Normal and abnormal full-term and premature neonatal electroencephalograms (EEGs) will be studied by computer methods that are specifically appropriate for the special and unique patterns that are encountered in the EEGs of this age group, with the eventual aim of developing indices of normality/abnormality and degree of maturation in relation to conceptional age. In the first instance, the method of adaptive segmentation will be used for analysis, a method that automatically recognizes the occurrence of a chance in the type of EEG activity. Segments of similar types of activity are brought together (clustered), and a temporal profile indicates the particular type of activity that obtained at any particular time during the recording. Power spectra of each principal type of activity can be computed if desired. The EEG recordings will be drawn from major centers having a particular interest in neonatal electroencephalography, either in the form of cassette tape recordings made simultaneously with the original clinical EEG recording, or in the form of ink recordings of the EEGs which will be reconverted to electrical form by means of a multi-channel photo-optical scanner. Attention will initially be directed to the normal full-term EEG. Although applied with considerable success in recent related work to a wide range of normal and abnormal adult EEG patterns, the method of adaptive segmentation has some limitations, so that a simpler but equally effective alternative approach will also be developed and evaluated for this project. Moreover, separate methods will be developed for automatic evaluation of particular aspects of the neonatal EEG, e.g., interhemispheric synchrony of """"""""burst"""""""" activity during particular phases of sleep, and the frequency of occurrence of a pattern unique to the neonatal EEG, """"""""delta brushes."""""""" Neonatal electroencephalography, which can be considered a special branch of neonatology, is an area of considerable importance in clinical electroencephalography that has come into its own only in relatively recent years, so that in some instances, norms are based on relatively small data bases since laborious manual determinations are entailed. At the same time, computer applications in neonatal EEG have been of little help with the establishment of such norms. This is one of the prime objectives of this project.
| Qian, J; Barlow, J S; Beddoes, M P (1988) A simplified arithmetic detector for EEG sharp transients--preliminary results. IEEE Trans Biomed Eng 35:11-8 |
| Barlow, J S (1985) Computer characterization of trace alternant and REM sleep patterns in the neonatal EEG by adaptive segmentation--an exploratory study. Electroencephalogr Clin Neurophysiol 60:163-73 |
| Moss, N G (1985) Electrophysiology of afferent renal nerves. Fed Proc 44:2828-33 |
