Classification of Comprehensive Neuro-Ophthalmologic Measures of Postacute Concussion | Traumatic Brain Injury | JAMA Network Open | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Comparison of Clinically Important Eye-Tracking Metrics in Concussed and Control Athletes
Comparison of Clinically Important Eye-Tracking Metrics in Concussed and Control Athletes

Truncated violin plots displaying the rotated kernel density for each of the 6 clinically important eye tracking metrics in participants with and without concussion, compared using unpaired t test with a Bonferonni-corrected P < .0012. Solid line indicates median; dashed lines, interquartile range.

Figure 2.  Vertical Saccades in a Representative Athlete With Concussion and Control Athlete
Vertical Saccades in a Representative Athlete With Concussion and Control Athlete

Although not meeting criteria as clinically important, 6 metrics related to vertical saccades had moderate discriminatory effect sizes (Cohen d, >0.50). The increased number of target misses can be seen in a representative athlete with concussion (B) compared with a representative control athlete (A). The saccadic pathways are also more irregular and less linear in the athlete with concussion, although this eye movement feature is not measured by the eye tracker.

Table 1.  Participant Characteristics
Participant Characteristics
Table 2.  Classification Accuracy of Clinically Important Eye Tracking Metrics
Classification Accuracy of Clinically Important Eye Tracking Metrics
Table 3.  Correlation Between Clinically Significant Eye-Tracking Metrics and Measures of Cognitive Performance in Participants with Concussion
Correlation Between Clinically Significant Eye-Tracking Metrics and Measures of Cognitive Performance in Participants with Concussion
1.
Langlois  JA, Rutland-Brown  W, Wald  MM.  The epidemiology and impact of traumatic brain injury: a brief overview.   J Head Trauma Rehabil. 2006;21(5):375-378. doi:10.1097/00001199-200609000-00001 PubMedGoogle ScholarCrossref
2.
Slobounov  S, Bazarian  J, Bigler  E,  et al.  Sports-related concussion: ongoing debate.   Br J Sports Med. 2014;48(2):75-76. doi:10.1136/bjsports-2013-092362PubMedGoogle ScholarCrossref
3.
Kroshus  E, Garnett  B, Hawrilenko  M, Baugh  CM, Calzo  JP.  Concussion under-reporting and pressure from coaches, teammates, fans, and parents.   Soc Sci Med. 2015;134(134):66-75. doi:10.1016/j.socscimed.2015.04.011 PubMedGoogle ScholarCrossref
4.
Elbin  RJ, Sufrinko  A, Schatz  P,  et al.  Removal from play after concussion and recovery time.   Pediatrics. 2016;138(3):e20160910. doi:10.1542/peds.2016-0910 PubMedGoogle Scholar
5.
Carson  JD, Lawrence  DW, Kraft  SA,  et al.  Premature return to play and return to learn after a sport-related concussion: physician’s chart review.   Can Fam Physician. 2014;60(6):e310-e315, e312-e315.PubMedGoogle Scholar
6.
Heitger  MH, Anderson  TJ, Jones  RD.  Saccade sequences as markers for cerebral dysfunction following mild closed head injury.   Prog Brain Res. 2002;140:433-448. doi:10.1016/S0079-6123(02)40067-2PubMedGoogle ScholarCrossref
7.
Drew  AS, Langan  J, Halterman  C, Osternig  LR, Chou  LS, van Donkelaar  P.  Attentional disengagement dysfunction following mTBI assessed with the gap saccade task.   Neurosci Lett. 2007;417(1):61-65. doi:10.1016/j.neulet.2007.02.038 PubMedGoogle ScholarCrossref
8.
Capó-Aponte  JE, Urosevich  TG, Temme  LA, Tarbett  AK, Sanghera  NK.  Visual dysfunctions and symptoms during the subacute stage of blast-induced mild traumatic brain injury.   Mil Med. 2012;177(7):804-813. doi:10.7205/MILMED-D-12-00061 PubMedGoogle ScholarCrossref
9.
Suh  M, Kolster  R, Sarkar  R, McCandliss  B, Ghajar  J; Cognitive and Neurobiological Research Consortium.  Deficits in predictive smooth pursuit after mild traumatic brain injury.   Neurosci Lett. 2006;401(1-2):108-113. doi:10.1016/j.neulet.2006.02.074 PubMedGoogle ScholarCrossref
10.
Maruta  J, Suh  M, Niogi  SN, Mukherjee  P, Ghajar  J.  Visual tracking synchronization as a metric for concussion screening.   J Head Trauma Rehabil. 2010;25(4):293-305. doi:10.1097/HTR.0b013e3181e67936 PubMedGoogle ScholarCrossref
11.
Thiagarajan  P, Ciuffreda  KJ, Ludlam  DP.  Vergence dysfunction in mild traumatic brain injury (mTBI): a review.   Ophthalmic Physiol Opt. 2011;31(5):456-468. doi:10.1111/j.1475-1313.2011.00831.x PubMedGoogle ScholarCrossref
12.
Gottshall  K, Drake  A, Gray  N, McDonald  E, Hoffer  ME.  Objective vestibular tests as outcome measures in head injury patients.   Laryngoscope. 2003;113(10):1746-1750. doi:10.1097/00005537-200310000-00016 PubMedGoogle ScholarCrossref
13.
Ventura  RE, Balcer  LJ, Galetta  SL.  The neuro-ophthalmology of head trauma.   Lancet Neurol. 2014;13(10):1006-1016. doi:10.1016/S1474-4422(14)70111-5 PubMedGoogle ScholarCrossref
14.
Nagib  S, Linens  SW.  Vestibular rehabilitation therapy improves perceived disability associated with dizziness postconcussion.   J Sport Rehabil. 2019;28(7):764-768. doi:10.1123/jsr.2018-0021 PubMedGoogle ScholarCrossref
15.
Ghajar  J, Ivry  RB; Cognitive and Neurobiological Research Consortium.  The predictive brain state: timing deficiency in traumatic brain injury?   Neurorehabil Neural Repair. 2008;22(3):217-227. doi:10.1177/1545968308315600 PubMedGoogle ScholarCrossref
16.
Barnes  GR.  Cognitive processes involved in smooth pursuit eye movements.   Brain Cogn. 2008;68(3):309-326. doi:10.1016/j.bandc.2008.08.020 PubMedGoogle ScholarCrossref
17.
Belanger  HG, Vanderploeg  RD.  The neuropsychological impact of sports-related concussion: a meta-analysis.   J Int Neuropsychol Soc. 2005;11(4):345-357. doi:10.1017/S1355617705050411 PubMedGoogle ScholarCrossref
18.
Snegireva  N, Derman  W, Patricios  J, Welman  KE.  Eye tracking technology in sports-related concussion: a systematic review and meta-analysis.   Physiol Meas. 2018;39(12):12TR01. doi:10.1088/1361-6579/aaef44 PubMedGoogle Scholar
19.
Ventura  RE, Jancuska  JM, Balcer  LJ, Galetta  SL.  Diagnostic tests for concussion: is vision part of the puzzle?   J Neuroophthalmol. 2015;35(1):73-81. doi:10.1097/WNO.0000000000000223 PubMedGoogle ScholarCrossref
20.
Samadani  U.  Will eye tracking change the way we diagnose and classify concussion and structural brain injury?   Concussion. 2015;1(1):CNC2. doi:10.2217/cnc.15.2PubMedGoogle Scholar
21.
Hunt  AW, Mah  K, Reed  N, Engel  L, Keightley  M.  Oculomotor-based vision assessment in mild traumatic brain injury: a systematic review.   J Head Trauma Rehabil. 2016;31(4):252-261. doi:10.1097/HTR.0000000000000174 PubMedGoogle ScholarCrossref
22.
Webb  B, Humphreys  D, Heath  M.  Oculomotor executive dysfunction during the early and later stages of sport-related concussion recovery.   J Neurotrauma. 2018;35(16):1874-1881. doi:10.1089/neu.2018.5673 PubMedGoogle ScholarCrossref
23.
Murray  NG, Szekely  B, Islas  A,  et al.  Smooth pursuit and saccades after sport-related concussion.   J Neurotrauma. 2020;37(2):340-346. doi:10.1089/neu.2019.6595 PubMedGoogle ScholarCrossref
24.
Cochrane  GD, Christy  JB, Almutairi  A, Busettini  C, Swanson  MW, Weise  KK.  Visuo-oculomotor function and reaction times in athletes with and without concussion.   Optom Vis Sci. 2019;96(4):256-265. doi:10.1097/OPX.0000000000001364PubMedGoogle ScholarCrossref
25.
Hunfalvay  M, Roberts  CM, Murray  N, Tyagi  A, Kelly  H, Bolte  T.  Horizontal and vertical self-paced saccades as a diagnostic marker of traumatic brain injury.   Concussion. 2019;4(1):CNC60. doi:10.2217/cnc-2019-0001 PubMedGoogle ScholarCrossref
26.
Kelly  KM, Kiderman  A, Akhavan  S,  et al.  Oculomotor, vestibular, and reaction time effects of sports-related concussion: video-oculography in assessing sports-related concussion.   J Head Trauma Rehabil. 2019;34(3):176-188. doi:10.1097/HTR.0000000000000437 PubMedGoogle ScholarCrossref
27.
Hunt  AWM, Mah  K, Reed  N, Engel  L, Keightley  M.  Oculomotor-based vision assessment in mild traumatic brain injury: a systematic review.   J Head Trauma Rehabil. 2016;31(4):252-261. doi:10.1097/HTR.0000000000000174 PubMedGoogle ScholarCrossref
28.
Voss  HU, McCandliss  BD, Ghajar  J, Suh  M; (CNRC-TBI).  A quantitative synchronization model for smooth pursuit target tracking.   Biol Cybern. 2007;96(3):309-322. doi:10.1007/s00422-006-0116-2 PubMedGoogle ScholarCrossref
29.
Samadani  U, Ritlop  R, Reyes  M,  et al.  Eye tracking detects disconjugate eye movements associated with structural traumatic brain injury and concussion.   J Neurotrauma. 2015;32(8):548-556. doi:10.1089/neu.2014.3687 PubMedGoogle ScholarCrossref
30.
Cifu  DX, Wares  JR, Hoke  KW, Wetzel  PA, Gitchel  G, Carne  W.  Differential eye movements in mild traumatic brain injury versus normal controls.   J Head Trauma Rehabil. 2015;30(1):21-28. doi:10.1097/HTR.0000000000000036 PubMedGoogle ScholarCrossref
31.
Maruta  J, Spielman  LA, Rajashekar  U, Ghajar  J.  Association of visual tracking metrics with post-concussion symptomatology.   Front Neurol. 2018;9:611. doi:10.3389/fneur.2018.00611 PubMedGoogle ScholarCrossref
32.
Hunfalvay  M, Roberts  CM, Murray  NP,  et al.  Vertical smooth pursuit as a diagnostic marker of traumatic brain injury.   Concussion. 2020;5(1):CNC69. doi:10.2217/cnc-2019-0013 PubMedGoogle ScholarCrossref
33.
Zaslow  T, Mueske  NM, Yu  GJ, Conrad-Forrest  A, Edison  B, Wren  TAL.  Oculomotor function in adolescent athletes following concussion.   Orthop J Sports Med. Published online April 20, 2020. doi:10.1177/2325967120S00150Google Scholar
34.
Samadani  U, Li  M, Qian  M,  et al.  Sensitivity and specificity of an eye movement tracking-based biomarker for concussion.   Concussion. 2015;1(1):CNC3.PubMedGoogle Scholar
35.
Bin Zahid  A, Hubbard  ME, Lockyer  J,  et al.  Eye tracking as a biomarker for concussion in children.   Clin J Sport Med. 2020;30(5):433-443. doi:10.1097/JSM.0000000000000639PubMedGoogle Scholar
36.
Kutcher  JS, Giza  CC.  Sports concussion diagnosis and management.   Continuum (Minneap Minn). 2014;20(6 Sports Neurology):1552-1569. doi:10.1212/01.CON.0000458974.78766.58PubMedGoogle Scholar
37.
US Centers for Disease Control and Prevention. Report to Congress on mild traumatic brain injury in the United States: steps to prevent a serious public health problem. Published September 2003. Accessed January 26, 2021. https://www.cdc.gov/traumaticbraininjury/pdf/mtbireport-a.pdf
38.
Murray  NP, Hunfalvay  M, Bolte  T.  The reliability, validity, and normative data of interpupillary distance and pupil diameter using eye-tracking technology.   Transl Vis Sci Technol. 2017;6(4):2. doi:10.1167/tvst.6.4.2 PubMedGoogle ScholarCrossref
39.
Murray  NP, Hunfalvay  M, Roberts  CM, Lange  B.  Reliability and normative data of computerized dynamic visual acuity tests.   Vision Dev Rehab. 2017;3(1):23-32.Google Scholar
40.
Iverson  GL, Lovell  MR, Collins  MW.  Validity of ImPACT for measuring processing speed following sports-related concussion.   J Clin Exp Neuropsychol. 2005;27(6):683-689. doi:10.1081/13803390490918435 PubMedGoogle ScholarCrossref
41.
Šimundić  A-M.  Measures of diagnostic accuracy: basic definitions.   EJIFCC. 2009;19(4):203-211.PubMedGoogle Scholar
42.
Lempke  LB, Howell  DR, Eckner  JT, Lynall  RC.  Examination of reaction time deficits following concussion: a systematic review and meta-analysis.   Sports Med. 2020;50(7):1341-1359. doi:10.1007/s40279-020-01281-0 PubMedGoogle ScholarCrossref
43.
Liversedge  SP, Findlay  JM.  Saccadic eye movements and cognition.   Trends Cogn Sci. 2000;4(1):6-14. doi:10.1016/S1364-6613(99)01418-7 PubMedGoogle ScholarCrossref
44.
Fukushima  K, Fukushima  J, Warabi  T, Barnes  GR.  Cognitive processes involved in smooth pursuit eye movements: behavioral evidence, neural substrate and clinical correlation.   Front Syst Neurosci. 2013;7:4. doi:10.3389/fnsys.2013.00004 PubMedGoogle ScholarCrossref
45.
Mullen  SJ, Yücel  YH, Cusimano  M, Schweizer  TA, Oentoro  A, Gupta  N.  Saccadic eye movements in mild traumatic brain injury: a pilot study.   Can J Neurol Sci. 2014;41(1):58-65. doi:10.1017/S0317167100016279 PubMedGoogle ScholarCrossref
46.
Wasserman  EB, Abar  B, Shah  MN, Wasserman  D, Bazarian  JJ.  Concussions are associated with decreased batting performance among Major League Baseball players.   Am J Sports Med. 2015;43(5):1127-1133. doi:10.1177/0363546515576130 PubMedGoogle ScholarCrossref
47.
Nordström  A, Nordström  P, Ekstrand  J.  Sports-related concussion increases the risk of subsequent injury by about 50% in elite male football players.   Br J Sports Med. 2014;48(19):1447-1450. doi:10.1136/bjsports-2013-093406 PubMedGoogle ScholarCrossref
48.
Guskiewicz  KM, McCrea  M, Marshall  SW,  et al.  Cumulative effects associated with recurrent concussion in collegiate football players: the NCAA Concussion Study.   JAMA. 2003;290(19):2549-2555. doi:10.1001/jama.290.19.2549 PubMedGoogle ScholarCrossref
49.
Zemper  ED.  Two-year prospective study of relative risk of a second cerebral concussion.   Am J Phys Med Rehabil. 2003;82(9):653-659. doi:10.1097/01.PHM.0000083666.74494.BA PubMedGoogle ScholarCrossref
50.
Zonner  SW, Ejima  K, Fulgar  CC,  et al.  Oculomotor response to cumulative subconcussive head impacts in US high school football players: a pilot longitudinal study.   JAMA Ophthalmol. 2019;137(3):265-270. doi:10.1001/jamaophthalmol.2018.6193 PubMedGoogle ScholarCrossref
51.
Kawata  K, Rubin  LH, Lee  JH,  et al.  Association of football subconcussive head impacts with ocular near point of convergence.   JAMA Ophthalmol. 2016;134(7):763-769. doi:10.1001/jamaophthalmol.2016.1085PubMedGoogle ScholarCrossref
52.
Kellar  D, Newman  S, Pestilli  F, Cheng  H, Port  NL.  Comparing fMRI activation during smooth pursuit eye movements among contact sport athletes, non-contact sport athletes, and non-athletes.   Neuroimage Clin. 2018;18:413-424. doi:10.1016/j.nicl.2018.01.025 PubMedGoogle ScholarCrossref
53.
Mack  DJ, Belfanti  S, Schwarz  U.  The effect of sampling rate and lowpass filters on saccades—a modeling approach.   Behav Res Methods. 2017;49(6):2146-2162. doi:10.3758/s13428-016-0848-4 PubMedGoogle ScholarCrossref
54.
Lange  B, Hunfalvay  M, Murray  N, Roberts  CM, Bolte  T.  Reliability of computerized eye-tracking reaction time tests in non-athletes, athletes, and individuals with traumatic brain injury.   Optom Vis Performance. 2018;6(3):167-182.Google Scholar
55.
Munoz  DP, Broughton  JR, Goldring  JE, Armstrong  IT.  Age-related performance of human subjects on saccadic eye movement tasks.   Exp Brain Res. 1998;121(4):391-400. doi:10.1007/s002210050473 PubMedGoogle ScholarCrossref
56.
Lapidot  MB.  Does the brain age uniformly? evidence from effects of smooth pursuit eye movements on verbal and visual tasks.   J Gerontol. 1987;42(3):329-331. doi:10.1093/geronj/42.3.329 PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Neurology
    March 3, 2021

    Classification of Comprehensive Neuro-Ophthalmologic Measures of Postacute Concussion

    Author Affiliations
    • 1University of Rochester School of Medicine and Dentistry, Rochester, New York
    • 2Medical College of Wisconsin, Milwaukee
    • 3University of Rochester, Rochester, New York
    • 4Department of Pediatrics, Boston Children’s Hospital, Harvard Medical School, Boston, Massachusetts
    • 5Department of Emergency Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York
    • 6Department of Kinesiology, Indiana University, Bloomington
    JAMA Netw Open. 2021;4(3):e210599. doi:10.1001/jamanetworkopen.2021.0599
    Key Points

    Question  Which computerized eye-tracking metrics most accurately differentiate athletes with and without concussion after acute injury?

    Findings  In this cohort study of 34 college-age athletes with concussion and 54 uninjured athletes, 6 eye-tracking measures that reflect slowed visual reaction time had a combined sensitivity of 77.8% and specificity of 92.6% for discriminating between the groups.

    Meaning  These findings suggest that accurate, objective measures of slowed visual reaction time have the potential to improve concussion recognition and reduce the disability associated with underdiagnosis.

    Abstract

    Importance  Symptom-based methods of concussion diagnosis in contact sports result in underdiagnosis and repeated head injury exposure, increasing the risk of long-term disability. Measures of neuro-ophthalmologic (NO) function have the potential to serve as objective aids, but their diagnostic utility is unknown.

    Objective  To identify NO measures that accurately differentiate athletes with and without concussion.

    Design, Setting, and Participants  This cohort study was conducted among athletes with and without concussion who were aged 17 to 22 years between 2016 and 2017. Eye movements and cognitive function were measured a median of 19 days after injury among patients who had an injury meeting the study definition of concussion while playing a sport (retrospectively selected from a concussion clinic), then compared with a control group of participants without concussion (enrolled from 104 noncontact collegiate athlete volunteers without prior head injury). Data analysis was conducted from November 2019 through May 2020.

    Exposure  Concussion.

    Main Outcomes and Measures  Classification accuracy of clinically important discriminator eye-tracking (ET) metrics. Participants’ eye movements were evaluated with a 12-minute ET procedure, yielding 42 metrics related to smooth pursuit eye movement (SPEM), saccades, dynamic visual acuity, and reaction time. Clinically important discriminator metrics were defined as those with significantly different group differences and area under the receiver operator characteristic curves (AUROCs) of at least 0.70.

    Results  A total of 34 participants with concussions (mean [SD] age, 19.7 [2.4] years; 20 [63%] men) and 54 participants without concussions (mean [SD] age, 20.8 [2.2] years; 31 [57%] men) completed the study. Six ET metrics (ie, simple reaction time, discriminate reaction time, discriminate visual reaction speed, choice visual reaction speed, and reaction time on 2 measures of dynamic visual acuity 2) were found to be clinically important; all were measures of reaction time, and none were related to SPEM. Combined, these 6 metrics had an AUROC of 0.90 (95% CI, 0.80-0.99), a sensitivity of 77.8%, and a specificity of 92.6%. The 6 metrics remained significant on sensitivity testing.

    Conclusions and Relevance  In this study, ET measures of slowed visual reaction time had high classification accuracy for concussion. Accurate, objective measures of NO function have the potential to improve concussion recognition and reduce the disability associated with underdiagnosis.

    Introduction

    It is estimated that 1.6 to 3.8 million sports-related concussions occur annually in the United States.1 Accurate diagnosis is important for managing safe return to play (RTP) and academic activities. However, the current symptom-based method of concussion diagnosis has been criticized for being subjective,2 allowing injured athletes to downplay symptoms to avoid being removed from play.3 This can result in exposure to another head injury before recovery is complete, resulting in prolonged recovery.4,5 An objective diagnostic aid has the potential to improve concussion recognition and reduce or prevent disability associated with premature RTP.

    Quantitative assessment of eye movements with eye tracking (ET) may be a potential solution. Concussion has been associated with abnormalities in all 4 basic eye movements, including saccades,6-8 smooth pursuit eye movement (SPEM),9,10 vergence movements,11 and vestibulo-ocular movements.12 These eye movements are executed by neural networks involving the midbrain, cerebellum, frontal eye fields, visual cortex, prefrontal cortex, and vestibular nuclear complex.13 Axonal injury in these networks is thought to underlie postconcussive imbalance, dizziness, and vision blurriness, especially when tracking objects in motion.10,14 Some eye movements (eg, predictive smooth pursuits) rely on intact working memory and attention to generate accurate prediction of future events, which is important for tracking moving objects.13,15 Thus, postconcussion impairments in cognitive processing may exacerbate eye movement performance.16,17 The associations between ET and cognitive processing are referred to as neuro-ophthalmologic (NO) function.13 Evaluation of NO response has been proposed as a way to detect neurologic injury after a head impact and aid in the diagnosis of concussion.18-20

    Computerized ET can provide objective measures of SPEM (slow eye movements used to track a moving target) and saccades (rapid movements used to look from 1 target to another).9,10,18,21-33 SPEM metrics have been identified as accurate concussion classifiers,29,34,35 but the discriminatory accuracy of non-SPEMs and the contribution of attention and memory to eye movement performance have not, to our knowledge, been evaluated. Addressing these gaps would be crucial for identifying the ET metrics most likely to provide added value in the evaluation of concussion.

    The objective of the current study was to identify ET metrics that most accurately distinguished athletes with and without concussion. Based on prior published work suggesting the potential clinical utility of SPEM, we hypothesized that reduced performance in these metrics would be the most accurate injury status classifiers. The secondary objective was to determine the contribution of attention and working memory to classifying ET metrics through correlation to validated measures of cognition. We hypothesized that reduced performance in classifying ET metrics after concussion would be associated with poorer visual memory or slower visual motor speed.

    Methods
    Participants and Setting

    In this cohort study, eye movements and cognitive function were measured after the acute injury period (median, 19 days; range 2-457 days) in a group of athletes with concussion and compared with results from a control group without concussion. The postacute period was defined as more than 1 day after injury. Participants without concussion were accrued between January and March 2017 from a volunteer sample of 104 noncontact National Collegiate Athletic Association Division III athletes without a prior head injury history. Noncontact sports included tennis, golf, swimming, crew/rowing, and squash. Participants with concussion were retrospectively selected from 101 new patients presenting to a university sport concussion clinic for ongoing symptoms who underwent ET as part of a standardized clinical intake process between September 2016 and June 2017. To provide matching to the control group based on age and mechanism, participants aged 17 to 22 years injured playing a sport were selected from 101 patients with ET data. Study participants provided informed consent, and the research protocol was approved by the University of Rochester review board. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

    Concussion diagnostic criteria consisted of a head injury associated with either (1) witnessed or self-reported brief loss of consciousness, amnesia, or confusion or (2) onset within 24 hours of injury of headache, short-term memory concerns, dizziness, or imbalance.36,37 Concussion diagnosis was made by an experienced clinician using these criteria. Participants with traumatic intracranial lesions on head computed tomography scan, who were hospitalized for more than24 hours, or in whom an alternative diagnosis was suspected were excluded. All participants were excluded for history of strabismus. Participants requiring corrective lenses were not excluded but were required to wear them during the ET procedure. Observers were not masked to patient grouping.

    ET Procedure

    Eye movements were evaluated with a commercially available eye tracker (RightEye) consisting of a 24-inch 3-dimensional vision monitor fitted with a System Management Interface 12-inch remote eye tracker (NVIDIA) connected to a gaming system (Alienware) and a wireless keyboard and mouse (Logitech).38 The eye tracker has a 120 Hz temporal resolution and an accuracy of 0.4°. The testing procedure lasted 12 minutes and produced 42 variables related to SPEM, saccades, dynamic visual acuity, and reaction time (eTable 1 in the Supplement). The testing parameters were factory preset and not modifiable. All participants underwent ET in a seated position after completing a 9-point calibration requiring fixation on at least 8 of 9 sequentially displayed points on the screen.

    For circular smooth pursuits, participants tracked a black target of 0.2° diameter on a white background, which followed a circular trajectory of 10° radius at a rate of 25° per second (0.4 Hz) for 6.5 cycles (15 seconds). For horizontal smooth pursuits, participants tracked a dot (same size and speed as the circular smooth pursuit test) moving back and forth in a straight line horizontally across the screen for 25 seconds.25 To be on target, the participant was required to follow the stimulus within 2.4°. The protocol for vertical smooth pursuit was identical but in a vertical plane.25 The eye tracker measures reaction time in 3 dedicated tests (simple reaction time, choice reaction time, and discriminate reaction time) and as part of dynamic visual acuity testing.39 The eAppendix and eTable 2 in the Supplement include descriptions of these methods as well as those related to saccades.39

    Cognitive Testing

    Cognitive performance was assessed using a software program (Immediate Post-Concussion Assessment and Cognitive Test [ImPACT Applications]) measuring total postconcussive symptom score and 4 domains of memory and attention (ie, verbal memory, visual memory, visual motor speed, and reaction time).40 Cognitive testing was administered immediately prior to ET in participants with concussion (eAppendix in the Supplement).

    Statistical Analysis

    Important discriminator metrics were identified by comparing group differences in the 42 available metrics using the Mann-Whitney U test for variables with skew values greater than an absolute value of 2.0 and an unpaired t test for all others. Bonferonni correction was applied to the a priori 1-tailed P value indicating significance, moving the threshold from P < .05 to P < .0012. Area under the receiver operator characteristic curve (AUROC) was calculated for individual metrics in which there was a statistically significant difference between groups; those that also had an AUROC of at least 0.70 were considered clinically important.41 ET metrics that did not have an AUROC of at least 0.70 but nonetheless had moderate to large effect sizes (ie, Cohen d, >0.50) were considered potentially important. A multivariable logistic regression model was used to determine the AUROC combining all clinically important metrics and the Youden J index to determine the optimal probability cutoff for discriminating between groups. The Youden J index assigns equal weights to the misclassification of events by identifying the probability cutoff that maximizes the difference between true-positives and false-positives.41 This cutoff was used to determine the sensitivity and specificity of the model.

    Because missing data were confined to the concussed group (eTable 1 in the Supplement), they were considered not missing at random and neither substituted nor imputed. Participants with missing data were included in the analysis, and each ET metric was analyzed with the available data. A sensitivity analysis using only participants with complete data was performed to estimate potential bias in results derived from participants with missing data.

    Clinically important ET metrics were correlated with 4 cognitive domains and total symptom score in participants with concussions using the Spearman correlation coefficient. Similarly, the Spearman correlation coefficient was used to determine the association between the injury-ET time interval and values for each clinically important metric in participants with concussion. Statistical analyses were performed using SAS software version 9.4 (SAS Institute).

    Results
    Participant Characteristics

    Thirty-four participants with concussion (mean [SD] age, 19.7 [2.4] years; 20 [63%] men) and 54 uninjured athletes (mean [SD] age, 20.8 [2.2] years; 31 [57%] men) completed the study. Of 101 available patients with concussion and ET data, 69 (69%) were either outside the age criteria or not injured during a sport, resulting in 32 (31%) participants with concussion. Of 104 healthy athlete volunteers, 50 (48%) were either ineligible or declined to participate, yielding 54 participants (52%) for the nonconcussed control group. Participants with concussion were most frequently injured playing soccer (6 [19%]), followed by football (7 [22%]) and basketball (4 [13%]) (Table 1). Compared with the control group, participants with concussion were younger and more likely to be White individuals (33 [61%] vs 28 [88%]). There was no significant group difference in the proportion of participants with attention-deficit/hyperactivity disorder. In participants with concussion, the median (interquartile range [IQR]) time between injury and ET procedure was 19 (10-39) days. Although the injury-testing interval ranged from 2 to 457 days, only 1 participant was tested after 58 days. Total symptom scores at the time of assessment ranged from 0 to 68 (median [IQR], 18 [4-39]). The 3 participants who reported a symptom score of 0 on cognitive testing had headache and/or dizziness with cognitive exertion. Because of ET software upgrades occurring during the course of study, a number of concussed participants (range, 1-13) were missing 1 or more reaction time variables (eTable 1 in the Supplement). In addition, cognitive test results were missing in 3 participants with concussion.

    ET Metrics in Participants With and Without Concussion

    The distributions of 8 of the 42 ET metrics (19%) were significantly different between participants with and without concussion after Bonferonni correction. Of these 8 ET metrics, 6 (75%) had an AUROC of at least 0.70 and were thus considered clinically important; all were measures of reaction time and none were related to SPEM (Table 2). They were simple reaction time, discriminate reaction time, discriminate visual reaction speed, choice visual reaction speed, and reaction time on 2 measures of dynamic visual acuity. Discriminatory accuracy ranged from an AUROC of 0.72 (95% CI, 0.61-0.84) for the reaction time metric in the dynamic visual acuity 2 test to an AUROC of 0.81 (95% CI, 0.70-0.93) for simple reaction time. All 6 clinically important measures of reaction time were slower among participants with concussion (Figure 1). For example, median (IQR) simple reaction time among participants with concussion was significantly slower than among participants without concussion (464.9 [387.9-502.7] milliseconds vs 373.5 [332.9-402.3] milliseconds). Despite the broad range in postinjury test assessment, none of these measures correlated with the injury-ET time interval (eTable 3 in the Supplement). The combination of these 6 clinically important metrics had an AUROC of 0.90 (95% CI, 0.80, 0.99) (Table 2). The Youden J Index identified 0.37 as an optimal probability cutoff, which yielded a sensitivity of 77.8% and specificity of 92.6%.

    A sensitivity analysis of participants with complete data for all ET metrics showed no change in statistical significance of the AUROC for the 6 clinically important metrics. There were small changes in AUROC point estimates that ranged from −0.02 to 0.05 (eg, choice reaction time, visual reaction speed, main analysis: AUROC, 0.76; 95% CI, 0.60-07.91; sensitivity analysis: AUROC, 0.74; 95% CI, 0.48-0.90; difference, −0.02; dynamic visual acuity 3, reaction time, main analysis: AUROC, 0.74; 95% CI, 0.62-0.86; sensitivity analysis: AUROC, 0.79; 95% CI, 0.58-0.90; difference, 0.05).

    Although 2 of the 8 variables (25%) with significantly different group means did not have an AUROC of at least 0.70, they nonetheless had large effect sizes (Cohen d, >0.80). An additional 8 variables with statistically nonsignificant group differences had moderate effect sizes (Cohen d, >0.50); 6 (75%) of these involved vertical saccadic eye movements (Figure 2). These 10 potentially important variables are summarized in eTable 4 in the Supplement.

    Correlation of ET Variables With Currently Used Cognitive Performance Measures

    Among participants with concussion, 3 of the 6 clinically important ET metrics (50%) had significant correlations to cognitive test measures of reaction time (Table 3). Simple reaction time also correlated with total symptom score. The remaining 3 clinically important ET metrics, including 2 measuring visual reaction speed, were not correlated with cognitive test–measured reaction time. For all 6 metrics, correlations to verbal memory, visual memory, and visual motor speed scores were weak and statistically nonsignificant.

    Discussion

    Contrary to our primary hypothesis, ET measures of reaction time, not SPEM, were the most accurate classifiers of athletes with and without concussion. Of the 6 most discriminatory, clinically important ET measures, all involved reaction time. Participants with concussion had significantly slower reaction time and impaired ability to see a fast-moving target during a dynamic visual acuity task compared with a control group without concussion. Participants with concussion also had slower reaction time during simple and discriminate reaction time tests as well as slower visual reaction speed during choice and discriminate reaction time tests. Moreover, slower simple reaction time was associated with higher total post concussive symptom scores. Individually, the AUROC for these 6 ET metrics ranged from 0.72 to 0.81, while their combined AUROC was 0.90.

    Our results indicate that concussion poses a significant challenge to responding to a rapidly appearing visual target and that the components of visual reaction time across multiple tests may offer an objective, clinical tool to diagnose concussion after the acute postinjury phase. This information may be especially useful for office-based health care professionals who evaluate patients with concussion days to weeks after injury. Pending confirmation in an independent cohort, visual reaction time measures could be used as a diagnostic aid when symptom minimization is suspected or when the clinical picture is otherwise unclear.

    Although others have reported postconcussive slowing in visual reaction time (often simply called reaction time),42 our finding of slowed visual reaction speed is novel. As the participant fixes on a central target, visual reaction time measures the time elapsed between presentation of a peripheral target and a manual motor response, such as pressing a mouse button or computer key. However, this single measure involves 3 separate processes: saccadic latency, visual reaction speed, and processing speed. Visual reaction speed (also called saccadic velocity) measures the time required for the eyes to travel from the central target to the peripheral target. Unlike overall visual reaction time, visual reaction speed does not involve a manual motor response and can only be measured by a device that monitors eye movements, such as an eye tracker. In our study, the finding that neither ET measure of visual reaction speed correlated with reaction time on cognitive testing suggests that this metric may be measuring aspects of visual reaction time unrelated to cognitive processes and more specific to disruption in oculomotor pathways, potentially providing unique clinical information for identifying concussion and determining the optimal timing of RTP. While slowed visual reaction speed after concussion has been reported by at least 1 other author,26 anatomic associations were not investigated nor was its value as a classifier or outcome predictor.

    Contrary to our secondary hypothesis, performance on the 6 clinically important classifying metrics was not correlated with visual memory or visual motor speed. There are several potential explanations for this. First, our secondary hypothesis was linked to the underlying assumption that metrics related to SPEM would be the most important injury classifies, not reaction time measures. Prior research indicated that variation in performance of SPEM and saccadic eye movements was associated with variation in cognitive processes, such as working memory and visual motor speed.43,44 Based on our results, these cognitive processes may have less of an influence on visual reaction time measures. Furthermore, our inability to identify a significant correlation to cognitive processes may be owing to relatively small sample size compounded by missing visual reaction time and/or cognitive test measures in several participants. However, the finding that visual reaction time variables on ET correlated fairly robustly to reaction time scores on cognitive testing suggests that the sample size was adequate.

    In addition to identifying 6 clinically important ET metrics, we identified an additional 10 potentially important metrics, most (6 of 10) related to vertical saccades. Participants with concussion displayed more target misses, fewer fixations, and fewer overall saccadic eye movements compared with their counterparts without concussion. Several authors have reported significant postconcussive deficits in vertical saccades.23,25,30,45 It is unclear why vertical saccade metrics were not statistically significant classifiers in our study. Small sample size combined with the relatively conservative Bonferonni method of correcting for multiple comparisons may have obscured a statistically significant group difference.

    While promising, additional research is needed to confirm our findings in an independent cohort and to determine the extent to which visual reaction time metrics improve at the time of RTP. Persistence of slowed visual reaction time after RTP could offer biologic insight into decrements in athletic performance and increases in injuries observed after concussed athletes RTP.46-49 In addition, future studies on the ability of ET metrics to predict prolonged (ie, >3 months) concussion recovery would provide clinicians and researchers with a useful risk stratification tool. Finally, research to identify structural correlates of slowed visual reaction time using advanced neuroimaging, such as diffusion tensor imaging, would be important for confirming a causal relationship to concussion.

    Limitations

    This study has several limitations. Unlike participants in the control group, those with concussion may have been exposed to repetitive head impacts (RHIs) which can affect oculomotor function and bias our results away from the null. Although RHIs have been associated with acute and transient worsening in near-point convergence,50,51 they have not been associated with significant changes in ET metrics.52 Athletes in the control group who played sports requiring advanced hand-eye coordination skills, such as tennis and squash, may have had better baseline visual reaction time than athletes with concussion who primarily played football, potentially accentuating group differences in ET measures. The low temporal resolution of the ET device (120 Hz) compared with laboratory-based research devices may have prevented detection of small reaction time differences. However, the temporal sampling error for a 120 Hz device is estimated to be approximately 4 milliseconds,53 well below the observed group differences in reaction time measures, which ranged from 23.277 to 185.018 milliseconds (eTable 1 in the Supplement). Moreover, the reliability of reaction time measures on the device used in our study was quite robust among noninjured young adults, with most intraclass correlation coefficients above 0.75.54 Because visual reaction time and other eye movement performance metrics are faster and more accurate in athletes than nonathletes23 and decline with advancing age,54-56 our results may not be generalizable to older or nonathlete populations. Furthermore, although observers were not masked to patient grouping, neither ET nor cognitive testing are subjective measures, likely limiting the effect of observer bias.

    Conclusions

    Contrary to our primary hypothesis, our results indicated that ET measures of visual reaction time, but not SPEM, were the best NO classifiers of athletes with and without concussion. Six ET measures reflecting slowed visual reaction time had a sensitivity of 77.8% and specificity of 92.6% for discriminating between the groups. None of these 6 metrics correlated with visual memory or visual motor speed, and 3 of the 6 did not correlate to reaction time on cognitive testing. Thus, not only were these metrics highly discriminatory, they may be measuring aspects of visual reaction time unrelated to cognitive processes and more specific to disruption in oculomotor pathways, potentially providing unique clinical information for identifying and managing a concussion. Providing objective measurements of visual reaction time may improve concussion diagnosis and management and reduce the disability associated with premature return to play.

    Back to top
    Article Information

    Accepted for Publication: January 13, 2021.

    Published: March 3, 2021. doi:10.1001/jamanetworkopen.2021.0599

    Correction: This article was corrected on March 31, 2021, to fix an error in the byline.

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Feller CN et al. JAMA Network Open.

    Corresponding Author: Jeffrey J. Bazarian, MD, MPH, Department of Emergency Medicine, University of Rochester School of Medicine and Dentistry, 265 Crittenden Blvd, Box 655C, Rochester, NY 14642 (jeff_bazarian@urmc.rochester.edu).

    Author Contributions: Drs Bazarian and Abar had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Dr Feller and Ms Goldenberg are co–first authors.

    Concept and design: Feller, Goldenberg, Asselin, Merchant-Borna, Mannix, Bazarian.

    Acquisition, analysis, or interpretation of data: Feller, Goldenberg, Asselin, Merchant-Borna, Abar, Jones, Kawata, Bazarian.

    Drafting of the manuscript: Feller, Goldenberg, Merchant-Borna, Abar, Kawata, Bazarian.

    Critical revision of the manuscript for important intellectual content: Goldenberg, Asselin, Merchant-Borna, Jones, Mannix, Kawata, Bazarian.

    Statistical analysis: Feller, Goldenberg, Merchant-Borna, Abar, Jones, Bazarian.

    Obtained funding: Merchant-Borna, Bazarian.

    Administrative, technical, or material support: Goldenberg, Merchant-Borna.

    Supervision: Asselin, Merchant-Borna, Bazarian.

    Conflict of Interest Disclosures: Dr Bazarian reported receiving grants from BrainScope and receiving personal fees from Abbott Point of Care Diagnostics and Q30 Innovations outside the submitted work. No other disclosures were reported.

    Meeting Presentation: This study was presented at the 3rd Joint Symposium of the International and National Neurotrauma Societies and American Association of Neurological Surgeons/Congress of Neurological Surgeons Section on Neurotrauma and Critical Care; August 13, 2018; Toronto, Ontario, Canada.

    Additional Contributions: The electronic database was created and maintained by Kimberly Kaukeinen, BA, (Department of Biostatistics, University of Rochester). She received no compensation beyond her salary for this contribution.

    References
    1.
    Langlois  JA, Rutland-Brown  W, Wald  MM.  The epidemiology and impact of traumatic brain injury: a brief overview.   J Head Trauma Rehabil. 2006;21(5):375-378. doi:10.1097/00001199-200609000-00001 PubMedGoogle ScholarCrossref
    2.
    Slobounov  S, Bazarian  J, Bigler  E,  et al.  Sports-related concussion: ongoing debate.   Br J Sports Med. 2014;48(2):75-76. doi:10.1136/bjsports-2013-092362PubMedGoogle ScholarCrossref
    3.
    Kroshus  E, Garnett  B, Hawrilenko  M, Baugh  CM, Calzo  JP.  Concussion under-reporting and pressure from coaches, teammates, fans, and parents.   Soc Sci Med. 2015;134(134):66-75. doi:10.1016/j.socscimed.2015.04.011 PubMedGoogle ScholarCrossref
    4.
    Elbin  RJ, Sufrinko  A, Schatz  P,  et al.  Removal from play after concussion and recovery time.   Pediatrics. 2016;138(3):e20160910. doi:10.1542/peds.2016-0910 PubMedGoogle Scholar
    5.
    Carson  JD, Lawrence  DW, Kraft  SA,  et al.  Premature return to play and return to learn after a sport-related concussion: physician’s chart review.   Can Fam Physician. 2014;60(6):e310-e315, e312-e315.PubMedGoogle Scholar
    6.
    Heitger  MH, Anderson  TJ, Jones  RD.  Saccade sequences as markers for cerebral dysfunction following mild closed head injury.   Prog Brain Res. 2002;140:433-448. doi:10.1016/S0079-6123(02)40067-2PubMedGoogle ScholarCrossref
    7.
    Drew  AS, Langan  J, Halterman  C, Osternig  LR, Chou  LS, van Donkelaar  P.  Attentional disengagement dysfunction following mTBI assessed with the gap saccade task.   Neurosci Lett. 2007;417(1):61-65. doi:10.1016/j.neulet.2007.02.038 PubMedGoogle ScholarCrossref
    8.
    Capó-Aponte  JE, Urosevich  TG, Temme  LA, Tarbett  AK, Sanghera  NK.  Visual dysfunctions and symptoms during the subacute stage of blast-induced mild traumatic brain injury.   Mil Med. 2012;177(7):804-813. doi:10.7205/MILMED-D-12-00061 PubMedGoogle ScholarCrossref
    9.
    Suh  M, Kolster  R, Sarkar  R, McCandliss  B, Ghajar  J; Cognitive and Neurobiological Research Consortium.  Deficits in predictive smooth pursuit after mild traumatic brain injury.   Neurosci Lett. 2006;401(1-2):108-113. doi:10.1016/j.neulet.2006.02.074 PubMedGoogle ScholarCrossref
    10.
    Maruta  J, Suh  M, Niogi  SN, Mukherjee  P, Ghajar  J.  Visual tracking synchronization as a metric for concussion screening.   J Head Trauma Rehabil. 2010;25(4):293-305. doi:10.1097/HTR.0b013e3181e67936 PubMedGoogle ScholarCrossref
    11.
    Thiagarajan  P, Ciuffreda  KJ, Ludlam  DP.  Vergence dysfunction in mild traumatic brain injury (mTBI): a review.   Ophthalmic Physiol Opt. 2011;31(5):456-468. doi:10.1111/j.1475-1313.2011.00831.x PubMedGoogle ScholarCrossref
    12.
    Gottshall  K, Drake  A, Gray  N, McDonald  E, Hoffer  ME.  Objective vestibular tests as outcome measures in head injury patients.   Laryngoscope. 2003;113(10):1746-1750. doi:10.1097/00005537-200310000-00016 PubMedGoogle ScholarCrossref
    13.
    Ventura  RE, Balcer  LJ, Galetta  SL.  The neuro-ophthalmology of head trauma.   Lancet Neurol. 2014;13(10):1006-1016. doi:10.1016/S1474-4422(14)70111-5 PubMedGoogle ScholarCrossref
    14.
    Nagib  S, Linens  SW.  Vestibular rehabilitation therapy improves perceived disability associated with dizziness postconcussion.   J Sport Rehabil. 2019;28(7):764-768. doi:10.1123/jsr.2018-0021 PubMedGoogle ScholarCrossref
    15.
    Ghajar  J, Ivry  RB; Cognitive and Neurobiological Research Consortium.  The predictive brain state: timing deficiency in traumatic brain injury?   Neurorehabil Neural Repair. 2008;22(3):217-227. doi:10.1177/1545968308315600 PubMedGoogle ScholarCrossref
    16.
    Barnes  GR.  Cognitive processes involved in smooth pursuit eye movements.   Brain Cogn. 2008;68(3):309-326. doi:10.1016/j.bandc.2008.08.020 PubMedGoogle ScholarCrossref
    17.
    Belanger  HG, Vanderploeg  RD.  The neuropsychological impact of sports-related concussion: a meta-analysis.   J Int Neuropsychol Soc. 2005;11(4):345-357. doi:10.1017/S1355617705050411 PubMedGoogle ScholarCrossref
    18.
    Snegireva  N, Derman  W, Patricios  J, Welman  KE.  Eye tracking technology in sports-related concussion: a systematic review and meta-analysis.   Physiol Meas. 2018;39(12):12TR01. doi:10.1088/1361-6579/aaef44 PubMedGoogle Scholar
    19.
    Ventura  RE, Jancuska  JM, Balcer  LJ, Galetta  SL.  Diagnostic tests for concussion: is vision part of the puzzle?   J Neuroophthalmol. 2015;35(1):73-81. doi:10.1097/WNO.0000000000000223 PubMedGoogle ScholarCrossref
    20.
    Samadani  U.  Will eye tracking change the way we diagnose and classify concussion and structural brain injury?   Concussion. 2015;1(1):CNC2. doi:10.2217/cnc.15.2PubMedGoogle Scholar
    21.
    Hunt  AW, Mah  K, Reed  N, Engel  L, Keightley  M.  Oculomotor-based vision assessment in mild traumatic brain injury: a systematic review.   J Head Trauma Rehabil. 2016;31(4):252-261. doi:10.1097/HTR.0000000000000174 PubMedGoogle ScholarCrossref
    22.
    Webb  B, Humphreys  D, Heath  M.  Oculomotor executive dysfunction during the early and later stages of sport-related concussion recovery.   J Neurotrauma. 2018;35(16):1874-1881. doi:10.1089/neu.2018.5673 PubMedGoogle ScholarCrossref
    23.
    Murray  NG, Szekely  B, Islas  A,  et al.  Smooth pursuit and saccades after sport-related concussion.   J Neurotrauma. 2020;37(2):340-346. doi:10.1089/neu.2019.6595 PubMedGoogle ScholarCrossref
    24.
    Cochrane  GD, Christy  JB, Almutairi  A, Busettini  C, Swanson  MW, Weise  KK.  Visuo-oculomotor function and reaction times in athletes with and without concussion.   Optom Vis Sci. 2019;96(4):256-265. doi:10.1097/OPX.0000000000001364PubMedGoogle ScholarCrossref
    25.
    Hunfalvay  M, Roberts  CM, Murray  N, Tyagi  A, Kelly  H, Bolte  T.  Horizontal and vertical self-paced saccades as a diagnostic marker of traumatic brain injury.   Concussion. 2019;4(1):CNC60. doi:10.2217/cnc-2019-0001 PubMedGoogle ScholarCrossref
    26.
    Kelly  KM, Kiderman  A, Akhavan  S,  et al.  Oculomotor, vestibular, and reaction time effects of sports-related concussion: video-oculography in assessing sports-related concussion.   J Head Trauma Rehabil. 2019;34(3):176-188. doi:10.1097/HTR.0000000000000437 PubMedGoogle ScholarCrossref
    27.
    Hunt  AWM, Mah  K, Reed  N, Engel  L, Keightley  M.  Oculomotor-based vision assessment in mild traumatic brain injury: a systematic review.   J Head Trauma Rehabil. 2016;31(4):252-261. doi:10.1097/HTR.0000000000000174 PubMedGoogle ScholarCrossref
    28.
    Voss  HU, McCandliss  BD, Ghajar  J, Suh  M; (CNRC-TBI).  A quantitative synchronization model for smooth pursuit target tracking.   Biol Cybern. 2007;96(3):309-322. doi:10.1007/s00422-006-0116-2 PubMedGoogle ScholarCrossref
    29.
    Samadani  U, Ritlop  R, Reyes  M,  et al.  Eye tracking detects disconjugate eye movements associated with structural traumatic brain injury and concussion.   J Neurotrauma. 2015;32(8):548-556. doi:10.1089/neu.2014.3687 PubMedGoogle ScholarCrossref
    30.
    Cifu  DX, Wares  JR, Hoke  KW, Wetzel  PA, Gitchel  G, Carne  W.  Differential eye movements in mild traumatic brain injury versus normal controls.   J Head Trauma Rehabil. 2015;30(1):21-28. doi:10.1097/HTR.0000000000000036 PubMedGoogle ScholarCrossref
    31.
    Maruta  J, Spielman  LA, Rajashekar  U, Ghajar  J.  Association of visual tracking metrics with post-concussion symptomatology.   Front Neurol. 2018;9:611. doi:10.3389/fneur.2018.00611 PubMedGoogle ScholarCrossref
    32.
    Hunfalvay  M, Roberts  CM, Murray  NP,  et al.  Vertical smooth pursuit as a diagnostic marker of traumatic brain injury.   Concussion. 2020;5(1):CNC69. doi:10.2217/cnc-2019-0013 PubMedGoogle ScholarCrossref
    33.
    Zaslow  T, Mueske  NM, Yu  GJ, Conrad-Forrest  A, Edison  B, Wren  TAL.  Oculomotor function in adolescent athletes following concussion.   Orthop J Sports Med. Published online April 20, 2020. doi:10.1177/2325967120S00150Google Scholar
    34.
    Samadani  U, Li  M, Qian  M,  et al.  Sensitivity and specificity of an eye movement tracking-based biomarker for concussion.   Concussion. 2015;1(1):CNC3.PubMedGoogle Scholar
    35.
    Bin Zahid  A, Hubbard  ME, Lockyer  J,  et al.  Eye tracking as a biomarker for concussion in children.   Clin J Sport Med. 2020;30(5):433-443. doi:10.1097/JSM.0000000000000639PubMedGoogle Scholar
    36.
    Kutcher  JS, Giza  CC.  Sports concussion diagnosis and management.   Continuum (Minneap Minn). 2014;20(6 Sports Neurology):1552-1569. doi:10.1212/01.CON.0000458974.78766.58PubMedGoogle Scholar
    37.
    US Centers for Disease Control and Prevention. Report to Congress on mild traumatic brain injury in the United States: steps to prevent a serious public health problem. Published September 2003. Accessed January 26, 2021. https://www.cdc.gov/traumaticbraininjury/pdf/mtbireport-a.pdf
    38.
    Murray  NP, Hunfalvay  M, Bolte  T.  The reliability, validity, and normative data of interpupillary distance and pupil diameter using eye-tracking technology.   Transl Vis Sci Technol. 2017;6(4):2. doi:10.1167/tvst.6.4.2 PubMedGoogle ScholarCrossref
    39.
    Murray  NP, Hunfalvay  M, Roberts  CM, Lange  B.  Reliability and normative data of computerized dynamic visual acuity tests.   Vision Dev Rehab. 2017;3(1):23-32.Google Scholar
    40.
    Iverson  GL, Lovell  MR, Collins  MW.  Validity of ImPACT for measuring processing speed following sports-related concussion.   J Clin Exp Neuropsychol. 2005;27(6):683-689. doi:10.1081/13803390490918435 PubMedGoogle ScholarCrossref
    41.
    Šimundić  A-M.  Measures of diagnostic accuracy: basic definitions.   EJIFCC. 2009;19(4):203-211.PubMedGoogle Scholar
    42.
    Lempke  LB, Howell  DR, Eckner  JT, Lynall  RC.  Examination of reaction time deficits following concussion: a systematic review and meta-analysis.   Sports Med. 2020;50(7):1341-1359. doi:10.1007/s40279-020-01281-0 PubMedGoogle ScholarCrossref
    43.
    Liversedge  SP, Findlay  JM.  Saccadic eye movements and cognition.   Trends Cogn Sci. 2000;4(1):6-14. doi:10.1016/S1364-6613(99)01418-7 PubMedGoogle ScholarCrossref
    44.
    Fukushima  K, Fukushima  J, Warabi  T, Barnes  GR.  Cognitive processes involved in smooth pursuit eye movements: behavioral evidence, neural substrate and clinical correlation.   Front Syst Neurosci. 2013;7:4. doi:10.3389/fnsys.2013.00004 PubMedGoogle ScholarCrossref
    45.
    Mullen  SJ, Yücel  YH, Cusimano  M, Schweizer  TA, Oentoro  A, Gupta  N.  Saccadic eye movements in mild traumatic brain injury: a pilot study.   Can J Neurol Sci. 2014;41(1):58-65. doi:10.1017/S0317167100016279 PubMedGoogle ScholarCrossref
    46.
    Wasserman  EB, Abar  B, Shah  MN, Wasserman  D, Bazarian  JJ.  Concussions are associated with decreased batting performance among Major League Baseball players.   Am J Sports Med. 2015;43(5):1127-1133. doi:10.1177/0363546515576130 PubMedGoogle ScholarCrossref
    47.
    Nordström  A, Nordström  P, Ekstrand  J.  Sports-related concussion increases the risk of subsequent injury by about 50% in elite male football players.   Br J Sports Med. 2014;48(19):1447-1450. doi:10.1136/bjsports-2013-093406 PubMedGoogle ScholarCrossref
    48.
    Guskiewicz  KM, McCrea  M, Marshall  SW,  et al.  Cumulative effects associated with recurrent concussion in collegiate football players: the NCAA Concussion Study.   JAMA. 2003;290(19):2549-2555. doi:10.1001/jama.290.19.2549 PubMedGoogle ScholarCrossref
    49.
    Zemper  ED.  Two-year prospective study of relative risk of a second cerebral concussion.   Am J Phys Med Rehabil. 2003;82(9):653-659. doi:10.1097/01.PHM.0000083666.74494.BA PubMedGoogle ScholarCrossref
    50.
    Zonner  SW, Ejima  K, Fulgar  CC,  et al.  Oculomotor response to cumulative subconcussive head impacts in US high school football players: a pilot longitudinal study.   JAMA Ophthalmol. 2019;137(3):265-270. doi:10.1001/jamaophthalmol.2018.6193 PubMedGoogle ScholarCrossref
    51.
    Kawata  K, Rubin  LH, Lee  JH,  et al.  Association of football subconcussive head impacts with ocular near point of convergence.   JAMA Ophthalmol. 2016;134(7):763-769. doi:10.1001/jamaophthalmol.2016.1085PubMedGoogle ScholarCrossref
    52.
    Kellar  D, Newman  S, Pestilli  F, Cheng  H, Port  NL.  Comparing fMRI activation during smooth pursuit eye movements among contact sport athletes, non-contact sport athletes, and non-athletes.   Neuroimage Clin. 2018;18:413-424. doi:10.1016/j.nicl.2018.01.025 PubMedGoogle ScholarCrossref
    53.
    Mack  DJ, Belfanti  S, Schwarz  U.  The effect of sampling rate and lowpass filters on saccades—a modeling approach.   Behav Res Methods. 2017;49(6):2146-2162. doi:10.3758/s13428-016-0848-4 PubMedGoogle ScholarCrossref
    54.
    Lange  B, Hunfalvay  M, Murray  N, Roberts  CM, Bolte  T.  Reliability of computerized eye-tracking reaction time tests in non-athletes, athletes, and individuals with traumatic brain injury.   Optom Vis Performance. 2018;6(3):167-182.Google Scholar
    55.
    Munoz  DP, Broughton  JR, Goldring  JE, Armstrong  IT.  Age-related performance of human subjects on saccadic eye movement tasks.   Exp Brain Res. 1998;121(4):391-400. doi:10.1007/s002210050473 PubMedGoogle ScholarCrossref
    56.
    Lapidot  MB.  Does the brain age uniformly? evidence from effects of smooth pursuit eye movements on verbal and visual tasks.   J Gerontol. 1987;42(3):329-331. doi:10.1093/geronj/42.3.329 PubMedGoogle ScholarCrossref
    ×