Brain dysmaturity as an index for the detection of infants considered to be at risk for apnea

We propose the definition of an EEG based brain dysmaturity index that coul d allow an automatic detection of neonates that deviate from expected ontog enetic patterns. The investigation was carried out in a group of 107 term a nd preterm born infants of 28-112 weeks of postconceptional age (PCA). For each neonate, one continuous 2-channel EEG of 1-6 hours duration was record ed. The cluster analysis of different age groups was performed with a self- referential neural network. The network performed a nonlinear discriminant analysis; the synaptic strength of input nodes indicates the relevance of a n individual EEG feature. The most relevant EEG features are given by the a verage amplitude in the Delta and in the Theta band and by the quotient of the amplitudes Alpha/Theta and Beta 1/Theta. respectively. The correlation between frequency shifts and postconceptional age agrees with measures of b rain dysmaturity in healthy preterm neonates (Scher 1997). Thus, the presen ted trend in early EEG development shows that it is possible to establish c linically relevant age dysmaturity scores.