Estimating a likelihood of neonatal severe compromise evident at birth

Abstract

A method for estimating a likelihood of neonatal severe compromise evident at birth includes receiving 302 fetal heart rate data of a fetus from a fetal monitor. At least a portion of the fetal heart

Description

[0001] A key challenge in the field of maternity-related healthcare is to reduce the rate of stillbirths, neonatal deaths and brain injuries obtained during or around childbirth. Fetal heart rate monitoring is regularly performed during childbirth to monitor a baby for signs of distress. However, the visual interpretation of heart rate data by clinicians is subject to high interobserver variability due to the inherent complexity of heart rate patterns.

[0002] Furthermore, it is difficult for clinicians to analyse the heart rate data in combination with other potential indicators of fetal wellbeing, since these factors and the relative weightings which should be assigned to the factors are not fully understood at this time. Such challenges naturally lead to less accurate judgements regarding a condition of the baby, and potentially negative outcomes for both mother and baby.

[0003] It is in this context that the invention is devised. BRIEF SUMMARY

[0004] In accordance with the present invention, there is provided an apparatus, a system and a computer-implemented method for estimating a likelihood of neonatal severe compromise evident at birth.

[0005] According to a first aspect, there is provided a computer-implemented method for estimating a likelihood of neonatal severe compromise evident at birth. The method includes receiving fetal heart rate data of a fetus. The fetal heart rate data is obtained from a fetal monitor and at least a portion of the fetal heart rate data is obtained by the fetal monitor within a predetermined time window of labour onset. The method also includes receiving risk factor data corresponding to the fetus. The risk factor data includes one or more of a maternal characteristic of a mother of the fetus, a characteristic of the fetus, a characteristic of the labour, and a characteristic of the pregnancy. The method also includes determining, using a data-driven computational model, a neonatal severe compromise risk score based on the received fetal heart rate data and the risk factor data. The method also includes outputting the determined neonatal severe compromise risk score. The determined neonatal severe compromise risk score is indicative of a likelihood of the fetus presenting with severe compromise at birth.

[0006] Advantageously, this method allows fetuses who have a higher risk of presenting with severe compromise at birth to be identified in advance. Standard fetal monitoring methods provide a clinician with an indication of a current condition of the fetus, from which the clinician may determine whether the fetus is currently in distress. However, the method of the present invention provides an estimation, using data obtained at or around an onset of labour, of a likelihood of the fetus being severely compromised at birth. That is, the present invention does not simply provide a snapshot of a current condition of a fetus, but provides an estimation of a future condition of the fetus.

[0007] Further advantageously, this method combines the fetal heart rate data with other risk factor data to provide an individualised and more accurate estimation of the fetus presenting with severe compromise at birth. Current visual interpretation methods of analysing fetal heart rate data are unable to reliably and consistently take into account the effects of prognostic risk factors relating to the mother, the fetus, the pregnancy, or the labour. The method of the present invention does consider, and weight, the input risk factors along with the fetal heart rate data to provide a more accurate and reliable risk score in a way which is not possible for a human clinician.

[0008] Using the determined neonatal severe compromise risk score, a clinician is able to make a more informed decision regarding whether to intervene with the labour or allow the labour to progress naturally. Moreover, the clinician is able to make this decision at an earlier stage of the labour, and even before the fetus begins to show typical symptoms of distress. This may result in an improved outcome for both the baby and the mother.

[0009] For example, if the risk score indicates that a fetus has a high likelihood of presenting with severe compromise at birth, the clinician may elect to perform a Caesarian section (or other assisted delivery method) at an earlier stage of labour. By performing this earlier intervention, the likelihood that the health of the baby will be damaged is reduced so subsequent long term health complications to the baby may be prevented. In addition, the earlier intervention enables the mother to have a more positive birth experience by avoiding the need for urgent intervention.

[0010] As another example, if the risk score indicates that a fetus has a low likelihood of presenting with severe compromise at birth, the clinician may choose to allow the labour to progress naturally. This may prevent the mother from unnecessarily undergoing major surgery without unduly compromising the health of the baby.

[0011] Furthermore, health complications originating from preventable obstetric complications provide a long term financial and time burden on health care services. By providing an estimated likelihood of a fetus being severely compromised at birth and taking appropriate action based on the estimation, the burden on health care services may be reduced.

[0012] Further advantageously, this method provides a systematic approach to fetal monitoring. The risk score does not rely on subjective judgements from clinicians, so results are consistent regardless of the equipment and clinician involved in the labour process. Patient care may therefore be improved and consistent nationally and internationally.

[0013] Optionally, the maternal characteristic comprises one or more of a maternal parity, maternal age, maternal body temperature, a maternal underlying health condition, a maternal ethnicity, a maternal body mass index or a maternal medical test result.

[0014] Optionally, the characteristic of the fetus comprises one or more of a fetal gestational age, a fetal underlying health condition, a fetal medical test result, or a sleep state cycle of the fetus.

[0015] Optionally, the characteristic of the labour comprises one or more of a type of onset of the labour and a thickness of meconium.

[0016] Advantageously, each of the maternal characteristics, the fetal characteristics, and the labour characteristics can be indicative of a likelihood of a fetus suffering severe compromise during childbirth. Incorporating one or more of these factors into the determination therefore provides a more accurate estimation of the severe compromise risk score so the clinician can make more informed decisions regarding the progression of the labour. Moreover, it is not currently understood how a clinician can themselves consider these characteristics in conjunction with the fetal heart rate data to determine a likelihood of the fetus being severely compromised. The present method therefore provides the clinician with an improved overall estimation of a fetus suffering severe compromise. Moreover, each of these characteristics is readily available, at least in most standard clinical care settings in the developed world. The clinician therefore does not need to perform additional tests to obtain the necessary data for the determination of the risk score.

[0017] Optionally, determining the fetal severe compromise risk score comprises extracting one of more fetal heart rate characteristic from the fetal heart rate data. The fetal heart rate characteristic comprises one or more of a baseline fetal heart rate, a short-term variability of the fetal heart rate, an acceleration number or amplitude, a deceleration capacity of the fetal heart rate, a number of prolonged decelerations of the fetal heart rate within a second predetermined time period, a number of fetal heart rate accelerations and a size of the fetal heart rate accelerations.

[0018] Advantageously, these fetal heart rate characteristics provide an indication of a current health of the fetus and a likelihood of the fetus presenting with severe compromise at birth. The use of this data is currently open to interpretation by clinicians, so the present method enables the data to be used in an objective manner to reliably and accurately obtain a risk score for the fetus.

[0019] Optionally, the data-driven computational model is a model fitted to data relating to a plurality of prior births, optionally wherein the plurality of prior births comprises at least 1000 prior births, 10000 prior births, 25000 prior births or 50000 prior births.

[0020] Advantageously, the use of real data provides more accurate, evidence-based estimations of a likelihood of the fetus suffering severe compromise.

[0021] Optionally, the data-driven computational model is configured to process the received fetal heart rate data and the categorical risk factor data according to: Risk score = —-—> where 1 + e* x = c^c^Mr + c^Np + c^ + c^PED + c^Npd + c^((Aa + c9) / cw) T eu*Dc + ci2Mn + c^((STV / cu)-2) + (c15*((G / c16)3)) + (C17*(((7 / Cis)'> )*log((?) / Cig)) + C2l*LK + C2i*(B / C22) + C23*Tm where Mt = maternal temperature, Np = nulliparity, Ma = maternal age, Ped = presence of pre-eclampsia or diabetes, Npd = number of prolonged decelerations of fetal heart rate, Aa = acceleration amplitude of fetal heart rate, Dc = deceleration capacity of fetal heart rate, An = number of accelerations of fetal heart rate, STV = short term variability of fetal heart rate, G = gestational age of fetus, Ls = labour stage, B = fetal baseline heart rate, Tm = presence of thick meconium and ci to C23 are constants.

[0022] Advantageously, this formula provides a good mapping of the obtained risk factor data and the likelihood of the fetus presenting with severe compromise at birth.

[0023] Optionally, the method further comprises outputting an alert if the determined risk score is greater than a predetermined threshold value.

[0024] Advantageously, this alerts the clinician if the risk of the fetus suffering severe compromise is above a threshold risk level. This enables the clinician to take immediate action to mitigate this risk and thereby improve the health outcome for the baby.

[0025] Optionally, the predetermined time period is a period of up to five hours before or after the onset of the labour.

[0026] Advantageously, using fetal heart rate and / or the clinical risk factors data at or around an onset of labour enables clinical decisions to be made at an earlier stage of the labour, and before urgent intervention becomes required. With earlier intervention, the fetus may avoid being severely compromised, resulting in significantly improved health outcomes for the fetus.

[0027] Optionally, the fetal heart rate data is recorded over a period of at least 5 minutes. Further optionally, the fetal heart rate data is recorded over a period of at least 15 minutes. Shorter fetal heart rate data collection times may be advantageous for ease of use, and obtaining a risk score quickly. Longer fetal heart rate data collection times may be advantageous for obtaining more accurate risk scores.

[0028] Optionally, the method further comprises receiving uterine contraction data relating to a plurality of uterine contractions, where the uterine contraction data is taken within the predetermined time period of the onset of labour. Determining the fetal severe compromise risk score further comprises processing the received uterine contraction data using the data-driven computational model.

[0029] According to another aspect of the invention, there is provided an electronic device for estimating a likelihood of neonatal severe compromise evident at birth. The electronic device includes a first receiving means configured to receive fetal heart rate data of a fetus. The fetal heart rate data is obtained from a fetal monitor and at least a portion of the fetal heart rate data is obtained by the fetal monitor within a predetermined time window of labour onset. The electronic device also includes a second receiving means configured to receive risk factor data corresponding to the fetus. The risk factor data includes one or more of a maternal characteristic of a mother of the fetus, a characteristic of the fetus, a characteristic of the labour, and a characteristic of the pregnancy. The electronic device also includes a processor and a memory storing instructions. When executed by the processor, the instructions cause the processor to determine, using a data-driven computational model, a neonatal severe compromise risk score based on the received fetal heart rate data and the risk factor data. The instructions also cause the processor to output the determined neonatal severe compromise risk score. The determined neonatal severe compromise risk score is indicative of a likelihood of the fetus presenting with severe compromise at birth.

[0030] According to a further aspect of the invention, there is provided a system for estimating a likelihood of neonatal severe compromise evident at birth. The system comprises an electronic device and a fetal monitor for obtaining fetal heart rate data. BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0031] To easily identify the discussion of any particular element or act, the most significant digit or digits in a reference number refer to the figure number in which that element is first introduced.

[0032] FIG. 1 illustrates an aspect of the subject matter in accordance with one embodiment.

[0033] FIG. 2 illustrates an aspect of the subject matter in accordance with one embodiment.

[0034] FIG. 3 illustrates an aspect of the subject matter in accordance with one embodiment. DETAILED DESCRIPTION

[0035] FIG. 1 shows an electronic device 100 for obtaining an estimated likelihood of a fetus presenting with severe compromise at birth. The electronic device 100 may also be referred to as an apparatus.

[0036] Severe compromise is a clinical term indicating that a newborn baby is in serious condition and requires immediate medical attention. A baby presenting with severe compromise may need to be admitted to a neonatal intensive care unit for specialised care. The severely compromised baby may require treatment such as surgery, oxygen therapy or medication. Outcomes such as fetal sepsis, haemorrhage, fetal or neonatal death within the first 28 days of life, brain injury, and injuries caused by oxygen deprivation, may be considered severe compromise. Severe compromise may indicate that the baby has medium to long term health complications. Medium to long term may be a time ranging from a few weeks, a few months, a few years, or life-long. The health complications may affect one or both of the physical and mental health of the baby over the medium to long term duration.

[0037] The electronic device 100 comprises a first receiving means 102 for receiving fetal heart rate data of a fetus. At least a portion of the fetal heart rate data is obtained by a fetal heart rate monitor within a predetermined time window of labour onset. The predetermined time window may be a window of 1,2, 3, 4 or 5 hours either side of labour onset. The first receiving means 102 may comprise an antenna, receiver, or transceiver for wirelessly receiving the fetal heart rate data from the fetal heart rate monitor. The first receiving means 102 may receive the fetal heart rate data through a wired connection to the fetal heart rate monitor.

[0038] Labour onset may occur at the point when a woman in labour is deemed to be in active labour rather than latent labour. Labour onset may occur when a dilation of the cervix reaches a threshold dilation. The threshold dilation may be 2cm, 3cm, 4cm or 5cm. Labour onset may occur when a frequency of contractions reaches a threshold frequency. The threshold frequency may be every 5 minutes, every 10 minutes, every 15 minutes or every 20 minutes. Labour onset may occur when a duration of contractions reaches a threshold duration. The threshold duration may be 30 seconds, 60 seconds or 90 seconds. Labour onset may occur when the contractions are regular in one or both of frequency or duration. Labour onset may occur when the waters of the labouring woman break. Labour onset may occur at any point based on the clinical opinion of an attending clinician.

[0039] The electronic device 100 comprises a second receiving means 104 for receiving risk factor data. The risk factor data includes one or more of a maternal characteristic of a mother of the fetus, a characteristic of the fetus, a characteristic of the labour, and a characteristic of the pregnancy.

[0040] The maternal characteristic may comprise a parity, age, body temperature, underlying health condition, ethnicity, body mass index of the mother or other physical characteristics. The maternal characteristic may include results of medical tests or investigations performed in relation to the mother. For example, the maternal characteristic may comprise results of ultrasound scans performed during pregnancy, a placental growth factor, or infection biomarkers relating to the mother.

[0041] The fetal characteristic may comprise a gestational age or underlying health condition of the fetus. The fetal characteristic may comprise results of medical tests or investigations performed in relation to the fetus. For example, the fetal characteristic may comprise results of ultrasound or other scans performed during pregnancy or infection or other biomarkers relating to the fetus.

[0042] The second receiving means 104 may be a means for enabling a clinician to manually input the risk factor data. For example, the second receiving means 104 may be a touch screen display, a keyboard, or any other input device. The risk factor data may be input by an attending clinician. Alternatively, or additionally, the second receiving means 104 may be a means for obtaining risk factor data from a stored database. At least some of the risk factor data may be stored in a database accessible by the second receiving means 104.

[0043] The electronic device 100 further comprises a processor 106 and a memory 108. The memory 108 may store the received risk factor data and fetal heart rate data. The processor 106, which may also be referred to as a controller, determines the estimated likelihood of the fetus presenting with severe compromise evident at birth. This estimated likelihood may be referred to as the neonatal severe compromise risk score. Throughout this application, the terms “neonatal severe compromise risk score” and “risk score” may be used interchangeably. In addition, the terms “fetus” and “baby” may be used interchangeably.

[0044] The processor 106 may wherein extract at least one fetal heart rate characteristic from the received fetal heart rate data. The at least one fetal heart rate characteristic may comprise one or more of a baseline fetal heart rate, a short term variability of the fetal heart rate, acceleration of the fetal heart rate, a deceleration capacity of the fetal heart rate, a number of prolonged decelerations of the fetal heart rate within a second predetermined time period, a number of fetal heart rate accelerations and a size of the fetal heart rate accelerations.

[0045] To speed up the processing of the received fetal heart rate data, the processor may process only a portion of the received fetal heart rate data. For example, the processor may analyse only a first 60 minutes' worth of obtained fetal heart rate data, where that 60 minutes occurs within the predetermined time period around labour onset. As another example, the processor may only analyse a first 20 minutes' or 15 minutes' worth of fetal heart rate data.

[0046] The processor 106 uses a data-driven computational model to determine the risk score. The data-driven model may be a statistical prognosing model , support vector machine, neural network, or any other suitable model for determining a risk score using data relating to previous births and their outcomes. . In this application, we will discuss the data-driven model in terms of a statistical prognosing model, but the skilled person would readily understand that the invention is not limited to this embodiment and that the statistical prognostic model is just one type of suitable model. Any computational model which is trained, using prior data, to predict a likely labour outcome (or generate a risk score of a particular labour outcome) when provided with new input data may be used to implement this invention.

[0047] The model of the present application is based on a statistical prediction model derived from real data relating to over 50000 prior births. That is, the statistical prognosing model has been fitted to the data of over 50000 previous births. The data included the parameters indicated in the equation below, including the fetal heart rate parameters and the risk factors, as well as whether the outcome of the birth involved the fetus presenting with severe compromise at birth. At least some of the fetal heart rate data was obtained using computerised analysis of the stored CTG traces relating to the fetus.

[0048] The statistical prognosing model may comprise a formula for determining the risk score. In particular, the processor 106 may determine the estimated neonatal severe compromise risk score according to the formula score. = __-___wherein 1 + x = citc2*Mt H-c3*Np+ c4*Ma / c5 + c^PED + c^Npd + c8*((A« + c9) / ci0) 4- c^Dc + ci2*An + cu*((STV / c^ T (ci5*((G / c16)3)) +(ci7*((G / ci8)3 )*log(G) / ci9)) + c^L, + c^{B / c22) 4- ¢23¾ where Mt = maternal temperature, Np = nulliparity, Ma = maternal age, Ped = presence of pre-eclampsia or diabetes, Npd = number of prolonged decelerations of fetal heart rate, Aa = acceleration amplitude of fetal heart rate, Dc = deceleration capacity of fetal heart rate, An = number of accelerations of fetal heart rate, STV = short term variability of fetal heart rate, G = gestational age of fetus, Ls = labour stage, B = fetal baseline heart rate, Tm = presence of thick meconium and ci to C23 are constants.

[0049] It is not necessary for all of the parameters to be known, or input, in order for the risk score to be determined. However, the more parameters that are input, the more accurate the risk score.

[0050] Moreover, the skilled person would recognise that additional parameters including, but not limited to those discussed in this application, may be further incorporated into this model to further improve the accuracy of the model.

[0051] The maternal temperature is input in degrees Celsius in the range 35 - 41. Nulliparity refers to whether the mother has previously given birth, and is input as a binary 1 or 0. The input is 1 if the mother has not previously given birth, and 0 if the mother has previously given birth. The maternal age is a number, in years, between 16 and 55. The presence of pre-eclampsia or diabetes is provided as a binary input, where the input is 1 if one or both of pre-eclampsia or diabetes is present in the mother and 0 if the mother does not suffer from either pre-eclampsia or diabetes. The number of prolonged decelerations of the fetal heart rate is an integer value of at least 0. The acceleration amplitude is measured in beats per minute and is input as a number between 5 and 45. The deceleration capacity of the fetal heart rate is measured in beats per minute and input as a number between 0 and 20. The number of accelerations of the fetal heart rate is input as an integer number of at least 0. The short term variability of the fetal heart rate is measured in milliseconds, and input as a number between 0 and 40. The fetal heart rate variability could also or instead be measured in beats per minute and input into the model accordingly. The gestational age of the fetus is measured in weeks, and is input as a number between 36 and 45. The labour stage is a binary input, where 0 is input if the stage of labour at the time that the fetal heart rate data is obtained is pre-labour, and 1 is input if the stage of labour at the time that the fetal heart rate data is active labour. The fetal baseline heart rate is input as a number between 50 and 230. The presence of thick meconium is a binary input with a value of 0 if thick meconium is not present and a value of 1 if thick meconium is present. The presence or absence of thick meconium is judged by a clinician attending the birth.

[0052] Some or all of the risk factor inputs relating to the fetal heart rate may be obtained directly from the fetal heart rate monitor. Some or all of the risk factor inputs relating to the fetal heart rate may be determined by the processor 106 based on the received fetal heart rate data.

[0053] One example for the values of the constants is: ci = -18.61, C2 = 0.96, C3 = 1.25, C4 = 0.53, C5 = 10, C6 = 0.55422, c? = 0.19; cs = -0.02, C9 = 0.1, cio = 10, cn = -0.14, C12 = 0.09, C13 = 0.088, C14 = 10, C15 = 686.74, Ci6 = 100, C17 = 1128.04, cw = 100, C19 = 100, C20 = 0.03, C21 = -0.44, C22 = 100 and C23 = 1.17. These particular coefficients have been derived from data relating to the over 50000 previous births. However, the skilled person would recognise that strict adherence to these precise figures is not necessary to implement the invention. Instead, these figures merely provide an indication of the order of magnitude of the constants for obtaining good results with the invention.

[0054] The determined risk score is a value between 0 and 1, where higher values signify increased risk of the fetus presenting with severe compromise at birth.

[0055] Although doctors have been using fetal heart rate monitors for decades, the relationship between fetal heart rate patterns I risk characteristics and clinical outcomes is still largely unknown. The ability to estimate a likelihood of severe compromise in a fetus, using data obtained at an onset of labour and including a variety of risk parameters and characteristics of the fetal heart rate, is a significant leap forward in fetal healthcare.

[0056] For example, current fetal monitoring performed by clinicians often focuses on fetal heart rate decelerations and fetal heart rate accelerations are seldom considered important. The present invention recognises the impact that the number and size of fetal heart rate accelerations can have on the outcome of the baby.

[0057] It has also been identified that high values of the decelerative capacity (DC-PRSA) parameter are related to a type of reduced oxygen. The present invention can consider the sleep state cycles of the fetus using the DC-PRSA parameter, and determines whether the DC-PRSA value is persistently low (as opposed to current practice, where the focus is on high DC-PRSA values).

[0058] In addition, the present invention focuses on severe compromise as the labour outcome. The outcome of labour can be measured or considered in many different ways, and considering the outcome as “severe compromise” at the onset of labour is unique to the present invention. A more typical labour outcome to try to predict is cord acidemia. However, cord acidemia is not a correct biomarker for obtaining an estimate of severe compromise as the labour outcome. In addition, severe compromise may broadly capture many different outcomes, as previously described, such as sepsis, infection, oxygen deprivation, fetal brain injury, and placental issues.

[0059] In addition, current monitoring methods aim to identify early signs of oxygen deprivation in a fetus, in order to trigger a decision to expedite the delivery of the fetus. This inherently requires the fetus to start showing signs of oxygen deprivation or distress before delivery is expedited. The fetus therefore suffers at least some oxygen deprivation or distress between the initial identification of the early signs of distress and the delivery of the baby, which may lead to negative health outcomes for the baby. The present invention enables the baby to be delivered before there are visual signs of distress, resulting in improved health outcomes for the baby.

[0060] The present invention considers, at the onset of labour, whether the baby is fit for the entire labour process. The parameters and physiology considered using the fetal heart rate data, and the risk factor data considered, is very different to that considered using standard fetal monitoring techniques, where the focus is on a current health condition of the baby.

[0061] FIG. 2 shows an example of a system 200 for obtaining an estimated likelihood of a fetus presenting with severe compromise at birth.

[0062] The system 200 comprises a monitoring apparatus 202 and the electronic device 100 of FIG. 1. The monitoring apparatus 202 monitors the fetal heart rate and outputs fetal heart rate data. The monitoring apparatus 202 may be a cardiotocogram, or any other device for monitoring a fetal heart rate.

[0063] The monitoring apparatus 202 may transmit the obtained fetal heart rate data to the electronic device 100 and, in particular, to the first receiving means 102 of the electronic device 100. The monitoring apparatus 202 may have a wireless or a wired connection to the electronic device 100.

[0064] The electronic device 100 may form a wireless connection with a plurality of monitoring apparatuses 202 and receive fetal heart rate data from each of the plurality of monitoring apparatuses. Each of the plurality of monitoring apparatuses 202 may monitor the heart rate of a different fetus, where each fetus belongs to a different mother. The electronic device 100 may also receive, at the second receiving means 104, corresponding risk factor data for each fetus of the plurality of fetuses. In this manner, the electronic device 100 may estimate a likelihood of severe compromise in a fetus for the plurality of fetuses at the same time. Furthermore, a single electronic device 100 may be used simultaneously by a plurality of monitoring apparatuses 202, thereby minimising overall costs of the system 200.

[0065] FIG. 3 shows an example of a method 300 for estimating a likelihood of a fetus presenting with severe compromise at birth. The method 300 may be performed by the electronic device 100 and, in particular, by the processor 106 of the electronic device 100.

[0066] In step 302, fetal heart rate data is received. The fetal heart rate data is obtained from a fetal monitor, such as the monitoring apparatus 202. At least a portion of the fetal heart rate data is obtained by the fetal monitor within a predetermined time window of labour onset.

[0067] In step 304, risk factor data corresponding to the fetus is received. The risk factor data relates to a characteristic of one or more of the mother, the fetus, the pregnancy or the labour.

[0068] The skilled person would readily recognise that step 302 and step 304 may occur in any order, or simultaneously. The risk factor data and the fetal heart rate data need to be received, but the order in which they are received is unimportant.

[0069] In step 306, a neonatal severe compromise risk score is determined based on the received fetal heart rate data and the risk factor data. The neonatal severe compromise risk score is determined using a pre-trained data-driven computational model, such as a statistical prognosing model. A statistical prognosing model may use the equation discussed in relation to FIG. 1.

[0070] In step 308, the determined neonatal severe compromise risk score is output. The output may be provided as a graphical user interface (GUI) for ease of use by the clinician attending the labour. The GUI may additionally output one or more of the main contributing clinical indicators to the risk score. This may enable a clinician to verify a likely accuracy of the determined risk score, or take steps to mitigate the risk without necessarily speeding up the labour process. Where the risk score indicates an above average likelihood of the fetus presenting with severe compromise at birth, displaying the main contributing clinical indicators may assist the clinician with determining an appropriate course of action for intervening with the labour. The determined risk score may alternatively or additionally be output audibly so that the clinician does not need to interrupt their care of the labouring mother to receive the determined risk score.

[0071] The determined risk score may be output as an assigned risk category, the risk category indicating whether the fetus has a low, average or high risk of being severely compromised at birth. The high-risk category may be further banded according to a severity of the risk to provide clinicians with additional information so that they may choose an appropriate intervention method.

[0072] That is, the risk score may be output in relation to a prior background risk derived from the dataset used to obtain the data-driven computational model. The risk score therefore not only indicates whether the risk of severe compromise in the fetus is above average, but includes a measured factor indicative of how much the risk is increased above an average baseline risk. That is, the risk score may be provided in the context of a background risk level to aid clinical decision making.

[0073] Optionally, in step 310, an alert is output if the determined neonatal severe compromise risk score is above a predetermined threshold risk score. The predetermined threshold risk score may be set at a value indicating that immediate medical intervention is required to prevent or reduce the likelihood of the fetus suffering severe compromise evident at birth. The predetermined threshold risk score may be set at a level indicating that the fetus has at least 1.5 times, at least 2 times, or at least 3 times a risk of suffering severe compromise compared to a background risk level.

[0074] The skilled person would readily recognise that step 308 and step 310 may occur in any order, or simultaneously, without impacting the invention.

[0075] The term 'processor' is to be interpreted broadly to include a CPU, processing unit, ASIC, logic unit, or programmable gate array etc, and may refer to a single processor or a combination of several processors. Certain aspects of the disclosure may be implemented using machine-readable instructions which may, for example, be executed by a general purpose computer, a special purpose computer, an embedded processor or processors of other programmable data processing devices to realize the functions described in the description and diagrams. In particular, a processor or processing apparatus may execute the machine-readable instructions. Thus, functional modules of the apparatus and devices may be implemented by a processor executing machine readable instructions stored in a memory, or a processor operating in accordance with instructions embedded in logic circuitry. The functional modules may be implemented in a single processor or divided amongst several processors.

[0076] It will be appreciated that embodiments of the present invention can be realised in the form of hardware, software or a combination of hardware and software. Any such software may be stored in the form of volatile or non-volatile storage such as, for example, a storage device like a ROM, whether erasable or rewritable or not, or in the form of memory such as, for example, RAM, memory chips, device or integrated circuits or on an optically or magnetically readable medium such as, for example, a CD, DVD, magnetic disk or magnetic tape. It will be appreciated that the storage devices and storage media are embodiments of machine-readable storage that are suitable for storing a program or programs that, when executed, implement embodiments of the present invention. Accordingly, embodiments provide a program comprising code for implementing a system or method as claimed in any preceding claim and a machine readable storage storing such a program. Still further, embodiments of the present invention may be conveyed electronically via any medium such as a communication signal carried over a wired or wireless connection and embodiments suitably encompass the same.

[0077] Throughout the description and claims of this specification, the words “comprise” and “contain” and variations of them mean “including but not limited to”, and they are not intended to (and do not) exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

[0078] Features, integers, characteristics, compounds, chemical moieties or groups described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and / or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and / or steps are mutually exclusive. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.

[0079] Unless stated otherwise, the numbers and values provided in the examples discussed throughout the description are included to assist with the understanding of the specification. The numbers used, and any relationships between the numbers, are merely examples and are not intended to be limiting.

[0080] It will be appreciated that all of the above-described embodiments, and their technical features, may be combined with one another in each and every combination, potentially unless there is a conflict between two embodiments or features. That is, each and every combination of two or more of the above-described embodiments is envisaged and included within the present disclosure. One or more features from any embodiment may be incorporated in any other embodiment, and provide a corresponding advantage or advantages.

Claims

1. A computer-implemented method for estimating a likelihood of neonatal severe compromise evident at birth, the method comprising:receiving fetal heart rate data of a fetus, wherein the fetal heart rate data is obtained from a fetal monitor and at least a portion of the fetal heart rate data is obtained by the fetal monitor within a predetermined time window of labour onset;receiving risk factor data corresponding to the fetus, the risk factor data comprising one or more of a maternal characteristic of a mother of the fetus, a characteristic of the fetus, a characteristic of the labour, and a characteristic of the pregnancy,determining, using a data-driven computational model, a neonatal severe compromise risk score based on the received fetal heart rate data and the risk factor data; andoutputting the determined neonatal severe compromise risk score, where the determined neonatal severe compromise risk score is indicative of a likelihood of the fetus presenting with severe compromise at birth.

2. The computer-implemented method of claim 1, wherein the maternal characteristic comprises one or more of a maternal parity, maternal age, maternal body temperature, a maternal underlying health condition, a maternal ethnicity, a maternal body mass index or a maternal medical test result.

3. The computer-implemented method of claim 1 or 2, wherein the characteristic of the fetus comprises one or more of a fetal gestational age, a fetal underlying health condition, a fetal medical test result, or a sleep state cycle of the fetus.

4. The computer-implemented method of any one of claims 1 to 3, wherein the characteristic of the labour comprises one or more of a type of onset of the labour and a thickness of meconium.

5. The computer-implemented method of any one of claims 1 to 4, wherein determining the fetal severe compromise risk score comprises extracting at least one fetal heart rate characteristic from the fetal heart rate data,wherein the at least one fetal heart rate characteristic comprises one or more of a baseline fetal heart rate, a short term variability of the fetal heart rate, acceleration of the fetal heart rate, a deceleration capacity of the fetal heart rate, a number of prolonged decelerations of the fetal heart rate within a second predetermined time period, a number of fetal heart rate accelerations and a size of the fetal heart rate accelerations.

6. The computer-implemented method of any one of claims 1 to 5, wherein the data-driven computational model is a model fitted to data relating to a plurality of prior births, optionally wherein the plurality of prior births comprises at least 1000 prior births, 10000 prior births, 25000 prior births or 50000 prior births.

7. The computer-implemented method of any one of claims 1 to 6, wherein the data-driven computational model is configured to process the received fetal heart rate data and the categorical risk factor data according to:Risk^score = -----1 + e*"whereinx - + +c^Pbd+c7*Npd + c^((Aa +c9) / c1Q)+ + cl2*An -|- c^^STV / c^-2) t- (ci5*((G / ei6)3))+(c17*((G / c18)3 )*log(G) / c19)) + e2{ALs + e21*(B / c-22) + c23*TMwhere Mt = maternal temperature, Np = nulliparity, Ma = maternal age, Ped = presence of pre-eclampsia or diabetes, Npd = number of prolonged decelerations of fetal heart rate, Aa = acceleration amplitude of fetal heart rate, Dc = deceleration capacity of fetal heart rate, An = number of accelerations of fetal heart rate, STV = short term variability of fetal heart rate, G = gestational age of fetus, Ls = labour stage, B = fetal baseline heartrate, Tm = presence of thick meconium and ci to C23 are constants.

8. The computer-implemented method of any one of claims 1 to 7, wherein the method further comprises outputting an alert if the determined risk score is greater than a predetermined threshold value.

9. The computer-implemented method of any one of claims 1 to 8, wherein the predetermined time period is a period of up to five hours before or after the onset of the labour.

10. The computer-implemented method of any one of claims 1 to 9, wherein the fetal heart rate data is recorded over a period of at least 5 minutes.

11. The computer-implemented method of any one of claims 1 to 10, further comprising: receiving uterine contraction data relating to a plurality of uterine contractions, wherein the uterine contraction data is taken within the predetermined time period of the onset of labour; whereindetermining the fetal severe compromise risk score further comprises processing the received uterine contraction data using the data-driven computational model.

12. An electronic device for estimating a likelihood of neonatal severe compromise evident at birth, the electronic device comprising:a first receiving means configured to receive fetal heart rate data of a fetus, wherein the fetal heart rate data is obtained from a fetal monitor and at least a portion of the fetal heart rate data is obtained by the fetal monitor within a predetermined time window of labour onset;a second receiving means configured to receive risk factor data corresponding to the fetus, the risk factor data comprising one or more of a maternal characteristic of a mother of the fetus, a characteristic of the fetus, a characteristic of the labour, and a characteristic of the pregnancy;a processor; anda memory storing instructions that, when executed by the processor, cause the processor to perform a method according to any one of claims 1 to 11.

13. The electronic device of claim 12, wherein the first receiving means comprises an antenna configured to receive, from an external apparatus, a signal comprising the fetal heart rate data.

14. A system for estimating a likelihood of neonatal severe compromise evident at birth, the system comprising:an electronic device according to claim 12 or claim 13; and a fetal monitor for obtaining fetal heart rate data.

Images