aEYE: A deep learning system for video nystagmus detection

Feasibility of video-based real-time nystagmus tracking: a lightweight deep learning model approach using ocular object segmentation

Background

Eye movement tests remain significantly underutilized in emergency departments and primary healthcare units, despite their superior diagnostic sensitivity compared to neuroimaging modalities for the differential diagnosis of acute vertigo. This underutilization may be attributed to a potential lack of awareness regarding these tests and the absence of appropriate tools for detecting nystagmus. This study aimed to develop a nystagmus measurement algorithm using a lightweight deep-learning model that recognizes the ocular regions.

Method

The deep learning model was used to segment the eye regions, detect blinking, and determine the pupil center. The model was trained using images extracted from video clips of a clinical battery of eye movement tests and synthesized images reproducing real eye movement scenarios using virtual reality. Each eye image was annotated with segmentation masks of the sclera, iris, and pupil, with gaze vectors of the pupil center for eye tracking. We conducted a comprehensive evaluation of model performance and its execution speeds in comparison to various alternative models using metrics that are suitable for the tasks.

Results

The mean Intersection over Union values of the segmentation model ranged from 0.90 to 0.97 for different classes (sclera, iris, and pupil) across types of images (synthetic vs. real-world images). Additionally, the mean absolute error for eye tracking was 0.595 for real-world data and the F1 score for blink detection was ≥ 0.95, which indicates our model is performing at a very high level of accuracy. Execution speed was also the most rapid for ocular object segmentation under the same hardware condition as compared to alternative models. The prediction for horizontal and vertical nystagmus in real eye movement video revealed high accuracy with a strong correlation between the observed and predicted values (r = 0.9949 for horizontal and r = 0.9950 for vertical; both p < 0.05).

Conclusion

The potential of our model, which can automatically segment ocular regions and track nystagmus in real time from eye movement videos, holds significant promise for emergency settings or remote intervention within the field of neurotology.

Capturing nystagmus during vertigo attacks using a smartphone: adherence, characteristics, pearls and pitfalls

OBJECTIVE

To investigate adherence, characteristics, and first clinical experiences of capturing ictal nystagmus at home, which can be performed to complement the diagnostic process in patients with episodic vestibular symptoms.

METHODS

Patients were recruited at a tertiary referral center in case capturing ictal nystagmus could contribute to the diagnostic process (e.g., to detect or rule out BPPV). They were asked to capture ictal nystagmus with their own smartphone at home, using a smartphone-based adapter (Nystagmocatcher, Balansdiagnos, Stockholm, Sweden). All recordings were analyzed by the last author (RvdB), and the adherence, characteristics, and first clinical experiences were evaluated.

RESULTS

Seventy patients with vestibular symptoms were asked to participate in this study. Sixty-two (89%) agreed to participate. The median period of participation was 86 days. Fifty-one patients experienced attacks during the study period. Eventually, 51% of them provided eye movement recordings sufficient for analysis. Different types of nystagmus were observed: positional nystagmus related to BPPV, positional nystagmus not related to BPPV, functional eye movements, and the absence of nystagmus or functional eye movements. Capturing ictal nystagmus could contribute to the diagnostic process in several ways, including to detect or rule out BPPV, to detect or rule out vestibular origin of symptoms, to determine the affected side, telemedicine, to monitor attack frequency, and to detect malingering. Furthermore, strict guidance of patients was necessary, which could be time-consuming.

CONCLUSION

Capturing ictal nystagmus can contribute to the diagnostic process in several ways, which motivates to rethink current clinical workflow in vestibular medicine. However, strict guidance is necessary and not all patients provide ictal recordings. In an outpatient setting, it would be advised to use ictal nystagmus recordings on indication, to complement the diagnostic process.

Neurological update: neuro-otology 2023

Much has changed since our last review of recent advances in neuro-otology 7 years ago. Unfortunately there are still not many practising neuro-otologists, so that most patients with vestibular problems need, in the first instance, to be evaluated and treated by neurologists whose special expertise is not neuro-otology. The areas we consider here are mostly those that almost any neurologist should be able to start managing: acute spontaneous vertigo in the Emergency Room-is it vestibular neuritis or posterior circulation stroke; recurrent spontaneous vertigo in the office-is it vestibular migraine or Meniere's disease and the most common vestibular problem of all-benign positional vertigo. Finally we consider the future: long-term vestibular monitoring and the impact of machine learning on vestibular diagnosis.

Comparative Analysis of Strabismus Measurement Using a Video Oculagraphy System and Alternate Prism Cover Test

PURPOSE

The aim of the study was to compare distance vision measurements obtained with video oculography (VOG) and an alternate prism cover test (APCT).

DESIGN

Single-center, retrospective, and cross-sectional.

METHODS

Eighty-seven subjects with strabismus were included. All patients underwent an optometric examination. The measurement of strabismus in distance vision was performed with the best optical correction using the APCT and the VOG GazeLab. Subjects were divided according to the type of strabismus; 41 were esotropic and 46 were exotropic.

RESULTS

The general comparison of APCT and VOG showed a mean difference of 0.60±2.74 prism diopters (Δ), not observing statistical differences between both methods ( P =0.059) and presenting a correlation of 0.976 ( P <0.001). Using APCT, the mean amount of esotropia for the group was 18.31±11.64 Δ and that of exotropia was 19.62±8.80 Δ. Compared with the VOG, the mean value of esotropia for the group was 18.65±11.65 Δ and that of exotropia was 19.48±8.62 Δ. The means are statistically different for the esotropia group ( P =0.005) but not for the exotropia group ( P =0.318). There was a high direct correlation between the methods of measurement for both the esotropic ( R =0.980; P <0.001) and exotropic patients ( R =0.975; P <0.001). Bland-Altman analysis shows a mean difference of 1.37±2.76 Δ for the esotropia patients and 0.31±2.66 Δ for the exotropic patients, respectively.

CONCLUSIONS

This study demonstrated comparable results in measuring strabismus between VOG and APTC for esotropia and exotropia, with an excellent correlation between both methods and good agreement, especially in subjects with exotropia.

Smartphone-Assisted Medical Care for Vestibular Dysfunction as a Telehealth Strategy for Digital Therapy Beyond COVID-19: Scoping Review

BACKGROUND

Dizziness and vertigo can be caused by various factors, such as peripheral vestibular and central disorders. Although consultations with specialists are advisable when necessary, patients with severe vertigo symptoms may have limited mobility, which may interfere with hospital visits. The spread of COVID-19 has further limited the number of hospital visits for patients with dizziness; therefore, a method of medical care that enables more accurate treatment under time and geographical constraints is needed. Telemedicine has become widespread, owing to the popularity of smartphone and tablet devices in recent years, and the use of devices and systems has made it possible to provide efficient medical care. However, no previous scoping review has mapped existing studies on telemedicine for vertigo and dizziness, and no recommendations have been made regarding which devices and systems should be used for specific diseases.

OBJECTIVE

The aim of this review was to map and assess previous studies on the use of information communications technology, smartphones, and apps for treating patients with vertigo and discuss the added value of introducing telemedicine to improve the quality of medical care and create an environment that builds security and trust among patients.

METHODS

A scoping review was conducted with the methodological framework of Arksey and O'Malley and in accordance with the of the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analysis extension for Scoping Reviews) guidelines. The PubMed, MEDLINE, and Cochrane Library databases were searched to retrieve previous reports on smartphone-assisted telemedicine treatment for vertigo published between January 2000 and May 2023. Two authors independently assessed eligibility and extracted data.

RESULTS

This review included 20 papers that reported devices or systems for telemedicine for vestibular dysfunction. Among studies that reported the use of a device or app, 2 were related to anamnesis and subjective symptoms, 12 were related to objective examination, 7 were related to remote diagnosis, and 7 were related to treatment and rehabilitation.

CONCLUSIONS

With the advancement of technology, the use of telemedicine in patients with dizziness may be feasible. In the future, it will be necessary to consider how telemedicine can be used in dizziness treatment and develop an effective treatment system combining in-person medical care and the effective use of devices for the management of severe vertigo and related diseases. The smooth introduction of telemedicine in vertigo treatment is expected to improve the quality of treatment, increase opportunities for patients to receive medical care, and reduce time and travel costs, leading to a sense of security and trust among patients.

Torsional nystagmus recognition based on deep learning for vertigo diagnosis

Introduction

Detection of torsional nystagmus can help identify the canal of origin in benign paroxysmal positional vertigo (BPPV). Most currently available pupil trackers do not detect torsional nystagmus. In view of this, a new deep learning network model was designed for the determination of torsional nystagmus.

Methods

The data set comes from the Eye, Ear, Nose and Throat (Eye&ENT) Hospital of Fudan University. In the process of data acquisition, the infrared videos were obtained from eye movement recorder. The dataset contains 24521 nystagmus videos. All torsion nystagmus videos were annotated by the ophthalmologist of the hospital. 80% of the data set was used to train the model, and 20% was used to test.

Results

Experiments indicate that the designed method can effectively identify torsional nystagmus. Compared with other methods, it has high recognition accuracy. It can realize the automatic recognition of torsional nystagmus and provides support for the posterior and anterior canal BPPV diagnosis.

Discussion

Our present work complements existing methods of 2D nystagmus analysis and could improve the diagnostic capabilities of VNG in multiple vestibular disorders. To automatically pick BPV requires detection of nystagmus in all 3 planes and identification of a paroxysm. This is the next research work to be carried out.

Deep learning in acute vertigo diagnosis

Recent advances in artificial intelligence are transforming healthcare and there are increasing efforts to apply these breakthroughs to the diagnosis of acute vertigo. Because the diagnosis of vertigo relies on the analysis of eye movements, there are several unique considerations that must be made when implementing deep learning approaches to vertigo. This review discusses the need for diagnostic aids for acute vertigo, the techniques used to preprocess eye movement data and adapt deep learning models to vertigo, and summarizes and analyzes all published models to date.