1 Introduction

Body orientation and balance are maintained by the integration of information from multiple sensory systems. Altered integration of this multisensory information can be elicited by endogenous maladies that alter established sensorimotor relationships (e.g., vestibular injury, disease, or aging), by exogenous alterations of the gravitoinertial force environment (e.g., aircraft flight, travel by sea, acceleration in laboratory acceleration devices, or spaceflight), or by visual-vestibular rearrangements (e.g., exposure to driving or flying simulators). Regardless of which challenge causes altered sensation and control of body orientation and motion, there is an initial period of disruption of functional abilities, leading to disorientation, imbalance, disruption of gaze control, and/or motion sickness symptoms. The process of disruption and subsequent adaptation is similar for patients recovering from certain kinds of endogenous vestibular or balance pathology as for healthy people adapting to the exogenous vestibular challenge posed by exposure to a vehicle simulator or a prolonged voyage at sea or into space [16]. Once adaptation is achieved by sensorimotor recalibration to these new processing demands, coordination and well-being is restored.

We sought to determine whether an exogenous balance perturbation (the aftereffect of body rotation followed by a sudden stop) or an endogenous balance perturbation (the effects of concussion followed by lingering dizziness) would degrade performance of an applied psychomotor performance task requiring postural equilibrium during goal-directed activity. The exogenous rotation perturbation we employed is a challenge to vestibular and balance functioning familiar to any child and has been studied at least since the time of Erasmus Darwin. The endogenous challenge we studied (variously called concussion or mTBI) is strongly associated with dizziness and imbalance in ways described by Lawson et al. [12,13,14].

The task chosen for study was simulated rifle shooting. While shooting has many degrees-of-freedom as a performance task and is susceptible to learning effects, it is a fundamental skill in many military occupational specialties. While negative findings for shooting performance can be difficult to interpret, when performance decrements are detected for shooting, there is a more direct intellectual pathway to the interpretation of operational or functional significance of the decrement than there is for most other psychomotor tests. Shooting is also a suitable analogue of goal-directed activity requiring balance plus psychomotor or “eye-hand” coordination. The shooting tasks and other methods are described below.

2 Methods

2.1 Participants

Volunteers were evaluated in two studies of exogenous (Study 1) and endogenous (Study 2) balance perturbation. Study 1 included 60 healthy participants (without history of head injury or pathologies affecting balance functioning) from Fort Rucker, AL (mean age 27, SD 5.4; 56 men and 4 women) asked to shoot immediately after a 20-second period of rotation at a constant velocity of 180 degrees-per-second, a stimulus intended to produce a vestibular aftereffect on balance and shooting.Footnote 1 Study 2 of endogenous balance perturbation included 40 participants from Fort Benning, GA (mean age 33, SD 6.9; 38 men and 2 women). Thirty of the 40 Fort Benning participants had been patients of a traumatic brain injury treatment center, who had a documented diagnosis of mTBI [1]. Diagnosis was confirmed by their medical history (DD Form 2807) and an interview with our study physicians (Major Timothy Cho and Dr. John Campbell). The remaining 10 Fort Benning participants were healthy (no history of head injury or balance pathologies). While our plan was to compare the Study 1 (healthy Fort Rucker) participants to the Study 2 (mTBI Fort Benning) participants, we included 10 healthy Fort Benning participants as a site-of-recruitment check. To ensure that participants were familiar with military marksmanship, the study recruited U.S. Army active-duty Soldiers, Natural Guard/Reserve Soldiers, or civilians with recent military experience.

2.2 Apparatus

The Engagement Skills Trainer 2000 was employed to evaluate performance following exogenous or endogenous balance perturbation (Fig. 1). The device is a small arms simulator used throughout the U.S. Army for rifle marksmanship training [2]. Shooting was performed with an M4 carbine using iron sights and pneumatically simulated recoil [2]. In the Army marksmanship qualification test using the EST, standard targets appear at virtual distances between 50–300 m, with the furthest targets subtending approximately 0.93° of visual angle. Our dynamic shooting-while-moving tasks were more challenging, so we employed virtual shooting distances of 15–75 m (with the furthest targets subtending approximately 3.72°). This approach made speed of performing accurately our focus rather than raw accuracy and minimized speed-accuracy trade-offs [10]. This approach also was more representative of dynamic, short-range, fast reaction shooting (e.g., during breaching) than long-range shooting would be.

Fig. 1.
figure 1

The EST 2000 apparatus [2].

Full size image

2.3 Psychomotor Performance Tasks

The participants executed four dynamic marksmanship tasks (Table 1) that were designed to be a compromise between the body movements known to challenge postural balance in established clinical balance gait tests and the skills needed to perform operationally relevant military weapons handling tasks. For example, established clinical assessments such as the Berg Balance Scale [3] and the Dynamic Gait Index [19] respectively challenge a patient’s ability to pick up objects from the floor or walk forward while turning his/her head in yaw. Similarly, we designed a task requiring participants to pick up a rifle and shoot at a target (called “pick up rifle and shoot”), and a task requiring participants to walk forward and pivot (in yaw) to shoot to the side (called “walk, head-swivel, and shoot”). The full set of tasks is shown in Fig. 2. The initial development of the four tasks and their test properties is detailed in Grandizio et al. [6]. The tasks were designed to challenge balance similarly to established balance tests, but to add a readily quantifiable goal-directed psychomotor performance task of clear military relevance, i.e., shooting. The main measure of shooting performance was shooting throughput, or number of accurate shots per minute. Details concerning the collection and analysis of this measure and the other performance variables collected in our shooting studies is provided in Lawson et al. [17].

Table 1. The four dynamic marksmanship tasks
Full size table
Fig. 2.
figure 2

The shooting tasks, from top to bottom: kneel and shoot; pickup rifle and shoot; walk, turn, and shoot; traverse beam and shoot (cable was kept clear of participant’s feet by a metal guide rod not depicted).

Full size image

2.4 Self-report Data

2.4.1 Perceived Workload

The National Aeronautics and Space Administration Task Load Index (NASA TLX) provides a workload assessment based on ratings for six dimensions, viz., mental demands, physical demands, temporal demands, own performance, effort, and frustration. It was used to assess perceived workload immediately after performing each shooting task. Detailed instructions can be found in the test administration guide [18]. A raw TLX scoring procedure was used [5, 7].

2.4.2 Dizziness

Participants completed the Dizziness Handicap Inventory (DHI). The DHI is a validated questionnaire that estimates functional disabilities related to vestibular dysfunction [8]. The questionnaire consists of 25 items related to the effects of dizziness, including physical (e.g., Do quick movements of your head increase your problem?), functional (e.g., Because of your problem, do you have difficulty reading?), and emotional effects (e.g., Because of your problem, are you embarrassed in front of others?). Answers are scored as no = 0, sometimes = 2, and yes = 4. The highest score possible is 100, with higher scores indicating a greater impairment. These data allowed us to characterize the participants’ general daily problems with dizziness.

2.5 Procedures

The study protocol was approved by the Headquarters, U.S. Army Medical Research and Materiel Command Institutional Review Board. Written informed consent was obtained, after which the participants filled out questionnaires concerning daily dizziness etc., before being introduced to the EST weapons simulator. Participants first zeroed the rifle, calibrating the laser sensor to the direction the rifle was pointing. They then completed the Army’s standard, relatively static marksmanship qualification test (involving shooting from the standing, kneeling, and prone positions). This permitted them to refamiliarize themselves with basic weapons handling and sight usage, and allowed us to use their static marsksmanship performance as a covariate during subsequent analyses.

Finally, the participants were introduced to the new dynamic marksmanship tasks. A member of the research team instructed participants in the proper execution of each task. Participants then practiced each shooting task prior to obtaining baseline performance data. The ultimate purpose of the shooting tests was to detect abnormal exogenous or endogenous vestibular/balance function, while avoiding sensitization or adaptation of concussion/mTBI patients. It was deemed important to limit head movement among mTBI participants, to avoid triggering unwanted symptoms. Therefore, while three sessions of practice were completed, the first was done at slower speed and with conscious limitation of head movement amplitude and speed. Upon completing the practice sessions, participants performed the dynamic marksmanship tasks. NASA TLX workload ratings were collected immediately following each of the shooting tasks. The performance order of the tasks in the dynamic battery was balanced.

3 Results

The main findings from these studies are summarized as follows:

3.1 We Identified the Most Reliable of the Shooting Tasks

The most reliable task of the four tasks assessed was kneeling while shooting at targets presented to the left and right (~30-degree sweep) of the visual field (test-retest Pearson r = 0.54) (Fig. 2).

3.2 We Identified the Shooting Task that Was Most Sensitive to Exogenous Balance Perturbation

Kneeling while shooting was the task that was most sensitive to changes in shooting performance caused by a brief period of rotation-induced dizziness followed by a sudden stop and an immediate shooting challenge (Pseudo R2 variance accounted = 0.54).

3.3 We Failed to Identify a Shooting Task that Was Sufficiently Sensitive to Shooting Performance Degradation Associated with Endogenous Balance Perturbation

Standard range qualification score (the number of accurate hits recorded during a standard or non-dynamic simulated range qualification task) was included as a covariate in each analysis to help account for individual marksmanship ability among the participants. Preliminary analyses of the dynamic shooting data found that when correcting for standard marksmanship ability, a significant effect [(F(7, 68) = 3.07, p = .007)] remained for the site at which the data was collected (Fort Rucker versus Fort Benning). Therefore, data collection site was also included as a covariate in the analyses. After correcting for variance due to initial (static) marksmanship qualification scores and site differences, a one-way MANCOVA nevertheless failed to detect an mTBI-specific difference in performance, with the overall main effect falling well short of significance [F(7,68 = 1.42, p = 0.13)].

Despite the failure to find a performance degradation due to the endogenous balance challenge we studied, the mTBI patients did feel dizzier than normal subjects while shooting and reported having to work harder to achieve a level of shooting performance that was comparable to normal subjects. These dizziness and workload findings are summarized below.

A one-way analysis of variance (ANOVA) demonstrated that DHI scores were significantly higher in the injured group than in the uninjured group [(F(1, 79) = 161.56, p < .001)]. Mean and standard error values for both groups are represented in Fig. 3 (note that the values are too small in the uninjured group to be visible in the Figure). The findings confirmed that the injured group was experiencing significantly greater dizziness-related handicaps during their daily activity, according to this established clinical instrument. In the uninjured group, participants’ DHI scores suggested no functional impairment (according to DHI clinical cut-offs), whereas DHI scores for the injured group suggested mild-to-moderate impairment.

Fig. 3.
figure 3

Mean and standard error for dizziness handicap index (DHI) scores by injury condition. The DHI characterizes dizziness-related impairment during various activities of daily living. Mean DHI Scores by Group were 0.04 (SEM 0.04) for the uninjured group and 26.53 (SEM 2.73) for the injured group.

Full size image

The mTBI group felt they needed to work harder than the healthy group to achieve comparable shooting scores. There was a significant overall between-subjects effect on subjective workload by condition demonstrating that, in general, injured participants reported significantly higher subjective workload scores than uninjured participants [F(1, 78) = 14.77, p < .001]. The main effect for condition is represented in Fig. 4.

Fig. 4.
figure 4

Mean and standard error for NASA TLX scores by injury condition. Mean NASA TLX scores by group were 24.95 (SEM 2.35) for the uninjured group and 39.91 (SEM 3.11) for the injured group.

Full size image

4 Discussion

4.1 Conclusions from the Present Study

We developed a prototype battery of dynamic shooting tasks that was intended to quantify goal-directed psychomotor performance during postural balance challenges and to pose a more realistic, operationally relevant balance challenge than traditional, static range marksmanship qualification tests. We then successfully identified a shooting task that was sufficiently reliable and sensitive to an exogenous balance perturbation to merit further refinement and study for potential future testing applications. However, more controlled research will be necessary to determine whether this or any of our other tests can distinguish healthy subjects from those affected by an endogenous balance challenge. The main limitation of our endogenous perturbation research was that our mTBI subjects were drawn from one site and most of our healthy subjects from another. We did not have access to mTBI subjects at our home site of Fort Rucker while initially developing the shooting tests, so we had to recruit our mTBI-affected subjects at Fort Benning during a second study. Unfortunately, the two groups could not be assumed to be from the same population (see Results), probably because the Fort Rucker participants were mostly aviation personnel, while the Fort Benning participants were mostly infantry personnel, who tend to be more practiced at rifle marksmanship.

4.2 Lessons Learned During Experimentation Which Are Relevant to Future Portable or Ambulatory Testing Applications

While the development of portable or ambulatory tests was not the purpose of this particular aspect of our research program, we learned useful lessons during the current study that are relevant to the future development of such portable systems. While the EST 2000 offers a versatile range of fairly realistic range training simulations and is highly relevant to military shooting performance, it required a fair amount of training and practice to use. It also required a utility van and several people to transport, and a large operating space within which to deploy. Furthermore, it required some trouble-shooting and repairs during our studies. A portable or ambulatory application would not necessarily have the same level of onsite support and labor.

Several modifications were necessary to adapt the system to dynamic and ambulatory shooting research [9, 15], which were outside the EST’s designed purpose. These included the addition of programs to permit sophisticated analyses of performance, repositioning of virtual targets (to increase subject movement), modification of the physical layout of the room to reduce trip hazards during ambulatory tests, increasing the spacing between subsequent targets to force the subjects to make larger head movements between shots, and the addition of an extended rifle cable and of a guide rod to direct the pneumatic cable away from the shooter’s feet. These modifications permitted the EST to yield some interesting findings concerning dynamic shooting. Nevertheless, the design purpose of the EST is to simulate range marksmanship training, rather than to support formal experiments requiring the analysis of dynamic shooting performance during rapid body/rifle movement or ambulatory shooting. While the company was supportive and the EST was adapted successfully to support the present experiment and to make interesting inferences concerning rifle kinematics during rifle movement and aiming [17], we found that some data could be lost during the rapid or large-amplitude rifle movements required by our research, due to the way the EST sampling algorithm prioritizes different channels of the incoming data. A research-focused dynamic shooting device would need to have high sampling rates to ensure all rifle kinematics data is acquired without gaps. Finally, while the EST measured goal-directed psychomotor activity (shooting) and could be adapted to indirectly infer a disruption in postural equilibrium in the present study, it was not designed to measure postural balance directly. A research-focused dynamic shooting device designed to challenge balance should include a dedicated measure of postural sway.

The optimal device for assessing dynamic shooting-while-moving in a situation that challenges balance would be small (or even portable or ambulatory), user-friendly, and robust. A small system would be advantageous because it could be rapidly transported and deployed to classroom or field settings using a mid-sized vehicle and without the need for multiple support personnel. A highly usable system would require less training time and less set-up time. Finally, a robust system would experience less down-time and require less time spent on troubleshooting or repair. Increased robustness might be obtained by reducing the number and variety of electrical and mechanical machines which make up the shooting simulator, for example, by sacrificing or redesigning the pneumatically driven mechanical simulation of rifle recoil.

We envision the future integration of a small part-task shooting simulator with an objective indicator of postural sway. The simulator could consist of a virtual headset display with appropriate targets and a rifle that is suitably realistic. The simulator would present targets requiring head and rifle movement and yield a measure of one’s speed of accurate shooting. The shooting simulator would be coupled with a direct, objective indicator of postural sway, such as an off-body motion tracking system or a body-mounted accelerometer package.

Some initial strides towards a dynamic moving-while-balancing shooting simulator could be made relatively rapidly by focusing solely upon our best-performing task, which was kneeling-while-shooting and engaging targets to the left and right. Since the subject would not be walking around the room during this test, room-mounted tracking systems or body-worn accelerometers would not be needed. All that would be required for postural sway measurement would be either a commercial camera system (such as was developed long ago by Kennedy [11] and more recently by Mortimer and colleagues, [4, 20]), or a balance platform capable of measuring center-of-pressure (of which, several small models have been developed, Lawson et al. [14]). Such a device would permit rapid testing, evaluation, and application of the best-performing test from the current study. However, a fully ambulatory balance and gait measurement system would be needed to obtain objective gait data when multiple steps are required across a large space. Ultimately, this would consist of a rifle, virtual display, and accelerometer/GPS-based wearable motion tracker. The tracker would need to be sophisticated enough to avoid the tracking errors that have been seen in some past models of wearable balance tracking systems [14].