Abstract
In this paper, we investigate an advanced monitoring system for a neonatal intensive care unit. The system intelligently detects abnormal neonatal cerebral Doppler ultrasound signals by means of principal component analysis and a non-normalised compensatory neuro-fuzzy rule based algorithm. Two hundred and ninety doppler ultrasound signals were recorded from the anterior cerebral arteries of 40 normal full-term babies and 14 mature babies with intracranial pathology. The features of the normal and abnormal groups were extracted from the maximum velocity waveforms using a principal component method. The non-normalised compensatory neuro-fuzzy rule based algorithm yielded the highest predictive accuracy of 76.21%. These results show that the proposed algorithm is superior to others, and could potentially be used to build an intensive neonatal care unit system for the intelligent detection of abnormal neonatal cerebral haemodynamics.
| Original language | English |
|---|---|
| Article number | 164 |
| Pages (from-to) | 1612-1614 |
| Number of pages | 3 |
| Journal | Annual International Conference of the IEEE Engineering in Medicine and Biology - Proceedings |
| Volume | 2 |
| DOIs | |
| Publication status | Published - 2001 |
| Externally published | Yes |
Keywords
- Blood flow velocity
- Compensatory fuzzy neural networks
- Decision-making systems
- Doppler ultrasound
- Neonatal cerebral arteries
- Pattern classification
- Principal component analysis
