nViso has obtained a grant from the Swiss Commission for Technology and Innovation (CTI) to extend its facial analysis technology to provide new levels of predictive analytics in healthcare. The project, valued at SFr 700,000 (US$750,000), will run for 18 months and initially focus on exploitng nViso's artificial intelligence-based face imaging capabilities to more accurately categorise patients who require tracheal intubation for surgeries involving general anaesthesia.
Initial results and a demonstration of the pilot project, which involves the anaesthesiology department at Lausanne University Hospital (CHUV) under the responsibility of Dr Patrick Schoettker, and the Signal Processing Laboratory of the Ecole Polytechnique Fédérale de Lausanne (EPFL) led by Prof. Jean-Philippe Thiran, will be displayed at the World Medtech Forum in Lucerne, Switzerland. The results of the pilot project, which collected data from 800 patients, include a 97 per cent classification accuracy in predicting the Mallampati score for patients requiring tracheal intubation, a score used by anaesthesiologists to predict difficult airway management.
nViso's image processing and computer vision algorithms, which to date have been primarily used to analyse emotional responses to marketing stimuli, will be extended to evaluate dynamic features of the face, mouth, and neck, which are critical to predicting potential intubation problems. The recorded biometric data is analysed with mathematical algorithms to generate a prediction score that is more accurate and efficient than existing methods.