Evaluating impaired consciousness after acquired brain injury using a virtual-reality-based eye-tracking system

doi:10.21203/rs.3.rs-4943480/v1

Download PDF

Research Article

Evaluating impaired consciousness after acquired brain injury using a virtual-reality-based eye-tracking system

https://doi.org/10.21203/rs.3.rs-4943480/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Virtual reality (VR) can provide an experimental basis for inferring consciousness using information obtained from the responses of persons with disorders of consciousness (DOC) to denoised exogenous stimuli. Although integration of eye-tracking technologies has been proposed for evaluating levels of DOC, the calibration process poses substantial challenges or may be infeasible for persons with DOC. We aimed to demonstrate the validity and clinical utility of biomarkers obtained from pupil movements in response to visuoauditory stimuli presented in a VR environment using eye-tracking technology, while addressing the limitations of uncalibrated pupil trajectories.

Methods

We enrolled persons with prolonged DOC caused by acquired brain injury who showed continuous eye-opening for at least 15 min, along with healthy individuals. Participants were shown nine visuoauditory stimuli in a three-dimensional VR space while pupil movements were measured using an eye-tracking system. We calculated the relative pupil tracking length for visual and auditory stimuli (RPTL-V and RPTL-A). We established their cut-off values based on their congruence with the evoked potential test result to ascertain the presence of a “visuoauditory response by the RPTL.” Based on these results and the “visuoauditory response by Coma Recovery Scale-Revised (CRS-R),” the individuals were classified into “overt tracking,” “covert tracking,” and “no sign of tracking” groups. After 1 year, we assessed whether the participants could obey a simple command.

Results

Fifteen persons with prolonged DOC (median age, 67 [interquartile range {IQR}, 64.5–72.5] years; 9 [60%] women) and six healthy individuals (median age, 55 [IQR, 52.3–58.3] years; 3 [50%] women) participated. The RPTL-V and RPTL-A distribution varied according to the level of DOC and integrity of the visual or auditory pathway. The RPTL-V and RPTL-A cut-off values were 14.737 and 30.019, respectively. Frequencies of simple command obeying in the groups were: overt tracking, 5/8 (62.5%); covert tracking, 2/4 (50%); and no sign of tracking, 0/3 (0%). In 1/15 persons with DOC, a visuoauditory response not detected through the CRS-R was identified via the RPTL.

Conclusions

A VR-based eye-tracking system can quantitatively assess DOC, offering valid and clinically useful support for diagnosis and prognosis in conjunction with the CRS-R.

Brain Injuries

Disorder of Consciousness

Virtual Reality

Eye-Tracking Technology

Calibration

Accurately diagnosing the level of disorders of consciousness (DOC) is clinically important because it can be used to determine whether the patient should be treated with pharmacologic therapeutics, noninvasive brain stimulation, or other therapeutic options [1]. It is also important for determining the prognosis of persons with DOC, which may reduce the burden on caregivers [2–4]. Furthermore, distinguishing between DOC status is important from a medico-ethical perspective, as minimally conscious state (MCS) and vegetative state/unresponsive wakefulness syndrome (VS/UWS) pose distinct ethical issues [5].

A variety of clinical tools have been developed to assess consciousness, such as the Glasgow Outcome Scale, Glasgow Outcome Scale-Extended, and Coma Recovery Scale-Revised (CRS-R) [6]. In the assessment using these scales, persons with DOC are provided with a variety of stimuli to identify meaningful behavioral responses to them. Among the various clinical assessment tools, the CRS-R is considered a standardized assessment tool for diagnosing DOC and has been widely used in research as well as in clinical practice [1, 3, 7]. However, it is prone to measurement error [8–10] as the level of response to a stimulus can vary depending on the circadian rhythm, medication use, comorbidities, and presence of coexisting sensory or motor impairments [11]. Moreover, unexpected distracting stimuli in the environment in which the assessment is being conducted and the skill level of the raters can affect the accuracy of the assessment results [5, 12]. To overcome these limitations, repeatedly assessing persons with DOC using the CRS-R has been proposed [10, 13]. However, owing to the complexity of the CRS-R assessment, it takes a substantial amount of time to conduct the assessment [14], making repeated use of the CRS-R infeasible in a clinical setting.

Including visual tracking as part of the diagnostic criteria for DOC was proposed due to the higher prevalence of visual response among persons with MCS than among persons with VS and its association with more favorable prognostic outcomes [15–17]. Furthermore, visual responses, such as visual pursuit or visual fixation, have been recognized as the earliest and most frequently observed signs of consciousness [18–21]. Another study showed multimodal evidence that, among persons with DOC, the group with cortically mediated visual responses had qualitatively different visual processing compared to the group with reflexive visual responses [22]. Therefore, there have been attempts to assess the level of consciousness using eye-tracking technology [23–26], which have demonstrated that utilizing eye-tracking technology to determine levels of consciousness achieves accuracy comparable to clinical assessments agreed upon by domain experts [24], suggesting its utility in predicting the emergence from DOC [23, 25, 26]. For these reasons, expert consensus obtained through the Delphi method has also presented a unanimous agreement on the necessity of employing eye-tracking (or pupil-tracking) technology in the assessment of DOC [1]. Pupil tracking technologies generally necessitate an initial calibration procedure to ensure accuracy in data collection [24]. For persons with DOC, the calibration process is challenging or even impossible. Consequently, previous studies either did not perform the calibration process or did not consider it [23, 26, 27]. Wannez et al. sought to overcome the inaccuracy of uncalibrated pupil trajectories by extracting meaningful values from the pupil trajectories of persons with DOC [24]. However, their method using the Pearson correlation value between the coordinates of the pupil trajectory and the mirror trajectory is inherently vulnerable to measurement errors due to the high granularity of the pupil trajectory coordinates, leading to errors during signal acquisition, filtering, and statistical processing.

Recently, virtual reality (VR) technology has been increasingly employed in clinical settings for assessment and therapy [28]. The integrated information theory, which explains consciousness in terms of measurable amounts of information, is a framework that quantifies the level of consciousness by calculating the amount of information generated by the brain [29]. The composition of stimuli perceived by the patient, influenced by exogenous stimuli, affects the amount of integrated information (φ-value), a phenomenon previously referred to as the contextuality of consciousness [30, 31]. Therefore, to determine the level of consciousness through the calculated amount of integrated information, it is essential to control the influence of exogenous stimuli. Exogenous stimuli can be divided into intended stimuli and noise. It is impossible to completely control the influence of exogenous stimuli in the real world due to the uncontrollable nature of noise. However, use of VR apparatus makes it possible to eliminate the noise and effectively control the intended stimuli by configuring the immersive environment. Consequently, in a VR environment, the level of consciousness can be estimated through the amount of integrated information calculated from responses to the same intended stimuli with minimal bias (represented in Fig. 1). This technology has the potential to function as a debiasing tool at a level unattainable by previous studies using computer screens or mirrors [23, 24, 26, 27]. Therefore, we expect that evaluation of the level of DOC using a VR-based eye-tracking system could complement the CRS-R to improve diagnostic accuracy. Accordingly, in this study, we aimed to devise a VR-based eye-tracking system using visuoauditory stimuli delivered through a head-mounted display (HMD) and to demonstrate the validity and clinical utility of the system in assessing consciousness while addressing the limitations of uncalibrated pupil trajectories.

Inclusion and exclusion criteria

We enrolled persons with DOC secondary to acquired brain injury from July 2020 to June 2021 and healthy individuals without a history of acquired brain injury during the same period. The inclusion criteria for persons with DOC were as follows: (1) ability to perform continuous eye-opening for at least 15 min; (2) no functional problems related to eyeball movement, as confirmed by an experienced physiatrist; and (3) presence of prolonged DOC (i.e., they had been in a DOC state for more than 4 weeks following an injury) [32]. The exclusion criteria for persons with DOC were as follows: (1) cranial nerve injury accompanied by brain injury; (2) a history of psychiatric disease; and (3) a history of neurodegenerative disease or cerebral neoplastic disease. When needed, the persons with DOC received medical and anticonvulsant treatments, as well as pharmacological interventions for consciousness improvement when possible. We also enrolled a control group of healthy individuals who were age- and sex-matched with the enrolled persons with DOC.

Visuoauditory stimuli

We devised nine types of visuoauditory stimuli to be displayed using a VR system (Fig. 2A). An additional movie file shows this in more detail [see Additional file 1]. The stimuli were provided to the enrolled persons and healthy individuals through an HTC Vive HMD (HTC Corporation, Taoyuan City, Taiwan). The visuoauditory stimuli were as follows: (1) a visual stimulus of a flame explosion; (2) a bright ball sequentially positioned that disappeared at the top, bottom, left, and right of the visual field; (3) an image of “Lenna” moving up, down, left, and right at a speed of π/4 rad/s; (4) directional auditory stimuli presented alternately on the left and right, saying “Please look here, patient”; (5) a checkerboard appearing in a random order from the left and right directions; (6) a checkerboard appearing in a random order from the left and right directions, accompanied by sound that occurred simultaneously with the appearance; (7) visual stimuli of a banana and an orange switching positions between left and right multiple times, accompanied by the auditory commands: “Look at the banana” and “Look at the orange”; (8) a bright ball that made a sound upon appearance and gradually faded away; the location of appearance moved progressively to the right; and (9) a ball moving left and right between the walls erected at the left and right ends, making a sound when it hit the walls. The second, third, fourth, and seventh stimuli were adapted to the VR environment from the stimuli used during the CRS-R assessment. All visuoauditory stimuli were presented within a three-dimensional VR space.

Pupil tracking system and experimental set-up

The binocular pupil movements of the enrolled participants were recorded and quantified using a computerized infrared pupil tracking module (Visual Camp Corporation, Seoul, Korea) integrated into an HMD, as utilized in a previous study [33]. An additional figure file shows this in more detail [see Additional file 2]. As calibration is unfeasible when applying eye-tracking technology to persons with DOC [23, 25], we utilized an eye-tracking system calibrated for healthy individuals to gather data on pupil movement. The system tracked the pupils simultaneously at a sampling rate of 60 Hz as visuoauditory stimuli were administered to the participants. An alarm system activated in the event of eye-tracking failure enabled the monitoring of participants’ eye-opening quality throughout the experiment, as illustrated in Fig. 2B. If eye closure persisted for more than 3 s, the arousal facilitation protocol from the CRS-R was employed to maintain eye-opening [34], The procedure was repeated until > 80% of a single session was successfully documented. If it was not possible to track the pupil movement in the enrolled individuals even after conducting three consecutive sessions, the evaluation was rescheduled for another day. All participants were tested while seated in an upright or semi-Fowler position, with the option of using a pillow or neck support as needed.

Devising biomarkers

The relative pupil trajectory length (RPTL) was defined as a value quantifying the pupil movement that occurred when the visuoauditory stimuli were presented compared with when the blank stimulus (“+” mark) was presented. After smoothing the pupil trajectory with a moving average for 1 s to minimize the effect of outliers, the RPTL was calculated using the following formula: RPTL = (total pupil trajectory length when the visuoauditory stimuli were presented/duration of visuoauditory stimuli)/(total pupil trajectory length when the blank stimulus was presented/duration of the blank stimulus). The RPTL-V was defined as the RPTL value when visual stimuli (stimuli no. 1, 2, 3, 5, 6, 7, 8, and 9) were presented to the individuals. The RPTL-A was defined as the RPTL value when the auditory stimuli (stimuli no. 4, 6, 7, 8, and 9) were presented to the individuals. We calculated the RPTL-V and RPTL-A using the tracking results from the pupil that exhibited greater movement among both eyes during the presentation of visuoauditory stimuli.

Acquiring and analyzing the visual evoked potential (VEP) and brainstem auditory evoked potential (BAEP)

The acquisition of evoked potentials (EPs) was performed in accordance with the standards for the use of EPs in persons with DOC [35, 36]. Flash VEPs, facilitated by Light Emitting Diode (LED) goggles, were utilized in this study because of their utility in assessing uncooperative persons [37]. The flash stimulus through LED goggles subtends a visual field of at least 200°, which was recommended by the International Society for Clinical Electrophysiology of Vision (ISCEV) standard for clinical visual EPs [38]. The goggles consisted of high-efficiency red LEDs (635 nm) in a 3 × 5 array in each eyepiece. The flash rate was 1.7 Hz, with a 200-ms duration. The VEP test was conducted on the enrolled persons with DOC positioned in either a semi-Fowler or supine posture in an isolated, quiet environment, and the patient’s eye opening was checked prior to the test. The VEP signals were obtained at a sweep duration of 300 ms. The active electrode was affixed at Oz, the reference electrode at Fz, and the ground electrode at Cz, in accordance with the ISCEV standards for clinical visual EPs [38]. Due to anticipated patient non-cooperation, measurements utilized binocular visual input concurrently. N300 implies the activation of basilar dendrites within the visual cortex [39], and a previous study employed the presence of P100 as a criterion for determining the existence of a VEP response [40]. Therefore, we established the absence of N300 and P100, as determined through visual inspection by two domain experts (LHH and HCH), as the criterion for assessing the absence of a VEP response.

In the BAEP test, auditory stimuli were administered via headphones with the following configurations: a click sound at a frequency of 11.29 Hz, with a duration of 0.1 ms, at an intensity of 75 dB nHL. White noise masking at 40 dB was applied on the contralateral side. During the test, the patient was positioned in either a semi-Fowler or supine posture, and the test was performed in an isolated and quiet setting. The electrode placement locations were as follows: the recording electrodes were positioned at A1 and A2, the reference electrode was situated at Cz, and the ground electrode was affixed at Fz. The BAEP signals were acquired under a 10-ms sweep duration. Auditory stimuli were sequentially administered to each ear, and the absence of auditory response was determined if the V/I ratio in both ears was < 0.5, which has been employed in previous studies that assessed BAEP in persons with DOC [41, 42]. The above examination procedure and decision criteria were also applied to healthy controls in the same manner.

Functional magnetic resonance imaging (fMRI) experiment and analysis

We conducted an fMRI experiment to verify whether the RPTL-V has physiological relevance. In the fMRI experiment, all the enrolled individuals received visual stimuli equivalent to those conveyed through the HMD. The provision of auditory stimuli was precluded by limitations pertaining to the MRI compatibility of the apparatus. The visual stimuli were programmed in nordicAktiva (NordicNeuroLab, Bergen, Norway) and shown on an MRI-compatible screen (40″ in diameter, with 3840 × 2160 pixels). Among the nine visuoauditory stimuli, excluding the fourth stimulus, which comprised solely auditory elements, eight visual stimuli (stimuli no. 1, 2, 3, 5, 6, 7, 8, and 9) were presented for a duration of 15 s each. Two visual stimuli were sequentially provided over a 30-s period, followed by a 30-s presentation of a blank stimulus (denoted as a “+” mark). This alternating sequence of visual and blank stimuli was maintained throughout the fMRI experiment for a total duration of 6 min. fMRI analysis was performed in two steps. In the first step, single-individual analysis involved a comparison of activated lesions in an individual to visual stimuli against blank stimuli. This step included the following specific preprocessing steps: (1) slice-time correction with a repetition time of 3 s; (2) motion correction; (3) co-registration with individual-specific T1 images; (4) normalization using the Montreal Neurological Institute template; and (5) application of spatial smoothing with a full-width half-maximum of 5 mm. In the second step, group-level analysis involved the comparison of activated lesions according to the presence of a visual response confirmed by the RPTL-V cut-off value using a random effect model adjusted for age. SPM12 software (available at: http://www.fil.ion.ucl.ac.uk/spm) was used for fMRI data analysis. The MR protocol and fMRI experiment design details are presented elsewhere [see Additional file 3].

Evaluation of obeying a simple command at the 1-year follow-up

Because persons with DOC cannot move without assistance, in-person assessments are not feasible unless they are readmitted or revisit our institution. Consequently, we evaluated the status of the enrolled persons at the 1-year follow-up through face-to-face assessments when they returned to our facility and via video calls after confirming the patient’s condition through family statements if readmission was infeasible. We presented the patient with one of the following commands: “Close your eyes for 2 s,” “Close your eyes twice in succession,” or “Stick out your tongue.” A consistent response to a given command on two consecutive occasions was considered a substantial execution of simple commands.

Statistical analysis

To investigate the differences in the RPTL-V and RPTL-A according to the level of DOC and the response to VEP or BAEP, we performed the t-test or Wilcoxon rank-sum test. We employed a top-down approach to determine the cut-off values for the RPTL-V and RPTL-A that ascertain the presence of a meaningful visuoauditory response. As VEP and BAEP responses are indicative of the integrity of the visual and auditory pathways [43], we selected the cut-off values of RPTL-V and RPTL-A that discriminate the presence of a VEP response and a BAEP response with the highest accuracy through receiver operating characteristic (ROC) curve analysis. Using these cut-off values to assess the presence of visual and auditory responses, it was determined that a visuoauditory response by RPTL was present if a response was confirmed using either of the two biomarkers. Using visuoauditory responses identified by the CRS-R and RPTL, we classified the enrolled persons into “overt tracking” (positive response with both indicators), “covert tracking” (identified by either a visuoauditory response by the CRS-R or a visuoauditory response by the RPTL), and “no sign of tracking” (negative for both the visuoauditory response by the CRS-R and the visuoauditory response by the RPTL) groups. Subsequently, we compared the frequency of obeying a simple command at the 1-year follow-up of these groups. A P value < 0.05 was considered statistically significant.

Fifteen persons with DOC and six healthy individuals were enrolled in the study (Table 1). Except for one patient with DOC who dropped out of the study due to death, all other individuals remained in the study during the 1-year follow-up period. The persons with DOC (n = 15) had a median age of 67 (interquartile range [IQR], 64.5–72.5) years, and nine (60%) were female. The six healthy individuals had a median age of 55 (IQR, 52.3–58.3) years, and three (50%) were female. There was a difference between the RPTL-V (MCS [n = 11], 99.73 ± 109.38; VS/UWS [n = 4], -10.05 ± 58.38; p = 0.083) and RPTL-A (MCS [n = 11], 105.00 ± 251.83; VS/UWS [n = 4], -8.36 ± 41.54; p = 0.412) according to the level of DOC; however, this difference was not significant (Fig. 3). There was a difference between the RPTL-V (VEP response [+] [n = 12], 185.19 ± 139.70; VEP response [-] [n = 9], -3.14 ± 36.66; p < 0.001) and RPTL-A (BAEP response [+] [n = 12], 275.39 ± 245.25; BAEP response [-] [n = 9], -1.71 ± 43.13; p = 0.001) according to the VEP and BAEP response (Fig. 4). We selected the cut-off values of the RPTL-V and RPTL-A that discriminated the presence of VEP and BAEP responses with the highest accuracy through ROC curve analysis (RPTL-V, 14.737 with a sensitivity of 1.000 and a specificity of 0.889; RPTL-A, 30.019 with a sensitivity of 0.889 and a specificity of 0.833; Fig. 5). We found that visual stimuli activated the visual cortex in persons with a visual response, which the RPTL-V confirmed, compared with persons without a response (Fig. 6). The differences in the frequency of obeying a simple command at the 1-year follow-up according to the level of visuoauditory response in the groups were as follows: overt tracking, 5/8 (62.5%); covert tracking, 2/4 (50%); no sign of tracking, 0/3 (0%) (Table 2). In 1/15 persons with DOC, a visuoauditory response not detected through the CRS-R was identified via the RPTL.

Table 1

Clinical characteristics of the enrolled persons with DOC (n = 15)
Age (years)	Sex	Diagnosis	Chronicity of disease (days)	Total CRS-R (A/V/M/O/C/A)	Level of DOC by the CRS-R	Visuoauditory response by the CRS-R	VEP response	BAEP response	RPTL-V	RPTL-A	Visuoauditory response by the RPTL	Category of visuoauditory response	Obeying a simple command at the 1-year follow-up
≥ 65	F	SAH	39	11 (2/3/3/1/0/2)	MCS	+	+	+	294.565	-11.305	+	Overt tracking	+^a)
≥ 65	M	SDH	560	7 (1/1/2/1/0/2)	VS/UWS	-	-	-	-88.310	-68.385	-	No sign of tracking	-^a)
< 65	F	ICH with IVH	32	8 (1/2/2/1/0/2)	MCS	+	-	+	13.530	840.450	+	Overt tracking	+^a)
≥ 65	M	ICH with IVH	300	11 (3/2/3/1/0/2)	MCS	+	-	-	1.900	4.049	-	Covert tracking	-
< 65	M	HBI	98	6 (1/0/2/1/0/2)	VS/UWS	-	-	-	-6.580	1.320	-	No sign of tracking	-^a)
≥ 65	F	SAH	114	11 (2/3/3/1/0/2)	MCS	+	-	-	-3.345	-42.545	-	Covert tracking	+^a)
< 65	F	ICH	102	5 (1/1/2/0/0/1)	VS/UWS	-	-	-	52.765	27.110	+	Covert tracking	+^a)
≥ 65	F	ICH	291	12 (2/2/3/1/1/3)	MCS	+	-	-	-2.670	-15.410	-	Covert tracking	N/A^b)
< 65	M	SAH	42	14 (3/2/4/2/1/2)	MCS	+	+	-	141.160	-15.155	+	Overt tracking	+
≥ 65	F	TBI	72	14 (3/3/3/1/1/3)	MCS	+	+	+	172.530	105.730	+	Overt tracking	-^a)
≥ 65	M	TBI	199	15 (3/3/3/2/1/3)	MCS	+	+	-	263.360	9.255	+	Overt tracking	-^a)
≥ 65	M	TBI	61	19 (3/4/5/3/1/3)	MCS	+	+	+	117.240	168.775	+	Overt tracking	+
≥ 65	F	SAH	131	15 (3/3/5/0/1/3)	MCS	+	+	-	96.2950	84.370	+	Overt tracking	+^a)
≥ 65	F	ICH with IVH	189	6 (0/1/2/1/0/2)	VS/UWS	-	-	+	1.943	6.521	-	No sign of tracking	-^a)
≥ 65	F	TBI	153	20 (3/5/5/3/1/3)	MCS	+	-	+	2.476	32.928	+	Overt tracking	-
^a) identified by face-to-face assessment; ^b) deceased before follow-up assessment DOC, disorders of consciousness; CRS-R, Coma Recovery Scale-Revised; A/V/M/O/C/A, auditory/visual/motor/oromotor and verbal/communication/arousal; VEP, visual evoked potential; BAEP, brainstem auditory evoked potential; RPTL-V, relative pupil tracking length for visual stimuli; RPTL-A, relative pupil tracking length for auditory stimuli; RPTL, relative pupil tracking length; SAH, subarachnoid hemorrhage; ICH, intracranial hemorrhage; IVH, intraventricular hemorrhage; HBI, hypoxic brain injury; TBI, traumatic brain injury; MCS, minimally conscious state; VS/UWS, vegetative state/unresponsive wakefulness syndrome

[Table 1 near here]

Table 2

Frequencies of obeying a simple command at 1 year according to the visuoauditory responses
		Visuoauditory response identified by the CRS-R
		Yes	No
Visuoauditory response identified by the RPTL	Yes	8 (overt tracking) 5/8 (62.5%) obeyed a simple command	1 (covert tracking) 1/1 (100%) obeyed a simple command
Visuoauditory response identified by the RPTL	No	3 (covert tracking) 1/3 (33.3%) obeyed a simple command (1 patient died before follow-up assessment)	3 (no sign of tracking) 0/3 (0%) obeyed a simple command

RPTL, relative pupil tracking length; CRS-R, Coma Recovery Scale-Revised

Clinical meaning of the results

This study demonstrated the validity and clinical usefulness of assessing impaired consciousness using a VR-based eye-tracking system. Clinical validity was demonstrated because the RPTL-V and RPTL-A were in concordance with the results of the CRS-R and the integrity of visual and auditory pathways, as evidenced by the EP test, and the visual cortex was activated in the group with visual response confirmed by the RPTL-V level. Clinical usefulness was demonstrated by the enhancement in prognostic accuracy through the visuoauditory response confirmed by RPTL. This may be attributed to the quantitative evaluation of visual responses simultaneously providing homogeneous and immersive stimuli in a VR-configured environment with less noise than in the real world. This could establish VR as an experimental tool to attenuate the contextual sensitivity of consciousness when inferring it through the amount of integrated information calculated using the response to the intended stimuli [30, 31]. Furthermore, the RPTL is robust against measurement errors that may arise from the high granularity of the pupil trajectory coordinates [24] as it is a relative metric that compares the pupil trajectory identified during several seconds of visual or auditory stimulation with the pupil trajectory during several seconds of blank stimulation. Previous studies utilizing eye-tracking technology have not used auditory stimuli [23–27]. Thus, delivering a combination of visual and auditory stimuli in an immersive environment also positions our system as an advancement over the existing methodologies. However, it is important to note that this system does not evaluate the motor, oromotor/verbal, or communication aspects included in the CRS-R. Therefore, using this system in conjunction with the CRS-R is more appropriate than as a stand-alone evaluation tool [8–10]. Furthermore, the automated assessment system, which provides stimuli and verifies responses in a VR environment, is expected to facilitate more frequent evaluations, which could mitigate potential errors in consciousness assessments that may arise from heterogeneity in the timing of measurements, especially due to diurnal variations in visual response [11]. Moreover, the ability to quantitatively measure visual responses highlights an increase in evaluation precision, facilitating the resolution of gray zones between each stage of the visual response assessed by the CRS-R. Therefore, this method allows for detecting subtle changes in consciousness, allowing for the early identification of emerging consciousness.

Limitations of this study

Further, in this study, we established cut-off values for the RPTL-V and RPTL-A by utilizing the essential condition of a substantial visual response to visuoauditory stimuli, which reflects the integrity of the visual and auditory pathways. Consequently, when the integrity of these pathways is compromised, a lower RPTL value can yield false negatives for detecting consciousness. Traumatic brain injury can cause damage to peripheral components of the visual or auditory pathways [44, 45]. Therefore, to assess consciousness accurately, the assessment system devised in this study necessitates a meticulous examination of cranial nerve damage caused by acquired brain injury, utilizing EP tests or neuroimaging studies.

In addition, due to the limitation of the sample size, only one individual among the enrolled persons who did not show visuoauditory responses in the CRS-R assessment could demonstrate visuoauditory responses by the RPTL. Therefore, the generalizability of the findings from this study necessitates verification using a larger-scale sample. Another limitation is that the confirmation of obeying a simple command at the 1-year follow-up was conducted in one of two ways: face-to-face assessments or video calls based on family statements. We planned to match the age and sex of persons with DOC with those of healthy controls but encountered a slight discrepancy in the age distribution. However, as the RPTL (V or A) data from persons with DOC and healthy individuals were used together to determine the cut-off values for the RPTL (V or A), the impact of age differences between the two groups on the findings may be negligible. Furthermore, although age influences the prognosis of DOC [46, 47], its effect on the diagnosis of DOC remains unclear. Lastly, previous studies have also utilized VEP and BAEP tests to examine the integrity of visual and auditory pathways in persons with DOC [40–43, 48, 49]. Yet, the assessment protocols for VEP and BAEP have not been standardized for this patient population. Moreover, since there are no established criteria for discriminating between normal and abnormal VEP and BAEP results in persons with DOC, previous studies have proposed various methods for defining diagnostic criteria for VEP and BAEP [40–42, 48, 49]. Therefore, determining the cut-off values for the RPTL-V and RPTL-A based on the integrity of the visual and auditory pathways evaluated through BAEP and VEP inherently carries the potential for error.

Remaining challenges and future directions

Clinical guidelines recommend forcibly opening the eyelids for consciousness assessment through eye movement observation [3], which is unfeasible in VR settings. Therefore, in this study, we evaluated the eye-opening quality in real-time and subsequently applied the arousal facilitation protocol of the CRS-R, if needed [34]. However, applying this protocol may pose potential risks for the accuracy of consciousness assessment by eye-tracking. Furthermore, VR requires specific equipment, such as the HMDs used in this study, which poses challenges in application to persons with DOC, especially those with movement disorders due to acquired brain injury [50, 51]. In future studies, we plan to improve the devised assessment system by considering the identified challenges and verify whether these research results can be replicated with a larger sample of persons with DOC of various etiologies.

We demonstrated that the level of DOC can be quantitatively measured with a VR-based eye-tracking system. We also found that the system is valid and clinically useful as an assistive diagnostic and prognostic tool that can be used with the CRS-R.

BAEP

brainstem auditory evoked potential

CRS-R

Coma Recovery Scale–Revised

DOC

disorders of consciousness

HMD

head-mounted display

ISCEV

International Society for Clinical Electrophysiology of Vision

MCS

minimally conscious state

ROC

receiver operating characteristic

RPTL-A

relative pupil tracking length for auditory stimuli

RPTL-V

relative pupil tracking length for visual stimuli

VEP

visual evoked potential

virtual reality

VS/UWS

vegetative state/unresponsive wakefulness syndrome

Ethics approval and consent to participate

Written informed consent was obtained from the legal guardians of eligible individuals and all healthy participants before enrollment in this study. The protocol was approved by the Ethics Committee of the Konkuk University Medical Center (IRB No. KUH2020-07-070).

Availability of data and materials

The datasets generated and/or analyzed during the current study and all custom-written R codes are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

This study was supported by the Rehabilitation Research & Development Support Program (#NRCRSP-EX20011) of the National Rehabilitation Center, Ministry of Health and Welfare, Korea.

Authors' contributions

The study was conceptualized, designed, and planned by HHL, WM and JL. HHL and CH developed the experimental apparatus. Data collection was conducted by HHL, CH, and CHK. Data analysis was performed by HHL. The manuscript was drafted by HHL. All authors contributed to the critical review and revision of the manuscript, and approved the final version.

Acknowledgements

Not applicable.

Overbeek BUH, Lavrijsen JCM, van Gaal S, Kondziella D, Eilander HJ, Koopmans RTCM. Towards consensus on visual pursuit and visual fixation in patients with disorders of consciousness. A Delphi study. J Neurol. 2022;269:3204–15.
Luauté J, Maucort-Boulch D, Tell L, Quelard F, Sarraf T, Iwaz J, et al. Long-term outcomes of chronic minimally conscious and vegetative states. Neurology. 2010;75:246–52.
Kondziella D, Bender A, Diserens K, van Erp W, Estraneo A, Formisano R, et al. European Academy of Neurology guideline on the diagnosis of coma and other disorders of consciousness. Eur J Neurol. 2020;27:741–56.
Boegle K, Bassi M, Comanducci A, Kuehlmeyer K, Oehl P, Raiser T, et al. Informal caregivers of patients with disorders of consciousness: A qualitative study of communication experiences and information needs with physicians. Neuroethics. 2022;15:24.
Young MJ, Bodien YG, Giacino JT, Fins JJ, Truog RD, Hochberg LR, et al. The neuroethics of disorders of consciousness: A brief history of evolving ideas. Brain Truog Rd: Hochberg LR. 2021;144:3291–310.
Pavlov YG, Spiegelsberger F, Kotchoubey B. Predicting outcome in disorders of consciousness: A mega-analysis. Ann Clin Transl Neurol. 2024;11:1465–77.
Giacino JT, Katz DI, Schiff ND, Whyte J, Ashman EJ, Ashwal S, et al. Comprehensive systematic review update summary: Disorders of consciousness: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology; the American Congress of Rehabilitation Medicine; and the National Institute on Disability, Independent Living, and Rehabilitation Research. Neurology. 2018;91:461–70.
Schnakers C, Vanhaudenhuyse A, Giacino J, Ventura M, Boly M, Majerus S, et al. Diagnostic accuracy of the vegetative and minimally conscious state: Clinical consensus versus standardized neurobehavioral assessment. BMC Neurol. 2009;9:35.
van Erp WS, Lavrijsen JCM, Vos PE, Bor H, Laureys S, Koopmans RTCM. The vegetative state: Prevalence, misdiagnosis, and treatment limitations. J Am Med Dir Assoc. 2015;16:85.e9-85.e14.
Wang J, Hu X, Hu Z, Sun Z, Laureys S, Di H. The misdiagnosis of prolonged disorders of consciousness by a clinical consensus compared with repeated coma-recovery scale-revised assessment. BMC Neurol. 2020;20:343.
Candelieri A, Cortese MD, Dolce G, Riganello F, Sannita WG. Visual pursuit: Within-day variability in the severe disorder of consciousness. J Neurotrauma. 2011;28:2013–7.
Vredeveldt A, Perfect TJ. Reduction of environmental distraction to facilitate cognitive performance. Front Psychol. 2014;5:860.
Wannez S, Heine L, Thonnard M, Gosseries O, Laureys S, Coma Science Group collaborators. The repetition of behavioral assessments in diagnosis of disorders of consciousness. Ann Neurol. 2017;81:883–9.
Bodien YG, Vora I, Barra A, Chiang K, Chatelle C, Goostrey K, et al. Feasibility and validity of the coma recovery scale-revised for accelerated standardized testing: A practical assessment tool for detecting consciousness in the Intensive Care Unit. Ann Neurol. 2023;94:919–24.
Giacino JT, Kalmar K. Diagnostic and prognostic guidelines for the vegetative and minimally conscious states. Neuropsychol Rehabil. 2005;15:166–74.
Giacino JT, Kalmar K. The vegetative and minimally conscious states: A comparison of clinical features and functional outcome. J Head Trauma Rehabil. 1997;12:36–51.
Shiel A, Horn SA, Wilson BA, Watson MJ, Campbell MJ, Mclellan DL. The Wessex Head Injury Matrix (WHIM) main scale: A preliminary report on a scale to assess and monitor patient recovery after severe head injury. Clin Rehabil. 2000;14:408–16.
Estraneo A, Pascarella A, Moretta P, Loreto V, Trojano L. Clinical and electroencephalographic on–off effect of amantadine in chronic non-traumatic minimally conscious state. J Neurol. 2015;262:1584–6.
Wannez S, Gosseries O, Azzolini D, Martial C, Cassol H, Aubinet C, et al. Prevalence of coma-recovery scale-revised signs of consciousness in patients in minimally conscious state. Neuropsychol Rehabil. 2018;28:1350–9.
Bagnato S, Boccagni C, Sant'angelo A, Fingelkurts AA, Fingelkurts AA, Galardi G. Longitudinal assessment of clinical signs of recovery in patients with unresponsive wakefulness syndrome after traumatic or nontraumatic brain injury. J Neurotrauma. 2017;34:535–9.
Martens G, Bodien Y, Sheau K, Christoforou A, Giacino JT. Which behaviours are first to emerge during recovery of consciousness after severe brain injury? Ann Phys Rehabil Med. 2020;63:263–9.
Sattin D, Rossi Sebastiano D, D’Incerti L, Guido D, Marotta G, Benti R, et al. Visual behaviors in disorders of consciousness: Disentangling conscious visual processing by a multimodal approach. Eur J Neurosci. 2020;52:4345–55.
Trojano L, Moretta P, Loreto V, Cozzolino A, Santoro L, Estraneo A. Quantitative assessment of visual behavior in disorders of consciousness. J Neurol. 2012;259:1888–95.
Wannez S, Hoyoux T, Langohr T, Bodart O, Martial C, Wertz J, et al. Objective assessment of visual pursuit in patients with disorders of consciousness: An exploratory study. J Neurol. 2017;264:928–37.
Aklepi G, Manolovitz B, Robayo LE, Sarafraz A, Blandino CF, Arwari B, et al. Covert tracking to immersive stimuli in traumatic brain injury subjects with disorders of consciousness. J Neurotrauma. 2024;41:646–59.
Alkhachroum A, Aklepi G, Sarafraz A, Robayo LE, Manolovitz BM, Blandino CF, et al. Covert tracking to visual stimuli in comatose patients with traumatic brain injury. Neurology. 2023;101:489–94.
Trojano L, Moretta P, Loreto V, Santoro L, Estraneo A. Affective saliency modifies visual tracking behavior in disorders of consciousness: A quantitative analysis. J Neurol. 2013;260:306–8.
Li L, Yu F, Shi D, Shi J, Tian Z, Yang J, et al. Application of virtual reality technology in clinical medicine. Am J Transl Res. 2017;9:3867–80.
Tononi G, Boly M, Massimini M, Koch C. Integrated information theory: From consciousness to its physical substrate. Nat Rev Neurosci. 2016;17:450–61.
Miyahara K, Witkowski O. The integrated structure of consciousness: Phenomenal content, subjective attitude, and noetic complex. Phenom Cogn Sci. 2019;18:731–58.
de Barros JA, Montemayor C, De Assis LPG. Contextuality in the integrated information theory; 2017. p. 57–70. http://link.springer.com/10.1007/978-3-319-52289-0_5
Giacino JT, Katz DI, Schiff ND, Whyte J, Ashman EJ, Ashwal S, et al. Practice guideline update recommendations summary: Disorders of consciousness: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology; the American Congress of Rehabilitation Medicine; and the National Institute on Disability, Independent Living, and Rehabilitation Research. Neurology. 2018;91:450–60.
Hyun C, Kim K, Lee S, Lee HH, Lee J. Quantitative evaluation of the consciousness level of patients in a vegetative state using virtual reality and an eye-tracking system: A single-case experimental design study. Neuropsychol Rehabil. 2022;32:2628–45.
Giacino JT, Kalmar K, Whyte J. The JFK Coma Recovery Scale-Revised: Measurement characteristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85:2020–9.
André-Obadia N, Zyss J, Gavaret M, Lefaucheur J-P, Azabou E, Boulogne S, et al. Recommendations for the use of electroencephalography and evoked potentials in comatose patients. Neurophysiol Clin. 2018;48:143–69.
Chatrian GE, Bergamasco B, Bricolo A, Frost JD, Prior PF. IFCN recommended standards for electrophysiologic monitoring in comatose and other unresponsive states. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol. 1996;99:103–22.
Taylor MJ, Menzies R, MacMillan LJ, Whyte HE. VEPs in normal full-term and premature neonates: Longitudinal versus cross-sectional data. Electroencephalogr Clin Neurophysiol. 1987;68:20–7.
Odom JV, Bach M, Brigell M, Holder GE, McCulloch DL, Tormene AP, et al. ISCEV standard for clinical visual evoked potentials (2009 update). Doc Ophthalmol. 2010;120:111–9.
Markand ON. Clinical evoked potentials: An illustrated manual. Springer Natl; 2020.
de Biase S, Gigli GL, Lorenzut S, Bianconi C, Sfreddo P, Rossato G, et al. The importance of polysomnography in the evaluation of prolonged disorders of consciousness: Sleep recordings more adequately correlate than stimulus-related evoked potentials with patients' clinical status. Sleep Med. 2014;15:393–400.
Fischer C, Luaute J, Morlet D. Event-related potentials (MMN and novelty P3) in permanent vegetative or minimally conscious states. Clin Neurophysiol. 2010;121:1032–42.
Cavinato M, Freo U, Ori C, Zorzi M, Tonin P, Piccione F, et al. Post-acute P300 predicts recovery of consciousness from traumatic vegetative state. Brain Inj. 2009;23:973–80.
Ragazzoni A, Cincotta M, Giovannelli F, Cruse D, Young GB, Miniussi C, et al. Clinical neurophysiology of prolonged disorders of consciousness: From diagnostic stimulation to therapeutic neuromodulation. Clin Neurophysiol. 2017;128:1629–46.
Armstrong RA. Visual problems associated with traumatic brain injury. Clin Exp Optom. 2018;101:716–26.
Fausti SA, Wilmington DJ, Gallun FJ, Myers PJ, Henry JA. Auditory and vestibular dysfunction associated with blast-related traumatic brain injury. J Rehabil Res Dev. 2009;46:797–810.
Howell K, Grill E, Klein A-M, Straube A, Bender A. Rehabilitation outcome of anoxic-ischaemic encephalopathy survivors with prolonged disorders of consciousness. Resuscitation. 2013;84:1409–15.
Pistoia F, Sacco S, Franceschini M, Sarà M, Pistarini C, Cazzulani B, et al. Comorbidities: A key issue in patients with disorders of consciousness. J Neurotrauma. 2015;32:682–8.
Rossi Sebastiano D, Panzica F, Visani E, Rotondi F, Scaioli V, Leonardi M, et al. Significance of multiple neurophysiological measures in patients with chronic disorders of consciousness. Clin Neurophysiol. 2015;126:558–64.
Wijnen VJM, Eilander HJ, de Gelder B, van Boxtel GJM. Visual processing during recovery from vegetative state to consciousness: Comparing behavioral indices to brain responses. Neurophysiol Clin. 2014;44:457–69.
Stasolla F, Vinci LA, Cusano M. The integration of assistive technology and virtual reality for assessment and recovery of post-coma patients with disorders of consciousness: A new hypothesis. Front Psychol. 2022;13:905811.
Jankovic J. Post-traumatic movement disorders: Central and peripheral mechanisms. Neurology. 1994;44:2006–14.

No competing interests reported.

Download PDF

Reviewers agreed at journal
19 Oct, 2024
Reviewers invited by journal
08 Oct, 2024
Editor assigned by journal
21 Aug, 2024
Submission checks completed at journal
21 Aug, 2024
First submitted to journal
20 Aug, 2024

You are reading this latest preprint version

Evaluating impaired consciousness after acquired brain injury using a virtual-reality-based eye-tracking system

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Inclusion and exclusion criteria

Visuoauditory stimuli

Pupil tracking system and experimental set-up

Devising biomarkers

Acquiring and analyzing the visual evoked potential (VEP) and brainstem auditory evoked potential (BAEP)

Functional magnetic resonance imaging (fMRI) experiment and analysis

Evaluation of obeying a simple command at the 1-year follow-up

Statistical analysis

Results

Discussion

Clinical meaning of the results

Limitations of this study

Remaining challenges and future directions

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1