Nederlands
English
Reyedar
Visual impairments and neurodegenerative disorders such as Glaucoma, Parkinson's Disease, and Multiple Sclerosis, severely impact the quality-of-life of hundreds of millions of people around the world. The accurate detection of early symptoms is critical for a correct treatment but is often not possible. The current screening tests are either invasive, time-consuming or unreliable in clinical populations such as young children or the elderly who cannot perform complex, tiring assessments. This causes two main problems: enormous healthcare expenditure for misdiagnosed disorders and severe deteriorations in patients that are not diagnosed in time. We are going to change this using artificial intelligence applied to eye movement recordings.
The quality of eye movements reflects the underlying state of the nervous system, as well as the integrity of visual functions. We combine high-speed oculography and machine learning to capture this relation and we translate it into actionable, easy to understand, medical insights. Our innovative algorithm extract biomarkers from the eye movements, allowing to indicate the possible presence of an underlying neurological condition and to perform visual assessments such as "perimetry" (i.e., the mapping of the retinal sensitivity in different locations of the visual field). In 5 minutes, we are able to test what normally would require a couple of hours of specialized assessment. Our method is based on signal processing techniques derived from neuroimaging and applied to eye movements recordings, and on artificial intelligence models based on deep recurrent neural networks.
