Decision of normal and growth retarded pregnancies by discriminant functions using umblical arterial blood flows velocity waveforms
Advances in Intelligent Systems and Computer Science, ss.186-190, 1999 (Scopus)
- Yayın Türü: Makale / Tam Makale
- Basım Tarihi: 1999
- Dergi Adı: Advances in Intelligent Systems and Computer Science
- Derginin Tarandığı İndeksler: Scopus
- Sayfa Sayıları: ss.186-190
- Anahtar Kelimeler: Artificial neural networks, Discriminant functions, Doppler measurements, Growth retarded pergnancies, Hypoxia, Umblical artery
- Boğaziçi Üniversitesi Adresli: Evet
Özet
Our study employs discriminant function based classifiers to make a decision about the health conditions of fetuses using the measurements by Doppler ultrasound. We use the measurements of umblical artery (UA) blood flow velocity waveforms and derive linear, quadratic and a nonlinear discriminant functions to learn and to decide the status of fetus. The measured parameters are gestational age in terms of weeks (week index: WI as a normalized value), pulsality index (PI), resistance index (RI) and Systolic/Diastolic ratio (S/D). With these indices and limited data, a decision success rate of up to 93.65 % is achieved. It has been demonstrated that linear disriminant analysis (LDA), quadratic disriminant analysis (QDA) and artificial neural network (ANN- nonlinear disriminant analysis) may be a reasonable medical aid to phycisians in diagnosis of fetal hypoxia.