Driving with binocular visual field loss? A study on a supervised on-road parcours with simultaneous eye and head tracking

Eye-Tracking as a Tool to Evaluate Functional Ability in Everyday Tasks in Glaucoma

To date, few studies have investigated the eye movement patterns of individuals with glaucoma while they undertake everyday tasks in real-world settings. While some of these studies have reported possible compensatory gaze patterns in those with glaucoma who demonstrated good task performance despite their visual field loss, little is known about the complex interaction between field loss and visual scanning strategies and the impact on task performance and, consequently, on quality of life. We review existing approaches that have quantified the effect of glaucomatous visual field defects on the ability to undertake everyday activities through the use of eye movement analysis. Furthermore, we discuss current developments in eye-tracking technology and the potential for combining eye-tracking with virtual reality and advanced analytical approaches. Recent technological developments suggest that systems based on eye-tracking have the potential to assist individuals with glaucomatous loss to maintain or even improve their performance on everyday tasks and hence enhance their long-term quality of life. We discuss novel approaches for studying the visual search behavior of individuals with glaucoma that have the potential to assist individuals with glaucoma, through the use of personalized programs that take into consideration the individual characteristics of their remaining visual field and visual search behavior.

Glaucoma and Driving Risk under Simulated Fog Conditions

Purpose

We evaluate driving risk under simulated fog conditions in glaucoma and healthy subjects.

Methods

This cross-sectional study included 41 glaucoma patients and 25 age-matched healthy subjects who underwent driving simulation. Tests consisted of curve negotiation without and with fog preview at 30 m of distance and two controlled speeds (slow and fast). Inverse time-to-line crossing (invTLC) was used as metric to quantify risk; higher invTLC values indicating higher risk, as less time is available to avoid drifting out of the road. Piecewise regression models were used to investigate the relationship between differences in invTLC in fog and nonfog conditions and visual field loss.

Results

Glaucoma patients had greater increase in driving risk under fog compared to controls, as indicated by invTLC differences (0.490 ± 0.578 s⁻¹ and 0.208 ± 0.106 s⁻¹, respectively; P = 0.002). Mean deviation (MD) of the better eye was significantly associated with driving risk under fog, with a breakpoint of −9 dB identified by piecewise regression. For values below the breakpoint, each 1 dB lower MD of better eye was associated with 0.117 s⁻¹ higher invTLC under fast speed (adjusted R² = 57.9%; P < 0.001).

Conclusions

Glaucoma patients have a steeper increase in driving risk under fog conditions when compared to healthy subjects, especially when the severity of visual field damage falls below −9 dB of MD in the better eye.

Translational Relevance

By investigating the relationship between driving risk and disease severity breakpoint, this study may provide guidance to clinicians in recognizing glaucoma patients who may be unfit to drive in complex situations such as fog.

Drivers' Visual Search Patterns during Overtaking Maneuvers on Freeway

Drivers gather traffic information primarily by means of their vision. Especially during complicated maneuvers, such as overtaking, they need to perceive a variety of characteristics including the lateral and longitudinal distances with other vehicles, the speed of others vehicles, lane occupancy, and so on, to avoid crashes. The primary object of this study is to examine the appropriate visual search patterns during overtaking maneuvers on freeways. We designed a series of driving simulating experiments in which the type and speed of the leading vehicle were considered as two influential factors. One hundred and forty participants took part in the study. The participants overtook the leading vehicles just like they would usually do so, and their eye movements were collected by use of the Eye Tracker. The results show that participants' gaze durations and saccade durations followed normal distribution patterns and that saccade angles followed a log-normal distribution pattern. It was observed that the type of leading vehicle significantly impacted the drivers' gaze duration and gaze frequency. As the speed of a leading vehicle increased, subjects' saccade durations became longer and saccade angles became larger. In addition, the initial and destination lanes were found to be key areas with the highest visual allocating proportion, accounting for more than 65% of total visual allocation. Subjects tended to more frequently shift their viewpoints between the initial lane and destination lane in order to search for crucial traffic information. However, they seldom directly shifted their viewpoints between the two wing mirrors.

Medical review licensing outcomes in drivers with visual field loss in Victoria, Australia

BACKGROUND

Good vision is essential for safe driving and studies have associated visual impairment with an increased crash risk. Currently, there is little information about the medical review of drivers with visual field loss. This study examines the prevalence of visual field loss among drivers referred for medical review in one Australian jurisdiction and investigates factors associated with licence outcome in this group.

METHODS

A random sample of 10,000 (31.25 per cent) medical review cases was extracted for analysis from the Victorian licensing authority. Files were screened for the presence of six visual field-related medical conditions. Data were captured on a range of variables, including referral source, age, gender, health status, crash history and licence outcome. Prevalence analyses were univariate and descriptive. Logistic regression was used to assess factors associated with licence outcomes in the visual field loss group.

RESULTS

Approximately 1.9 per cent of the 10,000 medical review cases screened had a visual field loss condition identified (n = 194). Among the visual field loss group, 57.2 per cent were permitted to continue driving (conditional/unconditional licence). Primary referral sources were the police, self-referrals and general medical practitioners. Key factors associated with licence test outcomes were visual field condition, age group, crash involvement and referral to the Driver Licensing Authority's Medical Advisors. Those who were younger had a crash involvement triggering referral and those who were referred to the Medical Advisors were more likely to have a positive licensing outcome.

CONCLUSION

The evidence base for making licensing decisions is complicated by the variable causes, patterns, progressions and measuring technologies for visual field loss. This study highlighted that the involvement of an expert medical advisory service in Victoria resulted in an increased likelihood that drivers with visual field loss will be allowed to continue driving. Further research is warranted to explore issues relating to severity of field loss and the capacity for compensation.

Driving with homonymous visual field loss: a review of the literature

Driving is an important rehabilitation goal for patients with homonymous field defects (HFDs); however, whether or not people with HFDs should be permitted to drive is not clear. Over the last 15 years, there has been a marked increase in the number of studies evaluating the effects of HFDs on driving performance. This review of the literature provides a much-needed summary for practitioners and researchers, addressing the following topics: regulations pertaining to driving with HFDs, self-reported driving difficulties, pass rates in on-road tests, the effects of HFDs on lane position and steering stability, the effects of HFDs on scanning and detection of potential hazards, screening for potential fitness to drive, evaluating practical fitness to drive and the efficacy of interventions to improve driving of persons with HFDs. Although there is clear evidence from on-road studies that some people with HFDs may be rated as safe to drive, others are reported to have significant deficits in skills important for safe driving, including taking a lane position too close to one side of the travel lane, unstable steering and inadequate viewing (scanning) behaviour. Driving simulator studies have provided strong evidence of a wide range in compensatory scanning abilities and detection performance, despite similar amounts of visual field loss. Conventional measurements of visual field extent (in which eye movements are not permitted) do not measure such compensatory abilities and are not predictive of on-road driving performance. Thus, there is a need to develop better tests to screen people with HFDs for visual fitness to drive. We are not yet at a point where we can predict which HFD patient is likely to be a safe driver. Therefore, it seems only fair to provide an opportunity for individualised assessments of practical fitness to drive either on the road and/or in a driving simulator.

Glaucoma and Driving: On-Road Driving Characteristics

Purpose

To comprehensively investigate the types of driving errors and locations that are most problematic for older drivers with glaucoma compared to those without glaucoma using a standardized on-road assessment.

Methods

Participants included 75 drivers with glaucoma (mean = 73.2±6.0 years) with mild to moderate field loss (better-eye MD = -1.21 dB; worse-eye MD = -7.75 dB) and 70 age-matched controls without glaucoma (mean = 72.6 ± 5.0 years). On-road driving performance was assessed in a dual-brake vehicle by an occupational therapist using a standardized scoring system which assessed the types of driving errors and the locations where they were made and the number of critical errors that required an instructor intervention. Driving safety was rated on a 10-point scale. Self-reported driving ability and difficulties were recorded using the Driving Habits Questionnaire.

Results

Drivers with glaucoma were rated as significantly less safe, made more driving errors, and had almost double the rate of critical errors than those without glaucoma. Driving errors involved lane positioning and planning/approach, and were significantly more likely to occur at traffic lights and yield/give-way intersections. There were few between group differences in self-reported driving ability.

Conclusions

Older drivers with glaucoma with even mild to moderate field loss exhibit impairments in driving ability, particularly during complex driving situations that involve tactical problems with lane-position, planning ahead and observation. These results, together with the fact that these drivers self-report their driving to be relatively good, reinforce the need for evidence-based on-road assessments for evaluating driving fitness.

Motor Vehicle Collision Involvement among Persons with Hemianopia and Quadrantanopia

Persons with homonymous quadrantanopia and hemianopia experience driving restrictions, yet there is little scientific evidence to support driving prohibition among persons with these conditions. This retrospective cohort study compares motor vehicle collision (MVC) rates among 27 current licensed drivers with hemianopic and quadrantanopic field defects, who were ≥6 months from the brain injury date with that of 27 age-matched drivers with normal visual fields. Information regarding all police-reported MVCs that occurred over a period of nine years was obtained. MVC rates per year and per mile travelled were calculated and compared using conditional Poisson regression. Drivers with hemianopia or quadrantanopia had more MVCs per mile driven compared to drivers with normal visual fields; specifically their overall MVC rate was 2.45-times (95% confidence interval (CI) 0.89–3.95) higher and their at-fault MVC rate was 2.64-times (95% CI 1.03–6.80) higher. This study indicates that drivers with hemianopia or quadrantanopia have elevated MVC rates. This is consistent with previous research despite studies showing wide individual variability from excellent to poor driving skills. Future research should focus on the functional and driving performance characteristics associated with superior driving skills and/or those that may be amenable to improvement via behavioral and/or engineering interventions.

On-road Driving Performance of Patients With Bilateral Moderate and Advanced Glaucoma

PURPOSE

To compare on-road driving performance of patients with moderate or advanced glaucoma to controls and evaluate factors associated with unsafe driving.

DESIGN

Case-control pilot study.

METHODS

A consecutive sample of 21 patients with bilateral moderate or advanced glaucoma from Washington University, St Louis, Missouri and 38 community-dwelling controls were enrolled. Participants, aged 55-90 years, underwent a comprehensive clinical evaluation by a trained occupational therapist and an on-road driving evaluation by a masked driver rehabilitation specialist. Overall driving performance of pass vs marginal/fail and number of wheel and/or brake interventions were recorded.

RESULTS

Fifty-two percent of glaucoma participants scored a marginal/fail compared to 21% of controls (odds ratio [OR], 4.1; 95% CI, 1.30-13.14; P = .02). Glaucoma participants had a higher risk of wheel interventions than controls (OR, 4.67; 95% CI, 1.03-21.17; P = .046). There were no differences detected between glaucoma participants who scored a pass vs marginal/fail for visual field mean deviation of the better (P = .62) or worse (P = .88) eye, binocular distance (P = .15) or near (P = .23) visual acuity, contrast sensitivity (P = .28), or glare (P = .88). However, glaucoma participants with a marginal/fail score performed worse on Trail Making Tests A (P = .03) and B (P = .05), right-sided Jamar grip strength (P = .02), Rapid Pace Walk (P = .03), Braking Response Time (P = .03), and identifying traffic signs (P = .05).

CONCLUSIONS

Patients with bilateral moderate or advanced glaucoma are at risk for unsafe driving-particularly those with impairments on psychometric and mobility tests. A comprehensive clinical assessment and on-road driving evaluation is recommended to effectively evaluate driving safety of these patients.

Ocular disease and driving

As the driving population ages, the number of drivers with visual impairment resulting from ocular disease will increase given the age-related prevalence of ocular disease. The increase in visual impairment in the driving population has a number of implications for driving outcomes. This review summarises current research regarding the impact of common ocular diseases on driving ability and safety, with particular focus on cataract, glaucoma, age-related macular degeneration, hemianopia and diabetic retinopathy. The evidence considered includes self-reported driving outcomes, driving performance (on-road and simulator-based) and various motor vehicle crash indices. Collectively, this review demonstrates that driving ability and safety are negatively affected by ocular disease; however, further research is needed in this area. Older drivers with ocular disease need to be aware of the negative consequences of their ocular condition and in the case where treatment options are available, encouraged to seek these earlier for optimum driving safety and quality of life benefits.

Blur, eye movements and performance on a driving visual recognition slide test

PURPOSE

Optical blur and ageing are known to affect driving performance but their effects on drivers' eye movements are poorly understood. This study examined the effects of optical blur and age on eye movement patterns and performance on the DriveSafe slide recognition test which is purported to predict fitness to drive.

METHODS

Twenty young (27.1 ± 4.6 years) and 20 older (73.3 ± 5.7 years) visually normal drivers performed the DriveSafe under two visual conditions: best-corrected vision and with +2.00 DS blur. The DriveSafe is a Visual Recognition Slide Test that consists of brief presentations of static, real-world driving scenes containing different road users (pedestrians, bicycles and vehicles). Participants reported the types, relative positions and direction of travel of the road users in each image; the score was the number of correctly reported items (maximum score of 128). Eye movements were recorded while participants performed the DriveSafe test using a Tobii TX300 eye tracking system.

RESULTS

There was a significant main effect of blur on DriveSafe scores (best-corrected: 114.9 vs blur: 93.2; p < 0.001). There was also a significant age and blur interaction on the DriveSafe scores (p < 0.001) such that the young drivers were more negatively affected by blur than the older drivers (reductions of 22% and 13% respectively; p < 0.001): with best-corrected vision, the young drivers performed better than the older drivers (DriveSafe scores: 118.4 vs 111.5; p = 0.001), while with blur, the young drivers performed worse than the older drivers (88.6 vs 95.9; p = 0.009). For the eye movement patterns, blur significantly reduced the number of fixations on road users (best-corrected: 5.1 vs blur: 4.5; p < 0.001), fixation duration on road users (2.0 s vs 1.8 s; p < 0.001) and saccade amplitudes (7.4° vs 6.7°; p < 0.001). A main effect of age on eye movements was also found where older drivers made smaller saccades than the young drivers (6.7° vs 7.4°; p < 0.001).

CONCLUSIONS

Blur reduced DriveSafe scores for both age groups and this effect was greater for the young drivers. The decrease in number of fixations and fixation duration on road users, as well as the reduction in saccade amplitudes under the blurred condition, highlight the difficulty experienced in performing the task in the presence of optical blur, which suggests that uncorrected refractive errors may have a detrimental impact on aspects of driving performance.

Driving with homonymous visual field loss: Does visual search performance predict hazard detection?

Introduction

Stroke often causes homonymous visual field loss, which can lead to exclusion from driving. Retention of a driving licence is sometimes possible by completing an on-road assessment, but this is not practical for all. It is important to find simple tests that can inform the assessment and rehabilitation of driving-related visual-motor function.

Method

We developed novel computerised assessments: visual search; simple reaction and decision reaction to appearing pedestrians; and pedestrian detection during simulated driving. We tested 12 patients with stroke (seven left, five right field loss) and 12 controls.

Results

The homonymous visual field defect group was split into adequately compensated or inadequately compensated groups based on visual search performance. The inadequately compensated group had problems with stimuli in their affected field: they tended to react more slowly than controls and in the driving task they failed to detect a number of pedestrians. In contrast, the adequately compensated group were better at detecting pedestrians, though reaction times were slightly slower than controls.

Conclusion

We suggest that our search task can predict, to a limited extent, whether a person with stroke compensates for visual field loss, and may potentially identify suitability for specific rehabilitation to promote return to driving.

Eye/head tracking technology to improve HCI with iPad applications

In order to improve human computer interaction (HCI) for people with special needs, this paper presents an alternative form of interaction, which uses the iPad's front camera and eye/head tracking technology. With this functional nature/capability operating in the background, the user can control already developed or new applications for the iPad by moving their eyes and/or head. There are many techniques, which are currently used to detect facial features, such as eyes or even the face itself. Open source bookstores exist for such purpose, such as OpenCV, which enable very reliable and accurate detection algorithms to be applied, such as Haar Cascade using very high-level programming. All processing is undertaken in real time, and it is therefore important to pay close attention to the use of limited resources (processing capacity) of devices, such as the iPad. The system was validated in tests involving 22 users of different ages and characteristics (people with dark and light-colored eyes and with/without glasses). These tests are performed to assess user/device interaction and to ascertain whether it works properly. The system obtained an accuracy of between 60% and 100% in the three test exercises taken into consideration. The results showed that the Haar Cascade had a significant effect by detecting faces in 100% of cases, unlike eyes and the pupil where interference (light and shade) evidenced less effectiveness. In addition to ascertaining the effectiveness of the system via these exercises, the demo application has also helped to show that user constraints need not affect the enjoyment and use of a particular type of technology. In short, the results obtained are encouraging and these systems may continue to be developed if extended and updated in the future.

The relationship between central visual field damage and motor vehicle collisions in primary open-angle glaucoma patients

Purpose

To investigate the relationship between visual field (VF) damage and history of motor vehicle collisions (MVCs) in subjects with primary open-angle glaucoma (POAG).

Methods

MVC history and driving habits were recorded using patient questionnaires in 247 POAG patients. Patients' driving attitudes (carefulness) were estimated using Rasch analysis. The relationship between MVC outcomes and 52 total deviation (TD) values of integrated binocular VF (IVF), better and worse visual acuities (VAs), age and gender was analyzed using principal component analysis and logistic regression.

Results

51 patients had the history of MVCs. Significant difference was observed between patients with and without history of MVCs only for: better VA, a single TD value in the superior-right VF, and the typical distance driven in a week (unpaired t-test, p = 0.002, 0.015 and 0.006, respectively). There was not a significant relationship between MVCs and mean deviation (MD) of IVF (p = 0.41, logistic regression). None of the principal components were significantly correlated with MVC outcome (p>0.05, polynomial logistic regression analysis). There was a significant relationship between IVF MD and Rasch derived Person parameter (R² = 0.023, p = 0.0095). There was also a significant positive relationship between MVCs and the distance driven in a week (p = 0.005, logistic regression).

Conclusions

In this study of POAG patients, MVCs were not related to central binocular VF damage. These results suggest the relationship between visual function and driving is not straightforward, and careful consideration should be given when predicting patients' driving ability using their VF.

Impact of superior and inferior visual field loss on hazard detection in a computer-based driving test

PURPOSE

Binocular visual field (VF) loss is linked to driving impairment, guiding authorities to implement fitness to drive requirements for VFs. Yet, evidence is limited regarding the specific types of VF defect that impede driving. This study used a novel gaze-contingent display to test the hypothesis that superior VF loss impacts detection of driving hazards more than inferior loss.

METHODS

The Hazard Perception Test (HPT) is a computer-based component of the UK examination for learner drivers. It measures the response rate for detecting hazards in a series of real-life driving films, yielding a score out of 75, calculated based on the efficiency of detecting 15 hazards. Thirty UK drivers with healthy vision completed three versions of the HPT in a random order. In two versions, a computer set-up incorporating an eye-tracker modified a simulated VF defect in the superior and inferior VFs, respectively, according to the users' real-time gaze as they completed the HPT. The other version was unmodified to measure the baseline performance.

RESULTS

Participants' mean score at baseline was 49/75 (SD=9). Mean (SD) performance fell by 18% (40(11)) when viewing films with a superior defect and 12% with an inferior defect (43(10)). These average differences were statistically significant (p<0.001; 95% CI for mean difference=1-7) CONCLUSIONS: In this study, simulated VF defects impaired the ability to detect driving hazards relative to participants' normal performances, with superior defects having more impact than inferior defects. These results could help inform the design of fairer tests of the VF component for fitness to drive.

Compensatory eye and head movements of patients with homonymous hemianopia in the naturalistic setting of a driving simulation

Homonymous hemianopia (HH) is a frequent deficit resulting from lesions to post-chiasmal brain structures with a significant negative impact on activities of daily living. To address the question how patients with HH may compensate their visual field defect in a naturalistic environment, we performed a driving simulation experiment and quantitatively analyzed both eye and head movements using a head-mounted pupil camera. 14 patients with HH and 14 matched healthy control subjects participated in the study. Based on the detection performance of dynamically moving obstacles, which appeared unexpectedly along the sides of the road track, we divided the patient group into a high- and a low-performance group. Then, we compared parameters of eye and head movements between the two patient groups and the matched healthy control group to identify those which mediate successful detection of potentially hazardous objects. Differences in detection rates could not be explained by demographic variables or the extent of the visual field defect. Instead, high performance of patients with HH in the naturalistic setting of our driving simulation depended on an adapted visual exploratory behavior characterized by a relative increase in the amplitude and a corresponding increase in the peak velocity of saccades, widening horizontally the distribution of eye movements, and by a shift of the overall distribution of saccades into the blind hemifield. The result of the group comparison analyses was confirmed by a subsequent stepwise regression analysis which identified the horizontal spread of eye movements as single factor predicting the detection of hazardous objects.

Homonymous Visual Field Loss and Its Impact on Visual Exploration: A Supermarket Study

PURPOSE

Homonymous visual field defects (HVFDs) may critically interfere with quality of life. The aim of this study was to assess the impact of HVFDs on a supermarket search task and to investigate the influence of visual search on task performance.

METHODS

Ten patients with HVFDs (four with a right-sided [HR] and six with a left-sided defect [HL]), and 10 healthy-sighted, sex-, and age-matched control subjects were asked to collect 20 products placed on two supermarket shelves as quickly as possible. Task performance was rated as "passed" or "failed" with regard to the time per correctly collected item (T_C -failed = 4.84 seconds based on the performance of healthy subjects). Eye movements were analyzed regarding the horizontal gaze activity, glance frequency, and glance proportion for different VF areas.

RESULTS

Seven of 10 HVFD patients (three HR, four HL) passed the supermarket search task. Patients who passed needed significantly less time per correctly collected item and looked more frequently toward the VFD area than patients who failed. HL patients who passed the test showed a higher percentage of glances beyond the 60° VF (P < 0.05).

CONCLUSION

A considerable number of HVFD patients performed successfully and could compensate for the HVFD by shifting the gaze toward the peripheral VF and the VFD area.

TRANSLATIONAL RELEVANCE

These findings provide new insights on gaze adaptations in patients with HVFDs during activities of daily living and will enhance the design and development of realistic examination tools for use in the clinical setting to improve daily functioning. (http://www.clinicaltrials.gov, NCT01372319, NCT01372332).