Antenatal fetal surveillance is a field of increasing importance in modern obstetrics. Measurements extracted (such as fetal heart rate) from antenatal fetal monitoring techniques have the potential to reduce the social, personal and financial burdens of fetal death on families, health care systems and the community. Techniques to monitor the fetus through pregnancy have been developed with the aim of providing information to enable the clinician to diagnose fetal wellbeing, characterise development and detect abnormality. An early diagnosis before delivery may increase the effectiveness of the appropriate treatment. Over the years, various research efforts have been carried out in the field of fetal electrocardiography by attaching surface electrodes to the maternal body. Unfortunately the desired fetal heartbeat signals at the electrode output are buried in an additive mixture of undesired interference disturbances. In this thesis, a non-invasive fetal electrocardiogram machine has been designed, constructed and implemented. This machine is composed of three modified electrocardiogram circuits and an external soundcard. Data was acquired from four surface electrodes placed on the maternal body. Eleven pregnant subjects, with a gestation age between the 30th and 40th weeks of pregnancy, were used to investigate the validity of this machine. Fetal R-waves were detected in 72.7 percent of subjects. The development of a non-invasive machine, capable of detecting and recording valuable anatomic and electrophysiological information of a fetus, represents an important tool in clinical and investigative obstetrics.

Magister Curationis - MCur / Fetal surveillance (FS) helps midwives to predict the well-being of the fetus during labour and its relationship with uterine contractions as labour progresses. Incorrect use of surveillance apparatus and wrong interpretation of data from electronic fetal heart rate monitoring have been identified as contributing factors to delayed interventions that might have prevented the development of hypoxic-ischemic encephalopathy, which is the fifth largest cause of death of children under five globally.

A case-control study was performed to evaluate the Swedish population-based symphysis fundus (SF) reference curves. The study included 242 small for gestational age (SGA) neonates (169 term and 73 preterm infants) as cases and 296 non-SGA infants as controls. Two Swedish SF curves were evaluated. In term pregnancies they showed a sensitivity of 32 % and 51 % and a specificity of 90 % and 83 %, respectively, at a cut-off level of < - 2 SD from the mean according to the SF reference curve. The sensitivity for SGA was higher in preterm pregnancies (49 % and 58 %, respectively) and the first alarm below – 2 SD was noted before 32 weeks in 37 % and 43 % of the preterm pregnancies, respectively. (Study I) A study of self-administered SF measurements was designed to achieve more regular and frequent SF measurements. Thirty-three women with singleton, ultrasound dated pregnancies performed SF measurements on average 14 weeks from gestational week 20 to 25 until delivery. Self-administered SF measurements were higher and had higher variance than midwives’ measurements. Four consecutive SF measurements on each occasion can compensate for higher variance. Reliable self-administered SF measurements can be obtained. (Study II) Self-administered SF measurements from 191 women were used to construct absolute and relative SF growth references. The influence of fetal sex, maternal obesity and parity was assessed in regression models. The lnSF growth was statistically influenced by maternal obesity, and a borderline significance was recorded for fetal sex and parity. Statistical analysis and graphical displays show no evidence that the relative lnSF growth should be dependent on these variables. (Study III) To improve detection of infants with intrauterine growth restriction (IUGR) rather than SGA a new statistical model (the SR method) was used. The SR method was evaluated with SF measurements from 1122 pregnant women. The sensitivity for neonatal morbidity and SGA was low, between 6 and 36 % for SGA (< -2SD). Neonates classified as SGA (< -2SD and < 10th percentile) had increased morbidity compared with the total study group. Neonates suspected to be SGA before delivery by the population-based SF measurement method had lower morbidity than those not suspected. The SR method was found not to improve detection of fetuses with increased morbidity or SGA neonates in this study. Better screening methods to detect IUGR and SGA prior to delivery are needed. (Study IV)