Research Studentship

Noninvasive detection of fetal asphyxia using ultrasound and heart rate variability

Complications of pregnancy, labour and delivery represent the main risk for perinatal asphyxia [1]. The Confidential Enquiry into Stillbirths and Deaths in Infancy reported a mortality rate of 0.62 per 1000 births in England, Wales and Northern Ireland in 1999 [2].

Neonates asphyxiated at birth who develop hypoxic-ischaemic encephalopathy (HIE) have a very high risk of death (20-50%) and as many as 25 % of the survivors show signs of cerebral palsy with major motor-cognitive impairment [1,3,4]. While some deaths are unavoidable, it is thought that more can be done to reduce this level of mortality and early detection of signs of asphyxia can improve outcome by prompting appropriate medical intervention.

Monitoring of labour and delivery is normally performed with cardiotocography (CTG) and assessment of fetal oxygen and acid-base status. Despite the continuous fall in the incidence of HIE [5], there is ongoing discussion about the effective contribution of CTG in uncomplicated pregnancies, due to its reduced sensitivity and specificity [1,3,4]. Fetal distress leads to well known decelerations in fetal heart rate (FHR), but difficulties in interpretation of FHR patterns, and lack of standardization, contribute to the poor specificity of CTG [6-8]. As a result, a large number of false-positives are subject to unnecessary intervention, such as caesarean delivery, thus contributing to increased costs and risk of complications. On the other hand, difficulties of interpretation and limitations in the sensitivity of FHR patterns contribute to inappropriate response times for staff to detect asphyxia and take action, which is normally in the range of 30-100 min [6]. Response times of this order are obviously inadequate since permanent brain damage can result from only 10 min of severe asphyxia. Therefore, improvements in the sensitivity and specificity of CTG could lead to earlier detection of intrapartum asphyxia, more immediate reaction and also a reduction in the false-positive rate leading to more appropriate use of resources and less iatrogeny.

Recent advances in the analysis and interpretation of heart rate time series have the potential to improve the classical pattern analysis of FHR. In particular, autoregressive spectral analysis of pulse interval signals has been shown to reflect autonomic nervous system disturbances associated with stroke, brain death, diabetes, heart failure and myocardial infarction [9,10]. Application of these methods to perinatal asphyxia remains largely unexplored, but recently, in collaboration with Prof N. Thakor (Johns Hopkins School of Medicine, USA), we (FSS) performed signal processing of heart rate signals from anaesthetized rats, subjected to different lengths of asphyxia, and were able to detect its onset within 1 min., as well as to obtain an indication of its severity [11]. These preliminary results suggest that our technique might be more sensitive than classical interpretation of FHR patterns in response to asphyxia. Rather than assessing the classical decelerations in FHR associated with uterine contractions, our approach focuses on the more stable periods and could use the signal between contractions. This approach can also be potentiated by including newer techniques of non-linear dynamic analysis that can take into account the complete information available in the FHR tracing. Ideally, rather than using the scalp ECG as the source of the FHR signal, the use of Doppler ultrasound would reduce the risk of infection and allow wider automatic monitoring using our approach (since most maternity units have an ultrasound monitor). A separate investigation would need to be completed to test whether its temporal resolution would be acceptable and also the reliability of recordings made during delivery. Following a study on the quality of ultrasound records [12], our preliminary results indicate this to be feasible [13].

In addition to umbilical cord or scalp pH, intrapartum pulse oximetry has been proposed, since the early 90's, as a more specific method to assess hypoxemia. The disadvantages of these techniques though are their cost, risk of complications, and technical difficulties [14-17]. If the hypothesis that spectral and non-linear analyses of FHR signals can yield parameters that correlate with scalp pH or pulse oximetry data this could represent a major breakthrough by leading to new technology that could considerably simplify and reduce the costs of monitoring high-risk labour/delivery.

Supervisor: Dr. F.S. Schlindwein (fss1@le.ac.uk), URL: http://www.le.ac.uk/eg/fss1/

References

[1] Avery GB, Fletcher MA, MacDonald MG Neonatology. Pathophysiology and Management of the Newborn, 4^th Edition. Philadelphia: JB Lippincot Co., 1994.

[2] 1999 CESDI (Confidential Enquiry into stillbirths and deaths in infancy in Wessex) 8th Annual Report, The Wessex Institute for health research and development.

[3] Robertson CMT, Finer NN, Grace MGA. School performance of survivors of neonatal encephalopathy associated with birth asphyxia at term. J Pediatr 1989;114:753-60.

[4] Vannucci RC. Perinatal hypoxic-ischemic encephalopathy. The Neurologist, 1995;1:35-52.

[5] Smith J, Wells L, Dodd K. The continuing fall in incidence of hypoxic-ischaemic encephalopathy in term infants. British J Obst Gynaec 2000;101:461-466.

[6] NICH. Electronic fetal heart rate monitoring: research guidelines for interpretation. Am J Obstet Gynecol 1997;177:1385-90.

[7] Nelson KB, Dambrosia JM, Ting TY, Grether JK. Uncertain value of electronic fetal monitoring in predicting cerebral palsy. NEJM 1996; 334:613-618.

[8] Murphy KW, Johnson P, Moorcraft J, Pattinson R, Russel V, Turnbull A. Birth asphyxia and the intrapartum cardiotacograph. British J Obstet Gynaecol 1990; 97: 470-79.

[9] Malik M and Camm AJ (editors). Heart Rate Variability. Futura Publishing Co., Armonk, NY,1995.

[10] Freitas J, Puig J, Rocha AP, Lago P, Teixeira J, Carvalho MJ Costa O, Freitas A. Heart rate variability in brain death. Clin. Autonomic Res. 1996; 6:141-146.

[11] Boardman, Anita; Schlindwein, Fernando S.; Thakor, Nitish V.; Kimura, Tetsu; Geocadin, Romergryko G., Detection of asphyxia using heart rate variability, Medical & Biological Engineering & Computing, vol. 40, No.6, pp.618-624, November 2002.

[12] Dawes GS, Visser GH, Goodman JD. Numerical analysis of the human fetal heart rate: the quality of ultrasound records, Redman CW., Am. J. Obstetr. Gynecol, vol.141, pp.43-52, 1981.

[13] Anita Boardman, Fernando S. Schlindwein, Jason Waugh, Monitoring fetal and perinatal stress: the challenges and potential of using heart rate variability derived from ultrasound, Assessment of the active cardiovascular system, IPEM Meeting, York, 16 May 2003.

[14] Dildy GA, Clark SL, Loucks CA. Intrapartum fetal pulse oximetry: past, present, and future. Am J Obstet. Gynecol. 1996;175:1-9.

[15] Carbonne B, Langer B, Goffinet F, Audibert F, Tardiff D, et al. Multicenter study on the clinical value of fetal pulse oximetry. Am. J. Obstet Gynecol 1997; 177:593-598.

[16] Dildy GA, Thorp JA, Yeast JD, Clarck SL. The relationship between oxygen saturation and pH in umbilical blood: Implications for intrapartum fetal oxygen saturation monitoring. Am. J Obstet. Gynecol. 1996; 175:682-687.

[17] Kühnert M, Seelbach-Göebel B, Butterwegge M. Predictive agreement between the fetal arterial oxygen saturation and fetal scalp pH: results of the German multicenter study. Am J Obstet. Gynecol. 1998; 178:330-335.

[19] Solum T, Ingemarsson I, Nygren A, The accuracy of ultrasonic fetal cardiography, Journal of Perinatal Medicine, 9(1), pp.54-62, 1981.