Primary Visual Pathway Changes in Individuals With Chronic Mild Traumatic Brain Injury

Abstract

Importance: Individuals with mild traumatic brain injury (TBI) often report vision problems despite having normal visual acuity and fundus examinations. Diagnostics are needed for these patients.

Objective: To determine if a battery of assessments or machine-learning approaches can aid in diagnosing visual dysfunction in patients with mild TBI.

Design, setting, and participants: This prospective, observational, case-control study was conducted between May 2018 and November 2021. The study setting was at a level 1 trauma research hospital. Participant eligibility included adult males and females with recorded best-corrected visual acuity and normal fundus examination. Individuals in the case group had a history of mild TBI; controls had no history of TBI. Exclusion criteria included a history of ocular, neurological, or psychiatric disease, moderate-severe TBI, recent TBI, metal implants, age younger than 18 years, and pregnancy. Cases and controls were sex- and age-matched. Data analysis was performed from July 2023 to March 2024.

Exposures: History of mild TBI in the case group.

Main outcomes and measures: The single-session visit included the Neurobehavioral Symptom Inventory and measurements of oculomotor function, optical coherence tomography, contrast sensitivity, visual evoked potentials, visual field testing, and magnetic resonance imaging.

Results: A total of 28 participants (mean [SD] age, 35.0 [12.8] years; 15 male [53.6%]) with mild TBI and 28 controls (mean [SD] age, 35.8 [8.5] years; 19 female [67.9%]) were analyzed. Participants with mild TBI showed reduced prism convergence test breakpoint (-8.38; 95% CI, -14.14 to -2.62; P = .008) and recovery point (-8.44; 95% CI, -13.82 to -3.06; P = .004). Participants with mild TBI also had decreased contrast sensitivity (-0.07; 95% CI, -0.13 to -0.01; P = .04) and increased visual evoked potential binocular summation index (0.32; 95% CI, 0.02-0.63; P = .02). A subset of participants exhibited reduced peripapillary retinal nerve fiber layer thickness, increased optic nerve/sheath size, and brain cortical volumes. Machine learning identified subtle differences across the primary visual pathway, including the optic radiations and occipital lobe regions, independent of visual symptoms.

Conclusions and relevance: Results of this case-control study suggest that the visual system was affected in individuals with mild TBI, even in those who did not self-report vision problems. These findings support the utility of a battery of assessments or machine-learning approaches to accurately diagnose this population.

Figure 1.. Deficits in Ocular Motor Function

A, Quantification of prism convergence test (PCT), represented as diopters for break point, recovery point, and prism fusion range. B, Quantification of the near point of convergence (NPC), with the normal range shown in the shaded gray area derived from age- and sex-adjusted normative data; values greater than 6 cm are considered abnormal. C, Quantification of near point of accommodation (NPA). D, NPA values were converted to diopters and plotted against age, overlaid with the Hofstetter equation. Horizontal lines between the upper and lower whisker marks indicate the mean, and the top and bottom whisker marks represent the 95% CIs. A 2-sided P value threshold of .01 was set for PCT and NPC, and .03 for NPA, using a Bonferroni correction to control for multiple comparisons, with a type I error rate of 0.05 per measurement. Analyses were conducted using a linear mixed-effects model and the Wilcoxon rank sum test. AA indicates accommodation amplitude; mTBI, mild traumatic brain injury.

Figure 2.. Structural and Functional Deficits in the Primary Visual Pathway

A, Quantification of contrast sensitivity (CS), represented as log contrast values. B, Quantification of the visual evoked potential (VEP) binocular summation index. C, Quantification of peripapillary retinal nerve fiber layer thickness across retinal regions obtained via optical coherence tomography (OCT). D, Quantification of optic nerve volume (mm³) via magnetic resonance imaging (MRI). Horizontal lines between the upper and lower whisker marks indicate the mean, and the top and bottom whisker marks represent the 95% CIs. A 2-sided P value threshold of .05 was set for CS, .007 for OCT, .01 for MRI of the optic nerve, and .03 for VEP, using a Bonferroni correction to control for multiple comparisons, with a type I error rate of 0.05 per measurement. Analyses were conducted using a linear mixed-effects model and the Wilcoxon rank sum test. mTBI indicates mild traumatic brain injury.

Figure 3.. Structural Deficits in the Posterior Visual Pathway

Tractography-based isolation of the optic radiations (arrowhead) was successful for participants in the control (A-C) and mild traumatic brain injury (mTBI) (D-F) groups. No gross abnormalities were visually apparent on diffusion magnetic resonance imaging. G, The random forest machine learning classifier demonstrated highest accuracy in the model including all metrics. H, The influence of each feature in the random forest machine learning classifier was evaluated by its contribution to decreasing Gini impurity (ie, the probability of misclassifying an observation). A higher decrease in Gini impurity associated with a feature indicates that the feature had a greater effect on correctly classifying an observation as belonging to the mTBI or control group. Machine learning models were created using Python 3.11.0 with packages such as pandas, sklearn, and scipy. AUC indicates area under the curve; OR, optic radiation.

Figure 4.. Battery of Visual Pathway Assessments

A, Heat maps illustrating prevalence of vision deficits in participants with mild traumatic brain injury (mTBI) with no self-reported vision problems or light sensitivity (ie, no vision symptoms; n = 9), self-reported vision problems or light sensitivity (n = 3), or both self-reported vision problems and light sensitivity (n = 16). Each row represents a participant with mTBI. Each column represents a vision metric tested. Navy blue shaded boxes (ie, a score of 2) depict that both eyes had a significant (>2 SD from control) deficit, the royal blue shaded boxes (ie, a score of 1) depict that only 1 eye had a significant (>2 SD from control) deficit, and the light blue shaded boxes (ie, a score of 0) indicate neither eye had a deficit. Boxes with an X depict that no measure was acquired for that specific metric. B and C, Histograms showing distributions of z scores for (B) standard and (C) all ophthalmic and orthoptic assessments. The control group was concatenated into 1 normative histogram (depicted in orange). Each participant with mTBI was depicted with their own histogram demonstrating how their distribution of z scores varied from the concatenated control. The solid blue lines indicate participants with mTBI who were identified as significantly different (false discovery rate–corrected P <.05) using a 2-sample Kolmogorov-Smirnov test. ^aNot including visual evoked potential or magnetic resonance imaging (MRI). Abnormal scores found in 64% of participants with mTBI. ^bAbnormal scores found in 78% of participants with mTBI.

Figure 5.. Summary Graphic of Individual and All Metrics

Due to inherent variability in participants with mild traumatic brain injury (mTBI), a variety of regions of the primary visual pathway can be affected, but only a minority of participants have changes in each area. However, when the entire primary visual pathway is assessed, along with ocular motor changes, 78% of participants with mTBI were identified as having abnormal findings. When machine learning was applied to the optic radiations and cortical visual areas, 70% of participants with mTBI were identified as having abnormal findings. This supports the use of a battery of assessments and/or machine learning approaches to diagnose these participants. RNFL indicates retinal nerve fiber layer; VEP, visual evoked potential. ^aAbnormal scores found in 78% of participants with mTBI.