Abstract Details

Several studies highlighted subclinical oculomotor abnormalities (OA) in VOG recording in MS patients even in the early stages of the disease. However, access to VOG is limited in clinical practice.

To assess the reliability of video oculography (VOG) recording using a home-developed iPad application called eVOG. eVOG uses the facial recognition of the iPad camera to track eye position and reproduces usual VOG paradigms.

We recruited patients with MS who had a normal clinical oculomotor exam.

Patients underwent on the same day a classical and an iPad VOG exam in a randomized order.

On each exam, we looked at the type and number of OA: square wave jerks, alteration of horizontal movements, alteration of vertical movements, alteration of pursuit and alteration of antisaccades. Additionally, some abnormalities could only be detected on the classical VOG: internuclear ophthalmoplegia, nystagmus, saccadic intrusions.

28 patients were included (mean age: 46 yrs, F/M ratio: 2, mean EDSS: 2)

The mean number of OA per patient was 4.1 on conventional VOG and 2.3 on eVOG.

Depending on the considered OA, the sensitivity of eVOG ranged from 57 to 100%, and specificity ranged from 71 to 89%. 

We found a correlation for saccades latencies (r=0.575 ; p<0.0001), saccades speeds (r=0.39 ; p<0.0001) and for the total number of OA (r=0.66; p<0.0001).

Patients who had 0-1 OA on eVOG had on average 2 OA on classical VOG whereas patients who had 2 or more OA on eVOG had on average 5 OA on classical VOG (p<0.0001).

eVOG can provide relevant information about OA in MS patients. It could help to investigate the predictive value of OA in MS patients at a larger scale, as it is an easy-to-deploy solution.