Automatic analysis of EEG pattern during sleep in Cheyne-Stokes respiration in heart failure

Cheyne-Stokes respiration (CSR) is a common sleep related breathing disorder in chronic heart failure (CHF), characterised by repeated hypopnoea or apnoea phases or both. Tthe associated transient changes in electroencephalography (EEG) may be hardly evaluated with standard sleep scoring techniques.To overcome these limits, we analysed EEG activityduring sleep using a new algorithm of spectral analysis: the Generalised Short Time Fourier Transform (GSTFT), representing a particular method of multiresolution analysis, optimising frequency and time resolution in order to discriminate EEG frequency bands.