PHYSIP is a European research unit developing innovative algorithms in the biomedical field using digital signal processing. Our main activity is the analysis of the electroencephalogram (EEG) for the study and monitoring of sleep and vigilance. Check our website .

Automatic Sleep EEG Analysis.

Physip researchers have been working on automatic sleep analysis for more than 15 years and our research is subject to ongoing clinical validation and publication.

It is single-channel based, using only the EEG signal, and only one EEG channel. It works and it has been clinically validated (see Aseega Performance Questions below).

Aseega is a single-channel sleep analysis method whereas the visual rules apply to poly-channel data. Nevertheless, Aseega follows the new AASM scoring rules but also provides R&K hypnograms.

No, the sleep analysis we provide is intended to be used as a scoring aid; it is not intended for the diagnosis or treatment of sleep disorders.

No, our service gives access to a sleep analysis scoring aid tool, it does not provide a diagnosis. Although our research is supported by an eminent medical advisory panel,we differentiate clearly between the two areas of competence.

No, only a single EEG channel.

Yes. Since Aseega involves neither expert prior tuning nor algorithm data-based learning rules, the recordings are systematically processed in the same way. If you input exactly the same EEG recording twice you will obtain exactly the same results.

We can analyze EDF and EDF+ format recordings (European Data Format ). Most if not all PSG software can export recorded data to EDF format. . We can also analyze some proprietary formats, contact us for more information.

Yes, as long as they meet the Aseega EEG requirements .

Yes, as long as they meet the EEG requirements, see previous question.

This recording did not meet Aseega recording requirements, either in terms of signal resolution, quality or duration. Rejected recordings are not invoiced.

Yes, 4 hours. Aseega uses this time to tune its auto-learning analysis features.

You need a computer with Internet access and a browser with Javascript, cookies and Flash enabled (they are enabled in a standard Windows installation). Supported browsers include Internet Explorer 6+, Firefox 1.5+ and Safari 2+.

Yes.

Yes, all results are provided in MATLAB-compatible, ASCII or Excel files.

We have produced guides to help you.

Yes, a quotation will be given systematically for approval before analysis begins. The final amount may be lower than this quotation if Aseega rejected recordings because of non-compliance with our recording recommendations.

Yes, preferential rates for volumes are available.

No, but there are payment advantages for volume uploads.

Your personal and organization data are kept strictly private. They will not be sold or given to any other entity (see Privacy Policy).

No. All you need is an account on AseegaOnline.

Yes, your login password will be issued after the setting up of your user profile.

Access the create account page and supply the details required for your user profile necessary for secure data transfer and invoicing.

We use the ultra-secure data transfer protocol SFTP. Encryption is carried out with a 2048-bit server key. Your information can only be viewed by authorized individuals.

No.

REM sleep information is partly conveyed in high EEG frequencies ([16-40] Hz). The distinction between REM sleep and Wake or sleep stage N1 is then possible by using accurate signal processing algorithms, assuming that the recording conditions meet the Aseega recording requirements.

The first clinical validation results can be found in: Automatic Analysis of Single-Channel Sleep EEG: Validation in Healthy Individuals, Berthomier et al. (SLEEP 2007; 30(11):1587-95 ). The obtained agreements between automatic and an expert-consensual scoring were 96.0% (wake/sleep scoring), 92.1% (wake/REM/NREM scoring), 84.9% (wake/REM/stages 1-2/SWS scoring), and 82.9% (wake/REM/stage 1/stage 2/SWS scoring). Finally, when classifying into 5 states, the sensitivity and positive predictive value of ASEEGA regarding wakefulness were 82.5% and 89.7%, respectively. Similarly, sensitivity and positive predictive value regarding REM state were 83.0% and 89.1%. More results are about to be published on new cohorts and with new visual reference.

These results have proved to be stable when the recording conditions are stable. You need to check the Aseega EEG requirements . In addition, epoch-by-epoch comparison between scorings is very delicate in terms of epoch timestamp. Especially rounding the sample frequency (Fe) will introduce epoch-start drift. For example, by rounding Fe from 200,01 Hz to 200,00 Hz, a shift between the compared epochs of 10 seconds (a third of an epoch) will occur after 8 hours of recording.