Sleep Analysis Software Sleep Sign for Animal We

Sleep Analysis Software Sleep. Sign for Animal We are developing a New Automatic Scoring Algorithm! Algorithm for sleep scoring in experimental animals based on fast Fourier transform power spectrum analysis of the electroencephalogram Collaborating with Osaka Bioscience Institute, Japan KISSEI COMTEC Co. , LTD. ( JAPAN )

Accepted Algorithm for sleep scoring in experimental animals based on fast Fourier transform power spectrum analysis of the electroencephalogram Sayaka Kohtoh, 1 Yujiro Taguchi, 1 Masashi Wada, 2 Naomi Matsumoto, 2 Zhi-Li Huang, 2 and Yoshihiro Urade 2 1 Department of Medical Systems, Kissei Comtec Co. , Ltd. , Nagano, and 2 Department of Molecular Behavioral Biology, Osaka Bioscience Institute, Osaka, Japan Sleep Biological and Rhythms Volume 6 Number 3 July 2008

Joint project with Osaka Bioscience Institute (OBI), Japan ØA top notch research institute dedicating to the sleep mechanism ØRanked as having one of the highest ratios of scientific Papers citing in the world ØDiscovered Prostaglandin D 2 in 2002 ØAn introduction of OBI appeared in ASBMB （ October 2006, P 10～ 12 ） http: //www. obi. or. jp/

Advantages Ø We selected 4 parameters EEG δ power , EEG θ / (δ+θ) ratio , EMG integral , Activity Ø Percent agreement 90. 9 ± 4. 0% for rats , 90. 0 ± 3. 2% for mice Ø Processing speed is 1/6 of the existing version Ø Multiple animals are scored at once

New Automatic Scoring Algorithm Activity EEG EMG Fast Fourier Transform Integral Scoring Logic Wake Non REM

Typical waveform and FFT spectrum Active wake EEG LOC FFT spectrum rat EMG EEG LOC FFT spectrum mouse EMG Quiet wake REM sleep NREM sleep

Non REM REM Non. REM REM Wake EEG q/(d+q) ratio (%) mouse rat 3 parameters to separate NREM, and Wakefulness EEG d power (m. V 2) EMG integral (m. V/sec)

Scoring Logic EEG δ wave Locomotor EEG θ wave Step : 1 Yes Activity ? No EEG EMG Step : 2 Active Wake Yes High EEG d power ? No EEG / EMG Step : 3 NREM Sleep Yes W R High EEG θ ratio and High EMG Integral No NR Hypnogram REM Sleep Quiet Wake

Time courses of parameters and hypnogram rat Dark phase EMG integral (m. V/sec) EEG q/(d+q) ratio(%) EEG d power (m. V 2) Locomotion activity (count) Light phase W R NR 12: 00 14: 00 16: 00 24: 00 Clock time 02: 00 04: 00

Time courses of parameters and hypnogram mouse Dark phase EMG integral (m. V/sec) EEG q/(d+q) ratio(%) EEG d power (m. V 2) Locomotion activity (count) Light phase W R NR 14: 00 16: 00 18: 00 22: 00 Clock time 24: 00 02: 00

Percent agreement between FFT algorithm and visuals rat ( n = 10 ) mouse ( n = 23 ) % Wake 93. 4± 5. 6 91. 6± 4. 8 REM 82. 5± 2. 4 65. 0± 19. 0 NREM 85. 0± 3. 7 91. 7± 6. 7 All States 90. 9± 4. 0 90. 0± 3. 2 %

Processing time in pooled 24 hr-data from 8 rats(min) Processing time 250 200 1/6 150 100 4 hr 1 / 80 50 0 Algorithm Waveform recognition Filtered ○ 40 min 3 min FFT ○ ×

: Quick Report Making in Excel SD Rat_2 SD Rat_1 After Scoring Control SD Rat_4 SD Rat_3 SD Rat_1. csv SD Rat_2. csv SD Rat_3. csv SD Rat_4. csv