WEAR · Behaviour Phenotyping using Inertial Sensors

Behavior is ultimately the observable output of the nervous system. Thus, to properly diagnose and monitor nervous system disorders it is crucial that we assess behavior comprehensively. However, every day clinical practice still relies mostly on subjective methods to evaluate behavior. The ability to adequately analyse behaviour is also critical during drug development, from the preclinical (animal models) to the clinical stages. Results from our ERC-funded work revealed that the combination of inertial sensor data with an unsupervised algorithm provides an optimal method for an easy to implement unbiased behavior classification that adequately captured the outcome of neural circuit’s computations. In this PoC we proposed to develop this technology into a product that provides a continuous, quantitative and comprehensive behavior assessment that is also versatile, covering the diverse spectrum from the pre-clinical to the clinical context. In this POC we will 1) refinement and validate of inertial sensors in animal models, 2) enable their integration into a Body Area Network, 3) Evolve our unsupervised algorithm into a stand-alone software that is versatile and easy to use. In addition to these technical aims, we propose to explore commercial opportunities and societal benefits, in particular in the medical and drug development sector. We will conduct market analysis and develop a business case for this product, while expanding industry contacts for production and commercialization. Our proposal will lead to a ground-breaking technical solution for quantitative, automated behavioral assessment in animal models of disease and humans that will have an important societal impact through innovation in diagnosis, disease monitoring, and drug development.