Validation of cognitive and neuromuscular latencies using an iPad-based tool to quantify simple reaction times in asymptomatic subjects

Methods

A total of 28 subjects were enrolled in the study between November 4, 2019 and November 25, 2019, with all 28 subjects finishing the study. The study was reviewed and approved by the Michigan State University Human Research Protection Program (STUDY00001620). Subjects consisted of an asymptomatic cohort of first- and second-year medical students recruited from Michigan State University’s College of Osteopathic Medicine (MSUCOM). Their involvement was voluntary, and no compensation was provided. At their first visit, the purpose of the study was explained to them. All potential subjects were then required to read the Informed Consent Form and encouraged to ask any questions that they may have had. Subjects were enrolled in the study after signing the Consent Form.

Two categories of tests were utilized. The first was a simple finger tapping test (FTT) designed to quantify pure NL (Figure 2). The second was a cognitive test (CT) that required a subject to solve a simple math equation (Figure 3). The equations were limited to addition and subtraction of numbers that would result in answers of less than 10. Subjects were allowed to familiarize themselves with each of the two tests.

Figure 3:

The simple reaction tool showing the simple math exercise screen.

Figure 2:

The simple reaction tool showing the orientation of the arm and forearm relative to the iPad screen.

The objective of the FTT was to measure the value of NL by having the subject simply tap a single target positioned on the iPad screen as fast as possible over an interval of 10 s. Measurements began when a target was displayed and ended when the target was tapped. The FTT had two levels of challenge: a small random displacement (SD) of the target of up to 6 cm, and a large random displacement (LD) of the target of up to 24 cm. After tapping the target, the iPad program presented the subject with a new target located in a new position on the iPad screen.

The objective of the choice test (CT) was for the subject to repeatedly solve a simple mathematical equation and then tap a target on the iPad screen containing the correct answer as fast as possible. Answers were presented to the subject on three targets, and only one of the targets contained the correct answer (Figure 3). Subjects were instructed to solve the equation and then to tap the target displaying the correct answer. The measurement began when a new math problem was displayed and ended when the target was tapped. The CT also had two levels of challenge: a small random displacement (SD) of all three targets of up to 6 cm and a large random displacement (LD) of all three targets of up to 24 cm with only one target displaying the correct solution to the equation. After tapping the target, the iPad program presented the subject with a new math problem and a new set of three targets was positioned on the iPad screen. The levels of displacement of the new targets were a function of the level of challenge.

Once the data were collected, SPSS (IBM SPSS Statistics for Windows, Version 28.0.1.0) was utilized to test for the statistical significance of the average latency between small and large displacements of targets for both the FTT and the CT tests.

Results

The following values of average latency for SRT (FTT) were measured for the FTT (Table 1). Utilizing the repeated-measures t-test analysis in SPSS, a significant difference was observed between the values of the average latency of the two target displacements at p<0.001 (Figure 4).

Figure 4:

A composite plot of the average latency for two levels of target displacement for the finger tapping test (FTT). There is a significant difference at p<0.001 between the average latency of the two target displacements.

The following values of average latency for SRT (CT) were measured for the CT (Table 2). Utilizing the repeated measures t-test analysis in SPSS, a significant difference was observed between the values of the average latency of the two target displacements at p<0.001 (Figure 5).

Figure 6:

A composite plot of the calculated cognitive latency (CL) for two levels of target displacement. There is not a significant difference between the value of CL for small random displacements (SD) and CL for large random displacements (LD) at p=0.696.

The magnitude of CL for both small and large displacements was calculated by subtracting the respective NL from the total SRT of the respective CT (Equations 2 and 3) (Table 3).

Utilizing the repeated-measures t-test analysis in SPSS, a significant difference was not observed between CL (SD) and CL (LD) at p=0.696 (Figure 6).

Figure 5:

A composite plot of the average latency for two levels of target displacement for the choice test (CT). There is a significant difference at p<0.001 between the average latency of the two target displacements.

Discussion

The results of this study are significant because they show that a modified FTT can be utilized to evaluate cognitive function as well as the neuromuscular reflex response. The study hypothesis was that cognitive latencies, related to a specific cognitive challenge, would remain constant even as neuromuscular latencies changed. In order to calculate NL, subjects were first presented with a target randomly positioned on an iPad screen. They were instructed to tap the target with their forefinger. After tapping the target, the iPad program presented the subject with a new target randomly positioned on the screen. Subjects were instructed to continue to tap the target as fast as possible for 10 s. Displacing the target between finger taps resulted in latencies that were a direct consequence of additional loading of the central nervous system (CNS). This resulted from the need to calculate the trajectory necessary to move the finger from the current location of the target to a new location.

To accomplish this, the eyes need to move rapidly to locate the target (otherwise known as “saccades”). The prefrontal cortex would then organize a motor plan for tapping the finger and then pass it to the motor cortex for execution. Working in conjunction with the cerebellum, the signal from the motor cortex would result in a smooth wrist flexion and finger tap. Mechanoreceptors in the fingertips, specifically Meissner’s corpuscles and Merkel’s discs, would detect the finger tap, producing action potentials that travel back up the spinal cord, through the thalamus, and ultimately into the somatosensory cortex in the parietal lobe [5], where the sensation of touch would be recognized by the brain, resulting in extension of the wrist and finger. Meanwhile, software written for the iPad sensed when the target had been tapped, causing the target to disappear and providing visual feedback to the subject. The iPad software then calculated the value of NL between the time when the target first appeared, and the time when the subject tapped the target. The process was then repeated until the end of the 10-s period.

Increasing the challenge presented to the subject from a small random displacement to a large random displacement requires an increase in time to acquire the target and an increase in time to plan/execute movement. We anticipated that the additional time required to process these variables contributed to the increased latency times seen in the FTT (SD) and FTT (LD) groups (Figure 4). Because the value of displacement for FTT (SD) is small, it was assumed that latency associated with movement of the mass of the hand and arm could be ignored. Utilizing our experimental values, the latency between a change in target position and the initiation of hand movement was calculated to be equal to 333 ms. This compares favorably with a reported value of 326 ms (±31, SD) for monkeys (n=2) [6].

Our study added a CT to the basic FTT protocol to expand its utility beyond calculating just neuromuscular latency (NL) in order to enable the calculation of cognitive latency (CL). The CT required a subject to solve a mathematical problem and then to visually locate a randomly positioned target that displayed the correct answer. Solving the problem had to occur before the subject’s brain could generate the trajectory necessary to move their forefinger to the new target position. This is an example of serial processing, as opposed to parallel processing [7], in which the SRT is equal to the sum of the average latency necessary to process the mathematical equation of the CT and the average latency to process the neuromuscular response of the FTT. To test the study hypothesis, the value of CL was calculated, for both SDs and LDs, by subtracting the respective NL from the total SRT.  Our values of CL (SD)=731 msec and CL (LD)=758 msec compare favorably with reported values (n)=20, average=888 msec (±72 SD) [8]. It was seen that there was no significant statistical difference between the values of CL for an SD or an LD of target displacement (Figure 6). The study hypothesis that cognitive latencies, related to a specific cognitive challenge, would remain constant even as latencies due to neuromuscular challenges changed, was verified. It is not surprising that CL associated with solving a mathematics equation would be independent of latencies due to target displacement because they depend upon different neural processes.

Now that COVID-19 issues have been reduced, we hope to recruit subjects diagnosed with TBI to test the iPad-based tool to see if it provides an advantage for the management of patient treatment. The ability to track cognitive and neuromuscular latencies over time should allow assessment of the effectiveness of a treatment plan as well as a metric to help gauge when patient progress has reached a plateau.

Conclusions

An analytic model for a specific SRT test has been formulated, and software programs have been developed, for an iPad-based tool that enables the separation and quantification of normative components of the SRT that are due to neuromuscular and cognitive latencies. Validating baseline values for subject-dependent latencies of the SRT utilizing asymptomatic subjects allows for future testing against a cohort of subjects diagnosed with TBI. By measuring latency under controlled conditions, it is hoped that this SRT tool will aid in the identification of regions of the brain that have a functional deficit due to a traumatic event, enabling focused rehabilitation efforts specifically tailored for the individual patient.