, T. Test-retest reliability and interpretation of common concussion assessment tools: Findings from the NCAA-DoD CARE Consortium. Sports Medicine. 48 (5), 1255-1268 (2018).
  9. Howell, D. R., et al. Examining Motor Tasks of Differing Complexity After Concussion in Adolescents. Archives of Physical Medicine and Rehabilitation. 100 (4), 613-619 (2019).
  10. Buttner, F., et al. Concussed athletes walk slower than non-concussed athletes during cognitive-motor dual-task assessments but not during single-task assessments 2 months after sports concussion: a systematic review and meta-analysis using individual participant data. British Journal of Sports Medicine. 54 (2), 94-101 (2020).
  11. Howell, D. R., Buckley, T. A., Lynall, R. C., Meehan, W. P. I. Worsening dual-task gait costs after concussion and their association with subsequent sport-related injury. Journal of Neurotrauma. 35 (14), 1630-1636 (2018).
  12. Howell, D. R., Kirkwood, M. W., Provance, A., Iverson, G. L., Meehan, W. P. Using concurrent gait and cognitive assessments to identify impairments after concussion: a narrative review. Concussion. 3 (1), 54 (2018).
  13. Lee, H., Sullivan, S. J., Schneiders, A. G. The use of the dual-task paradigm in detecting gait performance deficits following a sports-related concussion: a systematic review and meta-analysis. Journal of Science and Medicine in Sport. 16 (1), 2-7 (2013).
  14. Solomito, M. J., et al. Motion analysis evaluation of adolescent athletes during dual-task walking following a concussion: A multicenter study. Gait Posture. 64, 260-265 (2018).
  15. Stephens, J. A., Nicholson, R., Slomine, B., Suskauer, S. Development and pilot testing of the dual task screen in healthy adolescents. American Journal of Occupational Therapy. 72 (3), (2018).
  16. Ptak, R. The frontoparietal attention network of the human brain: action, saliency, and a priority map of the environment. Neuroscientist. 18 (5), 502-515 (2012).
  17. Jasper, H. Report of the committee on methods of clinical examination in electroencephalography: 1957. Electroencephalography and Clinical Neurophysiology. 10 (2), 370-375 (1958).
  18. Brigadoi, S., Cooper, R. J. How short is short? Optimum source-detector distance for short-separation channels in functional near-infrared spectroscopy. Neurophotonics. 2 (2), 025005 (2015).
  19. Sato, T., et al. Reduction of global interference of scalp-hemodynamics in functional near-infrared spectroscopy using short distance probes. Neuroimage. 141, 120-132 (2016).
  20. Scholkmann, F., et al. A review on continuous wave functional near-infrared spectroscopy and imaging instrumentation and methodology. Neuroimage. 85, 6-27 (2014).
  21. Baker, W. B., et al. Modified Beer-Lambert law for blood flow. Biomedical Optics Express. 5 (11), 4053-4075 (2014).
  22. Barker, J. W., Aarabi, A., Huppert, T. J. Autoregressive model based algorithm for correcting motion and serially correlated errors in fNIRS. Biomedical Optics Express. 4 (8), 1366-1379 (2013).
  23. Aguirre, G. K., Zarahn, E., D’Esposito, M. The variability of human, BOLD hemodynamic responses. Neuroimage. 8 (4), 360-369 (1998).
  24. Stephens, J. A., Liu, P., Lu, H., Suskauer, S. J. Cerebral Blood Flow after Mild Traumatic Brain Injury: Associations between Symptoms and Post-Injury Perfusion. Journal of Neurotrauma. 35 (2), 241-248 (2018).

Play Video

Cite This Article
Aumen, A. M., Oberg, K. J., Mingils, S. M., Berkner, C. B., Tracy, B. L., Stephens, J. A. Revised and Neuroimaging-Compatible Versions of the Dual Task Screen. J. Vis. Exp. (164), e61678, doi:10.3791/61678 (2020).

View Video