Validation of a new data-driven automated algorithm for muscular activity detection in REM sleep behavior disorder

Abstract

 Users may download and print one copy of any publication from the public portal for the purpose of private study or research.  You may not further distribute the material or use it for any profit-making activity or commercial gain  You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

External validation of a data‐driven algorithm for muscular activity identification during sleep

“…A more detailed explanation can be found in Cesari et al. (). (a) Training of the probabilistic models of atonia.…”

“…Figure a,b shows how the method was developed on Danish data (more details can be found in Cesari et al., and in the Figure caption). First, for each EMG signal (chin, TIBL and TIBR), we trained three probabilistic models of atonia (Figure a), using the assumption that HCs show atonia in REM sleep.…”

“…MAFs were extracted from each participant using the same procedure, and the features were then fed into the five previously trained classifiers and from each of them the three output probabilities were obtained. In order to generalize the algorithm application, we interpreted the probabilities (a) p [HC], (b) p [iRBD] and (c) p [PLMD] as follows: p [RBD(−)PLMS(−)] indicates the probability that a participant shows neither RBD‐like muscular activity nor increased PLMS index, as Danish HCs had neither RBD nor increased PLMS index; p [RBD(+)] indicates the probability that a participant shows RBD‐like muscular activity, without any further specification on PLMS, as 12 out of 29 Danish iRBD patients showed increased PLMS index (Cesari et al., ); and p [RBD(−)PLMS(+)] indicates the probability that a participant shows increased PLMS index without interference of RBD‐like muscular activity, as all Danish PLMD patients showed increased PLMS index and no RBD. …”

“…In our previous study, we found that the method achieved the highest performances without the need for removing muscular activity related to apneas and arousals; therefore, we also did not remove this type of activity in the German cohort (Cesari et al., ). Moreover, we applied the algorithm with the optimal set of parameters found in Cesari et al.…”

“…This study aims to validate our data‐driven algorithm for muscular activity detection during sleep (Cesari et al., ) using data recorded in a different centre and to investigate the inter‐clinical variability by analysing the differences in the algorithm performance. The results of this study show the applicability of our algorithm and further confirm that data‐driven methods can contribute to advancing research and clinical applications in sleep medicine.…”