Using Oculometrics to Identify and Evaluate Neurological Disorders with Micha Breakstone NeuraLight

Micha Breakstone is the CEO and Co-Founder of NeuraLight and points out the lack of objective and sensitive evaluation measures available to diagnose neurological diseases like multiple sclerosis, Alzheimer's, and Parkinson's.  Currently, evaluations using short questionnaires, visual observations, and self-reported behavior are further influenced by biases of the physician. NeuraLight is developing an approach that focuses on using images of the eye to produce objective, repeatable measures of disease progression. Micha elaborates, "We look at over 100 parameters at once, and we capture these by standard videos or anything from a webcam to a mobile phone. And we've developed a quite unique technology that marries, on the one hand, a kind of standard machine learning. But on the other hand, something that is entirely unique to us which is signal processing. So the ability to basically bring technologies that are available for satellites, in super-resolution, bring them down to earth into mobile phones in order to take these standard videos and extract parameters at a super-resolution or a sub-pixel resolution from standard video." "There are actually over 700 papers that were published in the last 20 years or so showing that oculometrics, namely, micro measurements of the eyes, are extremely good as a proxy both for currently used endpoints like UPDRS and EDSS that I mentioned earlier. But also in phenotyping patients, looking at the difference between secondary progressive MS and repeating, remitting MS or the subtypes, four or five subtypes of Alzheimer's, just being able to look at those oculometrics or ocular measures, being able to identify the subtype of the disease is a great proxy for the standard scale." @NeuraLight #NeuraLight #AI #Oculometric #OcularMeasures #NeurologicalDiseases #NeurodegenerativeDisease #Parkinsons #MS #MultipleSclerosis #Alzheimers #Dementia #ClinicalTrials #NeurologicalEvaluation NeuraLight.ai Download the transcript here