Tobii technologies, a Swedish Company, has recently released a new user interface device that uses eye tracking ability employing built in infrared sensors to accurately allow the user to merely look at different points of interest on a computer screen. Such technology will allow users to engage very accurately with their eyes to zoom in, scroll flip programs and all the varied activities that are becoming the App universe of computer programming. There is one minor problem, what happens if you have a concussion could you use the eye tracking interface?
In the latest Behaviour Research Methods paper published online September 7, 2012 titled Dynamic visuomotor synchronization: Quantification of predictive timing by Jun Maruta, Kristin Heaton, Elisabeth Kryskow, Alexis Maule and Jamshid Ghajar have accurately described the normal aspects that visual tracking works within our stable head gaze for any thing that captures the eyes attention that is moving within the perceptual visual field. Listen to their description of positioning of the eyes overlapping the control centers within our brain that organize attention: ” Visual tracking supports perceptual stability of the object of interest that is in motion. When visually tracking a moving target to maintain its image on the fovea, spatial and temporal predictions are used to circumvent the neural delay required for the visuomotor processing. In particular, the internally generated predictive drive must be synchronized with the external stimulus during continuous tracking, which highlights an important distinction between being able to predict that a target will appear at a specific location and being able to predict when that event will occur. Accurate predictive timing is the ability to synchronize what is expected with what is observed, which is considered to be a function of attention.” Losing this predictive capacity to synchronize in the vernacular of those suffering a concussion is typically described as, ‘I’m in the fog.’ Student athletes with concussions complain of having trouble following lecture writing on a chalk board the severity of following the assembly of written words is just to much to follow, the fatigued student blanks out his attention stream. It’s not because of motivation it’s his eye positioning capacity that is compromised.
Do you recall from a previous essay what author Ivan R Schwab described in his eloquent book, Evolution’s Witness how eyes evolved how he emphasized that the eye was in essence the very first brain developed sensing network, perhaps the first coherent brain network? When you are a concussion scientist investigator,s what do you pay attention to, what changes in the brain after a concussion? In terms of the overall capacity of our modern 2012 brains, the burden of interpreting the moment to moment visual panorama consumes 80% of our brain power. So if a concussion happens then trying to resolve what we might not see properly as a result of the concussion derangement, changes will surly involve some or parts of that 80% within the brain networks, since the eyes are the oldest network , from an evolutionary point within our brain. If paleontological fossil records are accurate for jellyfish with their multiple light sensitive eye like structures, then their eye timeline stretches back at least 800 million years ago. The problem is the jellies are soft bodied , so that any traces of their existence is virtually obliterated within any bed of sediments from that age. But eye structures already existed at the Cambrian explosion of life following the transformation 550 million years ago. The genes for vision had already become programmed, available for creatures like jelly fish to navigate within their surroundings at least 300 million years before the Cambrian diversity of life form explosion.
What kind of eye tracking deregulation may happen combined with head trauma?
Head trauma can involve nerve palsies in the event of localized facial injuries being commonest in the fourth cranial nerve even following trivial head trauma. The real dilemma as outlined in the Journal of Injury (1990) 21, 351-352 by authors J Kwartz, B. Leatherbarrow and H Davis in their paper Diplopia following Head Injury observed that a medical examiner might mistakenly decide that the eye movements are normal when in fact the eye movements are not. How can this be so? According to the authors, “The diagnostic problem is that the ocular (eye) position of a patient with a fourth nerve palsy may look normal in the primary (normal horizontal position). This is because all four recti muscles will move the eye horizontally or vertically when there is no associated third or sixth nerve palsy.
So what are the mechanism of possible injury that can cause diplopia ?
Diplopia from the Wikipedia web site is the effect of seeing a double image. The simultaneous perception of two images from a single object in the visual field may be displaced vertically or horizontally. ” It is usually the result of impaired function of the extraocular muscles (EOM’s), where both eyes are still functional but they cannot converge to target the desired object, from ‘Sullivan, S.B & Schmitz, T.J. (2007). Physical Rehabilitation. Philadelphia, PA: Davis. ISBN 978-0-8036-1247-1, where both eyes are still functional but they cannot converge to target the desired object. Problems with EOM’s may be due to mechanical problems, disorders of the neuromuscular junction, disorders of the cranial nerves III< IV< VI that stimulate the muscles, and occasionally disorders involving the supranucular oculomotor pathways.”
Concussion researchers have begun to piece together the various injuries associated with both mild and traumatic forms of brain injury within the subcortical white matter pathways. Traditional scanning technologies such as magnetic resonance imaging scans, do not readily detect then those subtle changes to white matter tracts in the form of diffuse axonal injury. However micro-structural changes in white matter are detectable using diffusion tensor imaging (DTI) employing fractional anistrophy as an indicator of white matter micro-structural integrity. What the DTI technique is able to discriminate is the diffusion of water molecules as directionally constrained in parallel fiber arrangements of white matter tracts. But like all techniques there are some difficulties since the underlying fiber directions determine the kind of scatter pattern to the signal. In other words, like a rope that may have various portions of twisted components within the rope itself, the discriminator result to resolve with precision is not matching the actual fiber orientation. These kinds of conflicts are healthy since the interpretation is evolving toward a better resolution of signal changes along the length of fiber tracts. What this I think boils down to is the underlying fiber anatomy needs to be respected for proper interpretation of signal changes along white matter tracts following concussive injury. Looking at the normal appearances of imaging data sets the preferred presentation is based on a series of congruent slices stacked, each slice on top of the neighboring slice. What is hard to appreciate is this artificial separation happens when viewing each slice. What is the point I am making here? Lets step back a minute. The entire brain evolved as an embryonic progression of guided positioning into a very intertwined series of tracts spaces with zones of separation yet all integrated in terms of function. If we get into the minute detail of each brain cell type the exercise gets lost very easily. So here is where studying concussions forces upon the observer to pay attention to the underlying rules of assembly of the brain tissue to arrive at the finished guided end product. The rules to accomplish assembling this massive architecture are based on the common final attachments for mobilizing into a final mobile elastic tissue. The very basic cellular assembly appears to be pivoting as a scripted tensegrity harmonized tension integrity structure. Concussions disturb the harmonized tensional integrity within the structure. White matter fiber tracts may be the window of observation of stretched changes occurring following a concussion. The question becomes: are the zones of diffusion anistrophic disparities showing the remnants of deformation that occurs in the dynamics of the concussion, deformations impacting into the entire brain structure? We have been educated to think of things inside the skull as a coup- contre-coup rebounding into deforming compression effects as the brain is jarred upon itself, like striking a bowl of jello, to energy rippling through the substance. That impression unfortunately is the simplistic aspect that appeals to the casual observer. Our brains however, are not bowls of jello, they are tensegrity structures of cellular architecture created and assembled with the heritage of billions of years of evolution. Yet the curious mind when it tries to match things up, sees a pattern in a brain image for measuring changes using affected fiber tracts as passing tension changes within a tension network of assembly. The white fiber tracts permit the eloquence of the observation to assess the concussive behavior into specific vulnerable cluster zones. So the question zeros in on these cluster zones of vulnerabilities: where are they located in terms of structure as linked effects from the spreading distortion of the concussive forces? What physiological changes might be linked to these distorted white tract zones? That’s the essence of our own research attempts to follow the minute tension changes of a concussed-disturbed brain. So far we have teased out oculomotor changes, eye vision tracking is off in its synchronization capacity, which also affects central balancing mechanisms, which I referred to in my previous essay Surfing on a concussed brain. So the stark conclusion becomes: the eyes themselves are decelerating in a different profile compared to the brain tissue decelerating because the eyes are not surrounded nor encased by bone. The eyes have mass and all objects with mass decelerate at their own specific rate of velocity but they have no solid surface to contact like the brain does. Therefore the eyes bulge out like Bernie’s eyes at the beginning of this essay. When this happens, now we have changes to measure, like the loss of the visual tracking synchronization. The fabric of supporting tissue structures holding the eyes in place in their bony open cup are stretched during the deceleration of the concussed forces. Each eyeball is a tethered structure moving away from its oriented normal position, that is perhaps the observation of diplopia following a mild blow to the head. The individual eye tracking capacity is no longer synchronized. The definition now reveals itself for us of the concussed eye as well as the concussed brain. I coin the term as: occulomotor concussion.
- White Matter Fiber Tracts: Visualizations of fiber track data from Betty Lee
“Despite the recent advances in elucidating the neural circuits that convey the visual information to generate pursuit eye movements, the precise localization and interrelationships of the neural substrates for the extra-retinal, cognitive components of visual tracking have yet to be determined. However it is generally assumed that the substrates for these components are broadly distributed. Thus, even a subtle neurocognitive dysfunction could impair visual-tracking behavior. Abnormalities in visual-tracking behaviors have been associated with various psychiatric and neurological disorders, brain lesions and pharmacological effects are broadly distributed; thus, even a subtle neurocognitive dysfunction could impair visual-tracking behavior. Abnormalities in visual-tracking behaviors have been associated with various psychiatric and neurological disorders, brain lesions and pharmacological effects.”
“Using video-oculography, eye movement can be monitored easily, precisely, and continuously. Furthermore, oculomotor paradigms are resilient to inconsistent or poor subject effort. However, to evaluate specific visual tracking abnormalities in a quantitative manner, characterization of normal behavior using a well-defined testing paradigm is necessary. Visual-tracking performance should then be objectively quantified using standardized parameters such as smooth pursuit velocity gain, phase error, and root-mean square (RMS) error. Impairments in visuomotor synchronization may also be assessed by variability of gaze positional error relative to the target.”
“Our interest in developing a rapid assessment of attention in concussion patients has led to the use of a circular visual tracking paradigm. The diagnosis of concussion, or mild traumatic brain injury (TBI), is made difficult by symptoms that are often subtle and transient. Although impaired attention is a hallmark of TBI , the impairment can go undetected by traditional neurocognitive measures that rely on verbal or motor responses to discrete stimuli and are sensitive to subject motivation and effort.
The use of a visual-tracking paradigm for attention assessment is based on the hypothesis that attention impairments in concussion patients are a consequence of reduced efficacy of predictive timing . Our approach is supported by the evidence that eye movement and attention processes are implemented by closely overlapping areas of the brain and that attention is required during visual tracking. Our previous study of circular visual tracking in concussed patients suggested that impaired predictive timing, rather than disengagement from prediction, can result in poor tracking . This study also supported that impaired visual-tracking performance was related to injury of attention-related anatomical locations and diminished neurocognitive performance.
The primary goal of this study is to describe the indices and normal variations of dynamic visuomotor synchronization during circular visual tracking in healthy, young adult subjects, from which the criteria for abnormal performance can be derived. In addition, because the clinical utility of a test is ultimately limited by the reliability of its measurements, we aim to establish the test–retest reliability of the visual-tracking measures.”
“The study was performed under the strict guidelines of the United States Army Research Institute of Environmental Medicine (USARIEM) located at the Natwick Soldier Center, Natwick Massachusetts. These research subjects were recruited within a larger ongoing study involving sleep deprivation induced fatigue on neurocognitive function.” The kind of exhausting sleepless nights my niece is ongoing during her current studies attending as a medical resident in a New York City Hospital in her Oral Maxilla Facial Surgery training. “The military volunteers were tested twice during a two week period while they were rested. Each session was a morning time slot to control for circadian effects plus meshing within the typical morning schedules for the volunteers. The eligibility criteria included no prior history of head injury with loss of consciousness, no known neurological nor psychiatric disorders (including attention deficit hyperactivity disorder), no substance abuse problems including treatment, no overt visual or hearing problems. Men and women between 18-50 years of age with at least 12 years of education plus the ability to abstain caffeine use for at least 26 hours.”
“Prospective subjects were interviewed using standard respected self-report tests. This included the Conners Adult ADHD Rating Scale-Self-Report:Short Version, the Post Traumatic Stress Disorder (PTSD) Checklist Civilian Version and the Brain Injury Screening Questionnaire, (BISQ). Volunteers were excluded for any positive result for brain injury on the BISQ or achieving a sufficient t-score for the ADHD .”
The eye is the way into the brain. The eye was the first sense network in evolution. How the eye is positioned, pivoting and tracking a rich scene reveals itself as a exquisite way to view how possible concussion affects eye motion control, plus how vision involves visual intention and attention, all overlapping within the visual interpreting network zones inside our brain. The researchers employed a visual-eye tracking protocol using a apparatus capable of integrating visual stimulus presentation with eye tracking motion. The eye tracking device is EyeLink Cl from SR Research based in Ontario. The eye test involved tracking a red circular target moving at 25 degrees per second traveling in a circular trajectory of 10 degrees radius at 0.4 Hz against a black background. The testing sequence lasted approximately 5 minutes involving a practice run, then a calibration with two separate test runs which were recorded.
” Calibration of the eye position was conducted by having the subject fixate on a target presented at eight locations on the circular path of the test stimulus and one additional fixation point at the center of the circular path. The fixation target was presented at these nine locations in a randomized order. When an error was suspected or detected at any location, the target was presented there again. The calibration was validated by presenting the fixation target at the nine locations in a similar fashion. ‘
“The verbal instructions given to the volunteer were, ‘ follow the movement of the target as closely as possible.’ With both the practice and test runs, the target was presented at the central location to serve as a visual fixation prior to and following the circular movement of the target. Each of the two test runs consisted of six cycles of 15 seconds in duration per test run compared with just two cycles of the practice runs.”
“A custom MATLAB program analyzed the eye movements. Any blinks or other perturbations involving the pupil becoming occluded were identified within the computer program and excluded. To compensate for any potential artifact caused by unwanted head drifts relative to the camera during eye movement recording, the differences between the recorded gaze positions and the central fixation point presented before and after the circular target movement were calculated. The offset in the horizontal and vertical eye positions caused by a head drift was estimated with a linear interpolation between the pre- and post-run fixation differences and digitally subtracted from the data. In practice, however, the drift measured during each 15-s trial had an average of 0.50° in total visual angle with a standard deviation (SD) of 0.49°; thus, a correction would have been unnecessary in most cases.”
“To visualize gaze positional errors relative to the target motion, the target position was expressed in polar coordinates, and both the target and eye positions were rotated so that the target was at the 12 o’clock position. In this reference frame, the distance between the origin and the gaze represented the instantaneous radius of the gaze trajectory, and the angle formed by the vertical axis and the gaze vector represented the phase difference between the target and the gaze—that is, phase error. Positive phase error was defined as the gaze leading the target.
We quantified intra individual variability in visual tracking behavior using the SD of gaze positional errors relative to the target. The variability in the radial direction was measured with the SD of gaze errors perpendicular to the target trajectory, whereas the variability in the tangential direction was measured with the SD of gaze errors along the target trajectory. To facilitate comparison, the error variability measures were expressed in visual angle for both the radial and tangential directions.The radial error corresponds to the deviation in the radius of the gaze trajectory from the circular trajectory of the target, and the tangential error is proportional to the phase error.”
“Horizontal and vertical eye position data were two-point differentiated to obtain eye velocity, which was smoothed with a ten-point moving average filter. The signal was further differentiated to obtain eye acceleration, which was smoothed with a five-point moving average filter. Saccades were detected with velocity and acceleration thresholds of 100°/s and 1,500°/s2, respectively, and the saccade segments in the velocity data, which were expressed as sharp spikes, were replaced with straight lines connecting the ends of the remaining segments. The saccade detection thresholds took into consideration that saccades were generated during pursuit rather than fixation. Eye position and velocity traces were visually displayed by the analysis program, and the accuracy of saccade detection was verified.”
“To measure the level of accuracy in matching the eye velocity to the target velocity, smooth pursuit velocity gain was computed. The amplitudes of horizontal and vertical velocity modulations were obtained by fitting the de-saccaded velocity traces with sine curves of the frequency of the circular movement of the target, using fast Fourier transformation. The fitted traces were overlaid on the eye velocity traces in the software interface and visually matched with the smooth pursuit velocity modulations. Horizontal and vertical gains were the ratios between the amplitudes of the respective components of eye and target velocities.”
“To obtain a metric equivalent to the combination of horizontal and vertical smooth pursuit gain, phase error data were two-point differentiated and smoothed with a ten-point moving average filter. Instantaneous angular velocity gain was expressed as unity plus the ratio of phase error velocity to the constant angular velocity of the target. Average smooth pursuit angular velocity gain was then calculated by excluding saccade segments.
To measure the level of positional precision of visual tracking performance in horizontal and vertical directions, RMS positional deviations of the gaze from the target were calculated for the respective directions. The SDs of radial and tangential errors, mean phase error, angular smooth pursuit gain, and RMS errors were computed from the combination of the two test trials included in each test sequence. The horizontal and vertical gain values were computed for each trial and then averaged. The data segments from the first cycle of each test run were discounted from the analysis so that the transient response to the initial target movement was excluded.”
“Eye movement was recorded binocularly. A pilot analysis of the day 1 data with Pearson’s r calculated for the five visual-tracking parameters showed a high correlation between the left and the right eyes (range .90–.99). However, only monocular data were pooled for further analyses. The use of monocular data was based on the following rationale: Generally, small radial error variability provides an indication of spatial accuracy in the recorded data, since it combines the effects of a high level of performance by the subject and accurate eye position calibration. The eye-tracking equipment utilized in this study employed a single camera to record both eyes; thus, the spatial accuracy of eye position calibration in our data may have been compromised by the placement of the camera relative to each eye. To focus on the records that likely better represented the subject’s performance, the data from the eye with the smaller SD of radial errors were used for further analyses. This routine is justified because ocular dominance may have little relevance to the level of visual-tracking performance.”
I have quoted these authors extensively for the interested reader toward appreciating the detailed, rigorous approach toward the tested eye measurements. The authors remarked upon the highly predictable nature of the target movement, the actual visual tacking was imperfect between volunteers. Essentially the eyes are involved in predicting the circular path of the target anticipating its movement, the most divisive performance variability amounts to a measure of each volunteer’s state of attention toward determining their eye tracking pursuit during the test tracking. ” The use of circular target motion provided spatial and temporal information of visuomotor prediction. The standard measures included smooth pursuit velocity gain, phase error and RMS (root mean square) error, the variability of gaze positional error relative to the target was also measured. Quantifying performance variability is essential since a dysfunction in predictive timing should increase performance variability. Positional error variability is a useful index in concussion studies since TBI is known to increase intra-individual performance variability on visuomotor tasks.”
“Although our subject cohort was limited to healthy enlisted soldiers with similarities in age, training, and physical conditioning, the spatial and temporal accuracy of prediction varied among the subjects. However, the intra-individual test–retest measurements that were taken 2 weeks apart were strongly correlated. Such stability over time suggests that inter-individual variations in visual-tracking performance are based on neurological differences. These variations in visual-tracking performance should provide insight into the spectrum of cognitive functioning between individuals. Furthermore, a change in visual-tracking performance within an individual may indicate a change in the person’s neurological state.”
“Visual tracking was more accurate in the horizontal than in the vertical direction and points to separate mechanisms of control. Because little noise is introduced in the final motor pathways, the difference between horizontal and vertical accuracies cannot be wholly explained by a difference in the brain-stem motor nuclei. The eye muscle geometry, however may place a larger computational load for vertical control to conform to Listing’s Law during motor planning. Therefore it is possible that this larger computational load at the pre-motor stage contributes to decreased accuracy. The difference between horizontal and vertical tracking may also be generated at the level of visual processing, since there is a large contribution of sensory errors to the noise of the visuomotor response. ”
“Although there were differences in horizontal and vertical tracking, performance levels in the horizontal and vertical directions were parallel within individuals. Similar results have been demonstrated in clinical populations, including people diagnosed with schizophrenia and with bipolar disorder. Research on infants also shows interdependence between the development of horizontal and vertical visual tracking mechanisms. Taken together, these findings suggest a hierarchy of visuomotor processing and the existence of a high-level mechanism of control for horizontal and vertical visual tracking whereby computations are carried out in the two-dimensional visual space. This argument is consistent with the notion that visual tracking requires complex cognitive processes that are mediated by the cerebral cortex.”
“Evidence for the functional linking of vertical and horizontal tracking lends validity to our use of visual tracking parameters based on polar coordinates. These parameters are uniquely associated with circular tracking, as opposed to linear or more complex two-dimensional tracking. With a precise method of eye position recording, large variability in the instantaneous radius of gaze trajectory (radial error variability) must indicate instability in the subject’s spatial control, while large variability in the instantaneous angular phase (tangential error variability) must indicate a compound effect of instabilities in spatial and temporal control. Mean phase error, on the other hand, is an indicator of overall temporal accuracy.
In a highly predictable circular tracking task, tangential error variability and mean phase error point to the individual’s ability to sustain the state of synchronization between the external stimulus and the internally generated predictive drive.”
“We found that increases in phase lead, not lag, were associated with decreases in tracking accuracy assessed by gaze error variability, gain, and RMS errors. During tracking, the phase error was modulated with a sawtooth pattern, interposed by forward saccades. Poor tracking was characterized not by the mere presence of forward saccades but by the large and variable amplitudes of these saccades. Large forward saccades were anticipatory rather than corrective, landing as much as >10° of visual angle ahead of the target in some subjects. While catch-up saccades—that is, corrective forward saccades—compensate for phase lag, anticipatory saccades produce phase lead. Since forward saccades repeatedly occurred before the gaze lagged the target sufficiently to offset the lead, the presence of large anticipatory saccades was associated with a large mean phase lead. In our healthy subject cohort, we found no evidence for consistent positional errors that could serve as a threshold for initiating forward saccades during circular tracking. The saccades could not have been generated in reaction to the target image falling out of the foveal range, because the degrees of phase lag were generally smaller than those corresponding to the known range of latency for reactive saccades. Thus, forward saccades must be triggered by an internal mechanism. It is possible that instability is induced when a high smooth pursuit eye velocity is generated, which can be ameliorated by generating large forward saccades, leading to slower velocities and greater stability.”
“Another possible explanation lies in the mechanism of attention. Attention is or can readily be allocated ahead of a moving target during predictive visual tracking. Such attention allocation is usually covert in that the gaze is maintained on the target; that is, the urge to shift the gaze to the center of attention away from the target is suppressed. It is possible that anticipatory saccades are the results of a failure in the top-down suppression mechanism, analogous to errors in antisaccade paradigms wherein suppression of reflexive automatic pro-saccades is required. In congruence with this hypothesis, the role of the right prefrontal cortex has been implicated in predictive visual tracking, antisaccade performances and attentional control. Thus, a visual-tracking performance marked by excessive anticipatory saccades would suggest a neurologic dysfunction distinct from those marked by an increase in phase lag. Visual tracking of patients with chronic concussive syndrome (PCS) typically includes anticipatory saccades and phase lead and, consistent with the hallmark symptom of PCS, attention impairments.”
I’ll leave the final words for the authors of this most important concussion study. ” Predictive visual tracking shows promise as an attention metric to assess severity of minor traumatic brain injury. Deficits seen during predictive visual tracking correlate with Diffusion Tensor Imaging findings and with observed damage to neural pathways known to carry out cognitive and affective functions that are vulnerable to minor traumatic brain injury. The paradigm we have developed for testing subjects is currently under 5 min in duration for the entire test, which is markedly shorter compared to most neuropsychological tests.”