REM behaviour disorder detection associated with neurodegenerative diseases
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
REM behaviour disorder detection associated with neurodegenerative diseases. / Kempfner, Jacob; Sorensen, Gertrud; Zoetmulder, Marielle; Jennum, Poul; Sorensen, Helge B D.
In: I E E E Engineering in Medicine and Biology Society. Conference Proceedings, Vol. 2010, 01.01.2010, p. 5093-6.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - REM behaviour disorder detection associated with neurodegenerative diseases
AU - Kempfner, Jacob
AU - Sorensen, Gertrud
AU - Zoetmulder, Marielle
AU - Jennum, Poul
AU - Sorensen, Helge B D
PY - 2010/1/1
Y1 - 2010/1/1
N2 - Abnormal skeleton muscle activity during REM sleep is characterized as REM Behaviour Disorder (RBD), and may be an early marker for different neurodegenerative diseases. Early detection of RBD is therefore highly important, and in this ongoing study a semi-automatic method for RBD detection is proposed by analyzing the motor activity during sleep. Method: A total number of twelve patients have been involved in this study, six normal controls and six patients diagnosed with Parkinsons Disease (PD) with RBD. All subjects underwent at least one ambulant polysomnographic (PSG) recording. The sleep recordings were scored, according to the new sleep-scoring standard from the American Academy of Sleep Medicine, by two independent sleep specialists. A follow-up analysis of the scoring consensus between the two specialists has been conducted. Based on the agreement of the two manual scorings, a computerized algorithm has been attempted implemented. By analysing the REM and non-REM EMG activity, using advanced signal processing tools combined with a statistical classifier, it is possible to discriminate normal and abnormal EMG activity. Due to the small number of patients, the overall performance of the algorithm was calculated using the leave-one-out approach and benchmarked against a previously published computerized/visual method. Results: Based on the available data and using optimal settings, it was possible to correctly classify PD subjects with RBD with 100% sensitivity, 100% specificity, which is an improvement compared to previous published studies. Conclusion: The overall result indicates the usefulness of a computerized scoring algorithm and may be a feasible way of reducing scoring time. Further enhancement on additional data, i.e. subjects with idiopathic RBD (iRBD) and PD without RBD, is needed to validate its robustness and the overall result.
AB - Abnormal skeleton muscle activity during REM sleep is characterized as REM Behaviour Disorder (RBD), and may be an early marker for different neurodegenerative diseases. Early detection of RBD is therefore highly important, and in this ongoing study a semi-automatic method for RBD detection is proposed by analyzing the motor activity during sleep. Method: A total number of twelve patients have been involved in this study, six normal controls and six patients diagnosed with Parkinsons Disease (PD) with RBD. All subjects underwent at least one ambulant polysomnographic (PSG) recording. The sleep recordings were scored, according to the new sleep-scoring standard from the American Academy of Sleep Medicine, by two independent sleep specialists. A follow-up analysis of the scoring consensus between the two specialists has been conducted. Based on the agreement of the two manual scorings, a computerized algorithm has been attempted implemented. By analysing the REM and non-REM EMG activity, using advanced signal processing tools combined with a statistical classifier, it is possible to discriminate normal and abnormal EMG activity. Due to the small number of patients, the overall performance of the algorithm was calculated using the leave-one-out approach and benchmarked against a previously published computerized/visual method. Results: Based on the available data and using optimal settings, it was possible to correctly classify PD subjects with RBD with 100% sensitivity, 100% specificity, which is an improvement compared to previous published studies. Conclusion: The overall result indicates the usefulness of a computerized scoring algorithm and may be a feasible way of reducing scoring time. Further enhancement on additional data, i.e. subjects with idiopathic RBD (iRBD) and PD without RBD, is needed to validate its robustness and the overall result.
U2 - 10.1109/IEMBS.2010.5626212
DO - 10.1109/IEMBS.2010.5626212
M3 - Journal article
VL - 2010
SP - 5093
EP - 5096
JO - Annual International Conference of the IEEE Engineering in Medicine and Biology - Proceedings
JF - Annual International Conference of the IEEE Engineering in Medicine and Biology - Proceedings
SN - 0589-1019
ER -
ID: 34190215