0432 A Deep Learning Approach for Automated Sleep-Wake Scoring in Pre-Clinical Animal Models

Abstract Introduction Experimental investigation of sleep-wake dynamics in animals is an important part of pharmaceutical development. It typically involves recording of electroencephalogram, electromyogram, locomotor activity, and electrooculogram. Visual identification, or scoring, of the sleep-wake states from these recordings is time-consuming. We sought to develop software for automated sleep-wake scoring capable of processing large databases of multi-channel signal recordings in a range of animal species. Methods We used a large historical database of signal recordings and scores in non-human primates, dogs, mice, and rats, to develop a deep Convolutional Neural Network (CNN) classification algorithm for automatically scoring sleep-wake states. We compared the performance of the CNN algorithm with that of a widely used Machine Learning algorithm, Random Forest (RF). Results In non-human primates and dogs, CNN accuracy in sleep-wake scoring of data was significantly higher than RF accuracy: 0.75 versus 0.66 for non-human primates and 0.73 versus 0.64 for dogs. In rodents, the difference between CNN and RF was smaller: 0.83 versus 0.81 for mice and 0.78 versus 0.77 for rats. The variability of CNN accuracy was lower than that of RF for non-human primates, dogs and mice, but similar for rats. Conclusion We recommend use of CNN for sleep-wake scoring in non-human primates and dogs, and RF for sleep-wake scoring in rodents. Support Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA