The Role of Oculomotor Function in the Assessment of Traumatic Brain Injury



By:  Marilyn F. Kraus




 The frontal cortices are particularly vulnerable to traumatic brain injury (TBI) across all severities of  injury. They are a common site of direct contusion. Also, diffuse axonal injury (DAI), which results in disruption of white matter tracts, can impair critical cortical-subcortical pathways, also resulting in frontal cortex dysfunction (1-3).

 Frontal-subcortical circuits are effector mechanisms that allow a person to act on the environment, modulating aspects of cognition, mood, and behavior. Dysfunction can result in a variety of disturbances in these functions.  Examples of mood and behavior changes include apathy/amotivation, impulsivity, irritability, aggression, depression, anxiety, and dysregulation of mood. Cognitive changes include decreased mental flexibility, trouble shifting sets, impaired attention, poor planning, lack of organization, problems with sequencing, impaired judgment, deficits in verbal fluency, problems with working memory, and impulsivity errors (4-6).


There are five major frontal-subcortical circuits  which are generally accepted (7).    


       Two motor circuits - one originates in the supplementary motor area, and an oculomotor circuit originating in the frontal eye field. 

       Three neurobehavioral circuits - dorsolateral prefrontal, orbital frontal (lateral and medial) and anterior cingulate (medial frontal) circuits

 A variety of neurobehavioral disorders, including types of dementia, Parkinson's, attention deficit disorder, and TBI involve disturbances of frontal-subcortical circuits. However, especially in milder severities of TBI, it can be difficult to find measures sensitive enough to detect dysfunction. There is a need for objective and quantifiable methodologies that can help better define neuropathology and resulting neurobehavioral deficits, particularly the more subtle problems that may be present with milder forms of TBI.  The potential clinical applicability would be to enhance diagnostic precision and ultimately assist with treatment planning, and assessing outcomes. Oculomotor function testing shows promise in this regard.


Oculomotor Function

 A saccade can be defined as a rapid intermittent eye movement, such as that which occurs when the eyes fix on one point after another in the visual field. The proper execution of saccades is dependent upon the integrity of regions including the frontal eye fields, dorsolateral prefrontal cortices, posterior parietal cortices, middle temporal region, supplementary motor regions, occipital lobes, thalamus, superior colliculus, cerebellum, and brain stem structures (8-10).     TBI can disrupt this network, producing functional deficits. Tasks designed to assess oculomotor function are one means of indirectly assessing neuronal damage/dysfunction.

 Oculomotor function is a general term, encompassing many aspects.  In this context, we are referring to specific oculomotor function testing, such as visually guided saccades, as well as more cognitively challenging tasks such as antisaccades or predictive saccades.  There are a number of different tasks that can be used, depending on the area of focus, including:

 Visually guided saccade (VGS).  This measure allows for an assessment of the basic sensorimotor system underlying all saccades (10).  The VGS task assesses the ability to quickly and accurately shift gaze toward targets presented unpredictably in the visual periphery. This type of saccade is considered more reflexive, as there is not much cognitive complexity to this function.  VGS is a reflexive saccade toward a suddenly appearing peripheral target.  These saccadic eye movements are completed quickly. The duration of a 10º saccade might typically be 45 msec.  Also, saccades are typically very accurate, bringing the eyes to within a fraction of a degree of the desired position.  The latency, or time it takes to mobilize the eyes to begin the saccade once the target is presented, is generally around 200 ms )  These  small values are why we often cannot detect subtle problems on the bedside neurological examination of eye movements.  Formal oculomotor testing allows for the accurate measurement of variables such as latencies and accuracy.

 Antisaccade (AS).  This task is a version of oculomotor testing that depends more on a cognitive component, as opposed to reflexive (as with the VGS).  In this task, the subject is instructed to look in the opposite direction of a visually presented stimulus. This is actually  harder than it sounds. Successful execution requires 2 things to occur. The individual must first inhibit a reflexive saccade toward a briefly appearing peripheral target and then reprogram and generate  a saccade to an equivalent point in the opposite hemifield.  If the individual fails to inhibit, and looks at the target, this is called a prosaccade error.  Neuroimaging research has provided strong support for the importance of prefrontal cortex and or/circuitry in the successful completion of this task (11-12).

Predictive Saccade (PRED). This is another modified version of a saccade task that  provides a measure of procedural learning that is also known to rely upon frontal striatal circuitry (13,14 ).  The PRED task typically involves a sequential presentation of 2 targets, each at a fixed location. This serial reaction time task rapidly induces procedural learning because, within 5 to 10 responses, subjects learn to “predict” target appearance and initiate saccades close to, if not anticipating, target appearance. This task has been shown in human neuroimaging studies to rely on regions within the dorsolateral prefrontal cortex, pre-supplementary motor area, anterior cingulate, hippocampus, mediodorsal thalamus, and striatum (13-14).


Oculomotor Function in TBI

 There have been several studies published to date using variations of oculomotor testing in TBI. Methodology has varied with regards to population characteristics (such as time out from injury and type of injury) and oculomotor methods (type of task and eye movement measures).

 Glass and colleagues (15) examined saccadic eye movements in response to pseudorandom and periodic stimuli in 9 patients with moderate and severe TBI and in healthy controls, but no mild TBI cases were studied. They reported reduced accuracy for the TBI group.   

Crevits et al (21) used AS and memory guided saccades to assess 25 acute mild TBI (MTBI) subjects within 24 hours of injury. Although they reported no statistically significant differences, they found trends for decreased accuracy and increased latency in MTBI. No assessment of neuropsychologic deficits was reported for either of these studies.

 Heitger et al (22) compared 30 MTBI subjects within 10 days of their injury with 30 controls on simple reflexive saccades, AS, and smooth pursuit. This acute group of MTBI subjects was found to have intact simple reflexive saccades, but prolonged AS latencies. They also showed an increased number of prosaccades. However, the MTBI group showed few deficits on neuropsychologic testing

 There is evidence that the PRED task is sensitive to MTBI (18, 20).   In a small sample of mixed chronic and subacute MTBI, Suh et al (18) demonstrated significant effects of TBI on accuracy of the PRED task.  Based on the small sample and period of recovery during testing, conclusions about the fronto-striatal deficits cannot be clearly delineated.

 In our lab we have used VGS, AS and PRED. In the first published study (19) we obtained oculomotor data and neuropsychological testing in chronic TBI  across all severities  using both VGS and AS tasks, comparing them to healthy controls.  The data showed that on the VGS task, which is a measure of basic attentional and sensorimotor function, only the moderate/severe TBI group (M/STBI)  showed significant impairment.  On the AS task, which is more reliant on prefrontal circuitry integrity, all severities of TBI showed impairments, and there were greater impairments seen on this task than on the VGS task. Prosaccade errors, which reflect a difficulty with voluntarily inhibiting responses, and prosaccade response latencies which reflect the ability to quickly disengage attention and voluntarily initiate behavior, were increased in all severities of TBI, including MTBI, compared to controls. Hence, data from the AS task may provide objective neurobehavioral evidence for a dysfunctional prefrontal systems, and appears to be sensitive to even a single MTBI.  In addition, all of the TBI subjects in this study were chronic.  Most previous studies of these tasks targeted acute TBI. The neuropsychologic testing results show a statistically significant difference between controls and the M/STBI group, but did not clearly differentiate MTBI  from the controls. So, taken alone, standard neuropsychologic testing may not always be sensitive to more subtle prefrontal deficits,  such as in MTBI.   MTBI appears to have more selectively prefrontal type impairments compared to more severe injuries. This could explain the differential sensitivities of various methods of assessments.   The data illustrate the sensitivity and specificity for differentiation of controls from MTBI , MTBI from M/STBI , and controls from M/STBI  for prosaccade error rates and for the executive function domain score.  The eye movement data are more sensitive and specific for differentiating MTBI from controls than the executive domain score.  However, for distinguishing MTBI from M/STBI, the findings are in the opposite direction. Namely, the executive domain score is more sensitive and specific than the eye movement data.  This differential ability to separate out milder TBI from controls, as opposed to separating MTBI and M/STBI, suggests that the neuropsychologic and oculomotor approaches may have different clinical utility. For MTBI, standardized neuropsychologic testing seems to have limited sensitivity for characterizing neurobehavioral deficits in many cases. In contrast, for separating MTBI and M/STBI, neuropsychologic data performed much better than the oculomotor data, suggesting that its psychometric sensitivity is better for detecting and scaling the severity of neurobehavioral deficits in M/STBI.

 Relationships were assessed between the oculomotor findings and cognitive domain scores. Across all TBI subjects, significant correlations were found between AS measures and executive, attention, and memory.

 In a second published study, we assessed the utility of the PRED task in 60 patients with a history of chronic TBI (at least 1 year from injury) of all severities (20).  The PRED data indicated that all severities of TBI result in slower procedural learning rates compared with the controls, and this impairment was proportional to the severity of injury. But, although TBI slows the rate of procedural learning, with greater severity of injury resulting in slower rates, it does not eliminate it. Importantly, the PRED was able to characterize a specific type of learning deficit in TBI via a neurophysiologic method, and hence adds another important dimension to our understanding of neuropathology in TBI across all severities, including MTBI.  A series of exploratory correlations were conducted between neuropsychological measures and oculomotor variables. No correlations were found for either the TBI patients or the controls between oculomotor variables (latencies, proportion of regular or anticipatory saccades) and any of the cognitive domains (memory, attention, and executive) in an exploratory analysis. This may be explained in part by a relative lack of sensitivity of many standard neuropsychologic testing batteries  for the assessment of procedural learning specifically, and could be a focus of further study.




The results of these oculomotor studies reflect the continuum of TBI neuropathology, and the effect of TBI on fronto-striatal circuitry. Neurophysiologic dysfunction logically follows that continuum, as the data demonstrate.  Oculomotor studies have utility in the assessment of all severities of chronic TBI, and this testing can document persistent deficits even with mild TBI.  Oculomotor assessment shows excellent promise as part of a multimodal assessment to better quantify and qualify deficits after TBI, with the potential to improve diagnostics as well as treatment planning.


Acknowledgements: Supported in part by NIH grant K23 MH068787 from the National Institute of Mental Health.



 1.  Levin HS, Williams DH, Eisenberg HM, et al. Serial magnetic resonance imaging and neurobehavioral findings after mild to moderate closed head injury. J Neurol Neurosurg Psychiatry. 1992; 55:255–262.
2.  Gennarelli TA, Thibault LE, Adams JH, et al. Diffuse axonal injury and traumatic coma in the primate. Ann Neurol. 1982;12: 564–574.
3. Kraus MF,  Susmaras T, Caughlin BP, Walker CJ,  Sweeney JA  Little DM, et al. (2007) White matter integrity and cognitive function in chronic traumatic brain injury: a diffusion tensor imaging study.  Brain 130: 2508-2519 
4.  Levin HS, Kraus MF. The frontal lobes and traumatic brain injury. J Neuropsychiatry Clin Neurosci. 1994;6:443–454.
5.  Miller E. The prefrontal cortex and cognitive control. Nat Neurosci Rev. 2000;1:59–65.
6.  Godefroy O. Frontal syndrome and disorders of executive functions. J Neurol. 2003;250:1–6.
7. Tekin S and Cummings JL. Frontal-subcortical neuronal circuits and clinical neuropsychiatry: an update.  J Psychosomatic Res 2002; 53:647 - 654. 
8. Becker W. The neurobiology of saccadic eye movements. Rev Oculomot Res Metrics. 1989;3:13–67.
9.  Berman R, Colby C, Genovese C, et al. Cortical networks subserving pursuit and saccadic eye movements in humans: an fMRI study. Hum Brain Mapping. 1999;8:209–225.
10.  Luna B, Thulborn K, Strojwas M, et al. Dorsal cortical regions subserving visually guided saccades in humans: an fMRI study. Cerebral Cortex. 1998;8:40–47.
11.  Munoz D, Everling S. Look away: the anti-saccade task and the voluntary control of eye movement. Nat Neurosci Rev. 2004; 5:218–228.
12.   Ploner C, Gaymard B, Rivaud-Pechoux S, et al. The prefrontal substrate of reflexive saccade inhibition in humans. Biol Psychiatry. 2005;57:1159–1165.
13.   Hubert V, Beaunieux H, Chételat G, et al. The dynamic network subserving the three phases of cognitive procedural learning. Hum Brain Mapping. 2007; 28:1415–1429.
14.   Simo L, Krisky C, Sweeney J. Functional neuroanatomy of anticipatory behavior: Dissociation beteen sensory driven and memory-driven systems. Cerebral Cortex. 2005; 15:1982–1991.
15.   Glass I, Groswasser Z, Groswasser-Reider I. Impersistent execution of saccadic eye movements after traumatic brain injury. Brain Injury. 1995;9:769–775.
16.   Crevits L, Hanse M, Tummers P, et al. Antisaccades and remembered saccades in mild traumatic brain injury. J Neurol. 2000; 247:179–182.
17.   Heitger M, Anderson T, Jones R, et al. Eye movement and visuomotor arm movement deficits following mild closed head injury. Brain. 2004;127:575–590. 18.   Suh M, Kolster R, Sarkar R, et al. Cognitive and neurobiological research consortium. deficits in predictive smooth pursuit after mild traumatic brain injury. Neurosci Lett. 2006; 401:108–113.
19.   Kraus MF, Little, DM, Donnell AJ, Reilly JL, Simonian N, Sweeney JA (2007). Oculomotor Function in Chronic Traumatic Brain Injury. Cog Behav Neurol 2007;20:170–178
20.   Kraus MF, Little DM, Wojtowicz SM, Sweeney JA. (2010)  Procedural Learning Impairments identified via Predictive Saccades in Chronic Traumatic Brain Injury.  Cogn Behav Neurol. 23(4): 210–217.