Stefano Bastianini1, Chiara Berteotti1, Alessandro Gabrielli2, Flavia Del Vecchio3, Roberto Amici3, Chloe Alexandre4, Thomas E Scammell4, Mary Gazea5, Mayumi Kimura5, Viviana Lo Martire1, Alessandro Silvani1, Giovanna Zoccoli6. 1. PRISM Lab, Alma Mater Studiorum - University of Bologna, Bologna, Italy. 2. Department of Physics and Astronomy, Alma Mater Studiorum - University of Bologna, Bologna, Italy. 3. Physiological Regulation in Wake-Sleep Cycle Lab, Department of Biomedical and Neuromotor Sciences, Alma Mater Studiorum - University of Bologna, Bologna, Italy. 4. Department of Neurology, Beth Israel Deaconess Medical Center, Boston, MA, USA. 5. Max Planck Institute of Psychiatry, Munich, Germany. 6. PRISM Lab, Alma Mater Studiorum - University of Bologna, Bologna, Italy. Electronic address: giovanna.zoccoli@unibo.it.
Abstract
BACKGROUND: Scoring of wake-sleep states by trained investigators is a time-consuming task in many sleep experiments. We aimed to validate SCOPRISM, a new open-source algorithm for sleep scoring based on automatic graphical clustering of epoch distribution. METHODS: We recorded sleep and blood pressure signals of 36 orexin-deficient, 7 leptin knock-out, and 43 wild-type control mice in the PRISM laboratory. Additional groups of mice (n=14) and rats (n=6) recorded in independent labs were used to validate the algorithm across laboratories. RESULTS: The overall accuracy, specificity and sensitivity values of SCOPRISM (97%, 95%, and 94%, respectively) on PRISM lab data were similar to those calculated between human scorers (98%, 98%, and 94%, respectively). Using SCOPRISM, we replicated the main sleep and sleep-dependent cardiovascular findings of our previous studies. Finally, the cross-laboratory analyses showed that the SCOPRISM algorithm performed well on mouse and rat data. COMPARISON WITH EXISTING METHODS: SCOPRISM performed similarly or even better than recently reported algorithms. SCOPRISM is a very simple algorithm, extensively (cross)validated and with the possibility to evaluate its efficacy following a quick and easy visual flow chart. CONCLUSIONS: We validated SCOPRISM, a new, automated and open-source algorithm for sleep scoring on a large population of mice, including different mutant strains and on subgroups of mice and rats recorded by independent labs. This algorithm should help accelerate basic research on sleep and integrative physiology in rodents.
BACKGROUND: Scoring of wake-sleep states by trained investigators is a time-consuming task in many sleep experiments. We aimed to validate SCOPRISM, a new open-source algorithm for sleep scoring based on automatic graphical clustering of epoch distribution. METHODS: We recorded sleep and blood pressure signals of 36 orexin-deficient, 7 leptin knock-out, and 43 wild-type control mice in the PRISM laboratory. Additional groups of mice (n=14) and rats (n=6) recorded in independent labs were used to validate the algorithm across laboratories. RESULTS: The overall accuracy, specificity and sensitivity values of SCOPRISM (97%, 95%, and 94%, respectively) on PRISM lab data were similar to those calculated between human scorers (98%, 98%, and 94%, respectively). Using SCOPRISM, we replicated the main sleep and sleep-dependent cardiovascular findings of our previous studies. Finally, the cross-laboratory analyses showed that the SCOPRISM algorithm performed well on mouse and rat data. COMPARISON WITH EXISTING METHODS: SCOPRISM performed similarly or even better than recently reported algorithms. SCOPRISM is a very simple algorithm, extensively (cross)validated and with the possibility to evaluate its efficacy following a quick and easy visual flow chart. CONCLUSIONS: We validated SCOPRISM, a new, automated and open-source algorithm for sleep scoring on a large population of mice, including different mutant strains and on subgroups of mice and rats recorded by independent labs. This algorithm should help accelerate basic research on sleep and integrative physiology in rodents.
Authors: Ioannis Exarchos; Anna A Rogers; Lauren M Aiani; Robert E Gross; Gari D Clifford; Nigel P Pedersen; Jon T Willie Journal: Sleep Date: 2020-05-12 Impact factor: 5.849
Authors: Đorđe Miladinović; Christine Muheim; Stefan Bauer; Andrea Spinnler; Daniela Noain; Mojtaba Bandarabadi; Benjamin Gallusser; Gabriel Krummenacher; Christian Baumann; Antoine Adamantidis; Steven A Brown; Joachim M Buhmann Journal: PLoS Comput Biol Date: 2019-04-18 Impact factor: 4.475