The detection of abnormal neonatal cerebral haemodynamics using principal component analysis of the Doppler ultrasound waveform

Doppler ultrasound recordings were made from the anterior cerebral arteries of 40 normal full-term babies and 14 mature babies with intracranial pathology. The maximum velocity waveforms were extracted using spectral analysis and further analyzed using the Pourcelot resistance index method and the principal component method. Principal component analysis was found to be superior to the currently widely used resistance index method and may significantly enhance the value of Doppler ultrasound for detecting abnormal cerebral haemodynamics in neonates.