‘EyePhone’ app evaluates eye movement to detect vestibular strokes
Click Here to Manage Email Alerts
Key takeaways:
- A smartphone app accurately identified a type of rapid uncontrollable eye movement associated with stroke.
- The EyePhone app performed with similar accuracy to currently used standard devices.
A novel smartphone app was shown to diagnose vestibular stroke symptoms via tracked eye movements in healthy volunteers with an accuracy similar to current clinical standard devices, researchers reported.
The app, “EyePhone,” which leverages the embedded facial recognition capabilities of the iPhone’s front camera, still needs to be tested in a real-world population and in emergency settings; however, researchers said the proof-of-concept study shows the technology could potentially reduce barriers to an accurate diagnosis for a common type of stroke that is easily misunderstood.
Image: Adobe Stock
“Smartphones can record eye movements accurately and can be a powerful mTech tool for neurological diagnosis,” Ali Saber Tehrani, MD, assistant professor of neurology at Johns Hopkins University School of Medicine, told Healio. “Smartphones can be a useful tool in the clinical assessment of patients, especially those with dizziness and vertigo. By enabling access to eye movement analysis and interpretation, we can democratize access to high-quality care everywhere using smartphones.”
Stroke vs. dizziness
Of the approximately 5 million annual dizziness visits to EDs in the U.S., vestibular strokes account for more than 250,000 diagnoses, Tehrani and colleagues wrote in the Journal of the American Heart Association. To distinguish vestibular strokes from peripheral dizziness, clinicians use the head impulse, nystagmus and test of skew (HINTS) eye examination; however, eye-movement signs are subtle and there is often a lack of familiarity and difficulty with recognition of abnormal eye movements in the ED.
Tehrani and colleagues analyzed data from 10 healthy volunteers with normal eye movements and no known vestibular disorders. The mean age of participants was 30 years and 10% were women. Researchers compared the accuracy of EyePhone with video oculography (VOG) goggles, used as a standard reference to quantify nystagmus. The study team used an iPhone 13 ProMax (Apple) with the EyePhone app installed for all the phone recordings; VOG traces were recorded by ICS Impulse goggles (Natus Medical Inc.) using OtoSuite Vestibular software.
Participants viewed optokinetic stimuli with incremental velocities (2° to 12° per second) in four directions, designed to induce nystagmus.
“This method relies on the optokinetic reflex in healthy subjects induced by movements of bars in the visual field, resulting in the slip of the images projected on the fovea (eg, looking at trees from the window of a moving car),” the researchers wrote. “To induce the optokinetic nystagmus, we used a set of videos that showed moving black and white strips, hence simulating a movement of the visual field. We aimed to induce nystagmus in four directions (right, left, up, and down) with incremental velocity (from 2° per second up to 12° per second).”
Researchers then extracted slow phase velocities from EyePhone data in each direction and compared them with the corresponding slow phase velocities obtained by the goggles. There was at least 10 minutes of rest between VOG and EyePhone recordings to prevent any potential remaining effect. Researchers calculated the area under the receiver operating characteristic curve (AUROC) for nystagmus detection by EyePhone.
High correlation between EyePhone, goggles
Researchers found that EyePhone highly correlated with the VOG recordings for horizontal slow phase velocities (r = 0.98; 95% CI, 0.97-0.99) and vertical slow phase velocities (r = 0.94; 95% CI, 0.92-0.96). Calibration increased the slope of linear regression for horizontal and vertical slow phase velocities.
Evaluating the EyePhone’s performance using VOG data with a 2° per second threshold showed an AUROC of 0.87 for horizontal and vertical nystagmus detection.
The researchers noted that the app’s performance for detecting vertical nystagmus was lower after average calibration compared with VOG, adding that the discrepancy between horizontal and vertical detection can be potentially explained by how eyelids might affect accurate detection of the pupils’ position, especially in downgaze. Such issues are mitigated in VOG recordings by having a very close image of the eye, providing higher spatial resolution, and the use of infrared light for pupil detection.
“While this is an important caveat that needs to be addressed, one has to consider that EyePhone’s performance in detecting the presence (rather than the velocity) of vertical nystagmus is what eventually helps in diagnosing strokes,” the researchers wrote. “Moreover, the presence of horizontal gaze-evoked nystagmus is more valuable in diagnosing vestibular strokes.”
Tehrani told Healio that research with real-world patients and in emergency settings is needed to assess the diagnostic accuracy of the smartphone application.
For more information:
Ali Saber Tehrani, MD, can be reached at ali.tehrani@jhmi.edu.
Collapse
Read more about
Click Here to Manage Email Alerts