WO2010038217A1

WO2010038217A1 - Neonatal brain well-being monitor

Info

Publication number: WO2010038217A1
Application number: PCT/IB2009/054316
Authority: WO
Inventors: Eli Morgan Barak; Johannes Bergman Nils
Original assignee: University Of Cape Town; South African Medical Research Council
Priority date: 2008-10-03
Filing date: 2009-10-02
Publication date: 2010-04-08

Abstract

A method and apparatus for monitoring the emotional state of an infant are disclosed. The apparatus includes an input circuit for receiving and pre-processing EEG and ECG data acquired from a subject, A processor running proprietary software, implementing respective modules of the apparatus, is arranged to analyse the EEG data and ECG data and generate respective output data related thereto. The respective modules determine asymmetry of the frontal lobe activity, and analyse the ECG data to determine several parameters including heart rate and at least one further parameter related to the interval between R-waves in an ECG waveform of the subject. The first and second output data are processed to generate display data, and the apparatus includes at least one display responsive to the display data to provide an indication in real time or near- real time of the emotional state of the subject.

Description

NEONATAL BRAIN WELL-BEING MONITOR

BACKGROUND OF THE INVENTION

THIS invention relates to a method of, and apparatus for, monitoring emotional or brain well-being in infants.

Current practice in many hospitals and neonatal facilities is to separate newborn infants from their mothers, as opposed to maintaining them in skin-to-skin contact with their mothers for extended periods. This applies to premature infants, who are kept in incubators, and also to many infants born by caesarean section and also natural birth, who are generally kept in cots near their mothers.

There is substantial evidence that human infants deprived of maternal contact are prone to develop greater negative affect (emotion) and poorer psychological and physiological resilience in both short and long term timeframes.

The negative biological effects of mother-infant separation in animals are well established. Broadly, animal studies support the thesis that the neuroendocrine corticotrophin release factor (CRF) system and hypothalamus-pituitary-adrenal (HPA) axis are sensitive to the early rearing environment (Meaney and Szyf 2005, Szyf et al. 2005, Thibodeau et al. 2006, Zhang et al. 2006,). As one example, Plotsky et al. (2005) reported clear differences at the molecular level in distributed CRF neural circuitry in the brains of rat pups that were separated from their mothers. They concluded: "Early experience-induced changes in CRF systems may lead to a cascade of adaptations including both the noradrenergic and serotonergic systems. This process serve[s] to 'tune' stress- sensitive neurocircuits in an adaptive or maladaptive manner. These findings may provide a mechanism whereby [early] life events, interacting with genetic predisposition, may increase vulnerability to pathology in later life." (from Plotsky et al. 2005, quote shortened)

In all animal studies mother-infant contact is considered the norm and separation is the experimental intervention. In contrast, human neonates are routinely and normatively separated from their mothers and placed into cots or incubators, ever the more so when requiring medical attention. However, recent evidence from studies in non-human primates and several other mammals including rats, mice and pigs, strongly suggests that separation may produce harmful effects. Furthermore the benefit in humans of avoiding separation from the mother has been substantiated by clinical studies. In a randomised controlled trial Bergman et al (2004) found significantly better cardiorespiratory function in low birth weight (120Og to 220Og) preterm neonates nursed against their mother's skin compared to infants nursed in an incubator. In this study, at 6 hours of age, 18/18 of SSC infants stabilised, while only 6/13 infants in incubators were completely stable. Researchers of non-human primates (regarding non- separation as normative) would conclude: "incubators destabilize newborns". Whether incubators are good or bad for newborn babies is therefore not a trivial question.

In short, recent neuroscience indicates that mammalian brain development is highly experience dependent and requires largely uninterrupted mother- infant physical contact for normal development. Recent reviews of the outcomes of human infants cared for in modern neonatal intensive care units provides support for these basic neuroscience findings. An alternative to separation of mothers and infants was first described in Colombia, as the "kangaroo mother-care method". This method was adapted by Bergman and Jurisoo working in Zimbabwe, using a technique that made the maternal-infant skin-to-skin contact (SSC) function as an ongoing incubator for premature infants. It was observed that premature infants in this method of care showed precocious development, and a five-fold improvement in survival. The above mentioned randomised controlled trial (Bergman, et al, 2004), conducted in Cape Town, showed that low birth weight babies who were not separated from their mothers stabilised in 6 hours, but became more unstable in incubators. These observations are entirely in keeping with mammalian separation studies.

It is now accepted that infant survival should not be the measure of good care, but rather the quality of survival as expressed in neurodevelopmental outcome. There is however no monitor that measures the neuroaffective (emotional) well-being of the brain during this development. In the current context the terms "neuroaffective well-being" or "emotional well-being" have a specific biological meaning. Emotions are the subjective aspects of global behavioural and internal physiological responses orchestrated by the brain that are crucial for survival of the organism. This is evident in the word itself, e-motion/emote which indicates motor action or physiological- behavioural response. In biology therefore, emotion means the adaptive response or "physiological-behavioural attitude" of the organism as a whole to particular circumstances. At bottom these circumstances are of two types: potentially rewarding or potentially threatening and elicit two basic emotional/affective responses: approach or avoidance/withdrawal.

These emotional/neuroaffective responses or "states" have a profound impact on neonate/infant development. The mammalian brain interprets mother-infant skin-to-skin contact as a safe, rewarding environment and orchestrates an approach-type neuroaffective state comprising nutritive (feeding and growth) and social (bonding and communication) components accompanied by pleasurable feelings. In contrast, the mammalian brain interprets mother-infant separation as an unsafe, threatening environment and orchestrates an avoidance-type neuroaffective state comprising withdrawal (immobilisation and resource-conservation) and social shutdown components accompanied by unpleasant feelings.

These states are subserved by different neural pathways in the brain and body and nerves which "fire together, wire together", i.e. the brain is continuously shaped by experience. Early neuroaffective/emotional well- being, therefore, has profound global implications for the well-being of body, brain and mind. These considerations however do not enter into standard current neonatal practise where babies are routinely separated from their mothers in 'incubators. Indeed according to modern western clinical practise it is considered almost "unethical" not to separate preterm/ill babies in this way.

It is an object of the invention to provide a method of monitoring emotional well-being in infants, and apparatus for implementing the method.

SUMMARY OF THE INVENTION

According to the invention there is provided a method of monitoring the emotional state of an infant, the method including:

acquiring EEG data from an infant subject indicative of frontal lobe activity in the subject;

acquiring ECG data from said subject indicative of heart rate and heart rate variability (HRV) in the subject;

analysing the EEG data to determine asymmetry of the frontal lobe activity and generating first output data related thereto;

analysing the ECG data to determine a plurality of parameters including heart rate and at least one further parameter related to the interval between R-waves in an ECG waveform of the subject, and generating second output data related thereto; and processing the first and second output data to provide an indication in real time or near-real time of the emotional state of the subject.

The at least one further parameter determined from the ECG data may include one or more of the following:

(a) Old vagus (very low frequency component of HRV).

(b) Sympathetic nervous system or SNS (low frequency component of HRV).

(c) New vagus (high frequency component of HRV).

(d) Standard deviation of normal R-R intervals (SDNN) as a moving average.

The method may include measuring and analysing at least one of the following parameters to determine the emotional state of the subject:

(e) Infant behavioural/vigilance state using the ABSS (eg. Asleep, quiet awake, crying);

(f) Instantaneous heart rate (=1/RR interval);

(g) Standard deviation of normal R-R intervals (SDNN) as a moving average.

The indication of the emotional state of the subject may comprise a graphical representation suggesting at least one of at least two possible emotional states.

The indication of the emotional state of the subject may comprise quantitative values obtained from recorded data, time lines of such data over a selected measurement period, or text-format interpretations of the subject's current state.

Further according to the invention there is provided apparatus for monitoring the emotional state of an infant, the apparatus including: an input circuit for receiving and pre-processing EEG and ECG data acquired from a subject;

a processor arranged to analyse the EEG data to determine asymmetry of the frontal lobe activity and to generate first output data related thereto; to analyse the ECG data to determine a plurality of parameters including heart rate and at least one further parameter related to the interval between R-waves in an ECG waveform of the subject, and to generate second output data related thereto; and to process the first and second output data to generate display data; and

at least one display responsive to the display data to provide an indication in real time or near-real time of the emotional state of the subject.

The apparatus preferably includes at least one module for analysing the EEG and ECG data, wherein a module for analysing the ECG data is arranged to determine at least one of the following further parameters from the ECG data:

(a) Old vagus (very low frequency component of HRV);

(b) Sympathetic nervous system or SNS (low frequency component of HRV);

(c) New vagus (high frequency component of HRV);

(d) Standard deviation of normal R-R intervals (SDNN) as a moving average.

Preferably, at least one module of the apparatus is arranged to measure and analyse one or more of the following parameters to determine the emotional state of the subject:

(e) Infant behavioural/vigilance state using the ABSS; (f) Instantaneous heart rate (=1/RR interval);

(g) Standard deviation of normal R-R intervals (SDNN) as a moving average.

Said at least one display is preferably capable of displaying a graphical representation suggesting at least one of at least two possible emotional states.

Alternatively, or in addition, said at least one display is preferably capable of displaying an indication of the emotional state of the subject comprising quantitative values obtained from recorded data, time lines of such data over a selected measurement period, or text-format interpretations of the subject's current state.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a simplified schematic block diagram of apparatus for monitoring the emotional state of an infant according to the invention;

Figures 2 and 3 are sets of graphs showing displays of ECG parameters generated by the apparatus of Figure 1 ;

Figure 4 is a simplified flow chart showing the functions implemented by modules of the apparatus; and

Figure 5 is a schematic diagram illustrating operational principles of the method and apparatus of the invention. DESCRIPTION OF AN EMBODIMENT

The present invention aims to provide a means of looking at the emotional well-being of an infant subject, that is, the well-being of the subject's brain, utilising both EEG and ECG data. The ECG data is complemented and corroborated by looking at upstream factors (EEG) in the frontal lobe (the heart being a downstream effect), the highest part of the "emotional brain" (the neural structures and networks responsible for processing emotional stimuli and generating emotional responses). The method of the invention gathers signals from these two sources, interprets them according to known as well as novel principles of "affective neuroscience" (the neurobiology of emotion), and provides real time or near-real time information to clinicians in order to guide newborn / infant care.

The first "emotional brain" to evolve was the autonomic nervous system which ensures appropriate behavioural responses and internal (metabolic, physiological) homeostasis necessary for survival. Survival, however, is more than momentary homeostasis, but involves procurement of resources for growth and development and avoidance of danger. The first component of the autonomic nervous system (ANS) to evolve (called the "old vagus") was therefore capable of orchestrating two emotional response modes: "nutrition & growth" in response to safe conditions and "immobilisation" in response to threat. This deepest layer is exemplified in reptiles and amphibians.

During subsequent evolution, the emotional brain became increasingly complex in a step-wise fashion discernible in the deeper layers of the mammalian brain. A later system evolved in early mammals, and is the well-known "fight or flight" of the sympathetic nervous system (SNS) which allows for "mobilisation" in response to threat (or when essential needs are not being met, eg. acute hunger). Together the old vagus and SNS constitute the autonomic nervous system (ANS). The latest layer is a social system (called the "new vagus"), which allows for the reading of the social environment ( (especially cues of emotional facial expressions: friendly or threatening). The ANS and new vagus together constitute the "emotional brain".

The emotional brain therefore comprises more primitive autonomic and more recent social-emotional components. These three systems generate distinctly different behaviours and make distinctly different metabolic demands on the body as a whole. The vegetative system switches between immobilisation (a passive withdrawal response to threat) and nutritive (feeding/growth response in absence of threat) modes. The SNS triggers an active fight or flight response to threat while the social system mediates internal and behavioural responses to changes in the social environment. The three systems communicate in bottom-up and top-down ways and mediate increasingly sophisticated and subtle physiological, behavioural and subjective (motivational/emotional) approach/avoid response patterns to rewarding or threatening situations.

The heart plays a crucial role in adapting the body according to the metabolic demands of each neuroaffective (emotional brain) subsystem. The emotional brain therefore sends appropriate messages to the heart through nerve pathways. Signals from the new social vagus have been extensively studied. At rest (safe conditions) it is actively on and produces a slowing effect on the heart (if it was not there, human hearts would beat a lot faster than they do.) This effect is extremely finely tuned, and can change the interval from one beat to the next with microsecond precision. Metabolic demand changes can therefore be managed extremely efficiently.

This control mechanism is also directly linked to the breathing centre of the brain as evidenced by heart rate increases during inspiration and decreases during expiration (a pattern known as respiratory sinus arrhythmia, RSA). RSA and heart rate are so closely coupled that the breathing rate can be calculated accurately from a particular frequency band component in the heart's beat to beat variation. Variation of the interbeat interval on a beat-to-beat basis in the high frequency (HF) component of heart rate variability (HRV) reflects activity of the most recent layer of the emotional brain, the new social vagus.

In analysing frequencies in the above type of research, a method called a Fast Fourier Transformation is used. This extracts the HF frequency band, and calculates a power value (area under the graph, more area = more HF modulation = more new social vagus activity). Variation in the interbeat interval (IBI) also occurs over longer time periods. The HRV Task Force, based on adult research, has regarded these slower periods (= lower frequencies) as being a mixture of sympathetic and parasympathetic, and therefore not useful. We have however shown that in newborns, using a cycle of between 5 and 30 seconds, this power spectrum analyzed by FTT represents the SNS well. In contrast, periods less than 30 seconds have been called VLF (Very Low Frequency) and we have shown this to correspond with the old vegetative vagus.

The old vegetative vagus, the first layer of the emotional brain to evolve, having such old origins, has been regarded by many as having "vestigial" cardiac effects in the human being. In reptiles and amphibians, a stress response will make this vagus slow the heart rate very dramatically. As amphibians (and reptiles) require only 20% of the oxygenation of mammals, this allows for submerged immobilisation in water as a good defence. Such a slow heart rate would be lethal for a human, thus the system has been thought to be inactive.

The sympathetic nervous system represented a significant advance in terms of defence. It allows for very rapid mobilisation of resources, and a rapid increase in heart rate. But the social vagus has branches that switch this off (allowing for more subtle responses to a social threat, which does not necessarily mean a physical threat, where a fight or flight response is appropriate). The SNS can therefore only show its effect once the social vagus is inactive. In the presence of real danger, the social vagus does indeed switch off, allowing for an increase in heart rate (vagal brake is off), and on top of this increase, sympathetic activation increases heart rate as required. Whether in safe conditions or under threat, mammalian research shows that all the above systems interact in various ways. For example, the old vagus, while presumed to have no cardiac effect, will under threat actively shut off the gut.

The social vagus allows for the rapid emotional assessment of faces which requires rapid approach or avoid decisions, whether to engage or disengage. This involves both cortical or conscious judgement and subcortical unconscious responses. The subcortical amygdala is a key part of the emotional brain, and sends emotion-charged information to the frontal lobe. The left frontal lobe is responsible for approach to reward, and the right responsible for avoidance of threat. This activity can be measured in the EEG, and by comparing left and right a measure called "Frontal EEG Asymmetry" can be calculated (Davidson, 1998). Approach asymmetry has been shown to correlate well with psychological and physiological health in adults, and also in infants more than four weeks old. It has mostly been interpreted as a personality trait, rather than as an emotional state.

The incubator was invented some 100 years ago, and active management of prematurity started some 50 years ago. In the incubator environment, technologies were developed that supported temperature regulation, breathing and cardiac functions, and metabolic and nutritional support. A variety of monitoring techniques developed with these technologies. Increasingly sophisticated technology is now able to achieve survival at extreme low birth weights and gestational ages.

An unquestioned assumption in this development was the need for maternal infant separation, which became widely practised even for normal newborns. As previously discussed, mammalian neuroscience conducted during this same time period shows however that such separation is extremely damaging for the newborn brain.

In brief, the present invention utilises both frontal EEG data which is analysed to provide an indication of dynamic frontal lobe asymmetry, and ECG data from which heart rate variability (HRV) data is extracted and analysed, to determine an overall emotional state of the subject.

In preliminary work, a Place Model was developed. This states that the newborn in skin-to-skin contact with the mother (SSC; the expected place) expresses a nutritional and neurodevelopmental program that leads to good quality attachment, breastfeeding and optimal development. When separated from the mother (Maternal-Infant Separation or MIS; any other place) a survival program is expressed, in which psycho-physiological dysregulation is accompanied by poor quality attachment (e.g. dissociation), which hampers neurocognitive, emotional and social development.

A pilot study was devised, that hypothesised that the effects predicted in the Place Model could be detected, and ultimately monitored. Such monitoring would allow for environmental adjustments that improve emotional brain/neuroaffective status, the primary means being return of mother as provided for by SSC. However, in the unavoidable absence of mother, other environmental adjustments may or may not ameliorate separation effects, and working with these could improve neurodevelopmental outcomes.

In the pilot study, 12 infants were studied. They were two days old, full term and healthy, and awaiting the discharge of their mothers after caesarean section. Each spent one hour in a cot (MIS), and one hour on mother's chest (SSC), while HRV data and frontal EEG data were collected for subsequent analysis.

The HRV data has in the past been regarded as a static model, though recent research has shown there are dynamic aspects. The novel approach to making this data useful entails a synthesis of three things: firstly, emotional brain effects on the heart are dynamic. Secondly, they reflect the real time inputs and outputs of the emotional brain. Thirdly, the deeper and older vegetative vagus was identified at very low frequencies (VLF) buried in the HRV recording.

The Heart Rate Variability (HRV) frequencies which best reflect the different components of the emotional brain in neonates were predicted and explored in the data, after the data sets had been cleaned using a manual visual filtering technique. Following spectrographic analysis and adjustment of epoch lengths, the data was extracted and digitalised for further analysis. Correlations were made with the Anderson Behaviour State Scale or ABSS (Anderson 1986), heart rate and standard deviation of inter beat intervals of infants studied in cots and during maternal infant skin- to-skin contact.

Likewise, the EEG data were analysed. The data sets were cleaned, and epochs were extracted that allowed for dynamic state analysis. The aim now was to examine the asymmetry of the left and right frontal lobes as a dynamic phenomenon. In adults, the asymmetry has mostly been investigated as a trait, and is calculated from recordings averaged over 4 or more minutes. The hypothesis made instead was that in the newborn, such asymmetry is the result of the current experience (e.g. of skin-to-skin contact or MIS), which is known to stimulate the firing and wiring of the specific pathway that interconnect the emotional brain (e.g. amygdala and the frontal lobes). Frontal lobe activity asymmetry was investigated at different frequencies from those used in adults (i.e. not only alpha).

The behavioural results are in keeping with the Place Model. There is much more sleep and better quality of sleep during SSC, much more crying in MIS, and overall greatly increase activity in all components of the ANS during MIS.

During cot sleep, the babies appeared to be sleeping well, though regular breathing, evidence of the deepest phase of sleep, was reduced. This deepest sleep is essential for brain wiring, as consolidation of memory circuits occurs during such sleep. However, this research demonstrates that these babies are in fact not simply "asleep" when separated. They are in a state of heightened vigilance, with elevated stress effects from all three parts of the ANS. There is a heightened level of old vagus, which is causing an immobilisation defence, which is not actual sleep, only sleep mimicking. Babies sleeping with mothers (SSC) show a distinctive cyclicity of all three ANS systems, with the deep sleep phase is evident also on the heart rate and breathing.

Further, a number of crying episodes were analysed. These show distinctive ANS patterns, allowing interpretation of at least three cries, namely hunger, protest, social distress (each the vocalisation of a different system in the emotional brain). It is possible that combinations of these may signal other needs. It is obvious that the ability to interpret the newborns signals is essential to providing for its needs and wellbeing. These signals may be evident even without active crying - infants in intensive care technology may not be able to cry, yet will show such neuroaffective signals detectable from brain and cardiac monitoring.

Breastfeeding behaviours similarly appear identical to the external observer. This technology does however distinguish between at least two different feeding behaviours, "nutritive" (high old vagus) and "comforting/soothing", (high new social vagus). In a single episode a high sympathetic pattern may also be indicating breastfeeding as pain relieving behaviour.

Following the pilot study, a methodology and apparatus for implementing the findings were developed. The apparatus was designed to utilise output signals from currently available EEG and ECG equipment, and to process these signals using algorithms based on the findings of the pilot study, to provide a real time or near-real time indication of the emotional state of an infant subject. The prototype apparatus is shown in a simplified schematic form in Figure 1. The apparatus is utilised with a conventional EEG data acquisition unit 10, and a conventional ECG data acquisition unit 12.

The EEG data acquisition unit used in the pilot study was a battery powered research-only EEG unit such as units made by Biosemi of Amsterdam, The Netherlands or by G-Tec of Graz, Austria. The electrodes of the unit are easily adhered onto the mid-frontal areas of the scalp (F3 and F4) and left and right mastoids (as reference) using disposable adhesive rings or elasticised Velcro (trade mark) bands (e.g. Biosemi infant electrode caps). If the Biosemi-type active electrode system is used, a CMS (common mode signal) and a DRL (driven right leg) reference electrode will be attached to the scalp as well. The electrodes are washed in warm water and air dried between each subject.

The ECG data acquisition unit 12 can be the equivalent to an ambulatory ECG data acquisition system (e.g. a unit made by AMS of Vrije Universiteit, Netherlands) customised to work with three disposable neonatal ECG electrodes placed on the chest at the left mid-clavicular line (4^th intercostal space), at the sternal notch, and at the right mid-axillary line at the level of the last rib.

The outputs of the EEG and ECG data acquisition units 10 and 12 are fed to an A/D converter 14 for digitization. The output of the A/D converter is fed to a digital signal processing (DSP) circuit 16, which pre-processes the digitized data and utilises noise separation software to clean the acquired data.

The output of the DSP circuit is fed to a microprocessor 18 running proprietary software which implements the algorithms of the invention. The software provides an output which can drive a display 22 and an associated alarm or indicator 24. The apparatus of Figure 1 may include a plurality of components or modules which correspond to the functional tasks to be performed by the apparatus. In this regard, "module" in the context of the specification will be understood to include an identifiable portion of code, computational or executable instructions, data, or computational object to achieve a particular function, operation, processing, or procedure. It follows that a module need not be implemented in software; a module may be implemented in software, hardware, or a combination of software and hardware. Further, the modules need not necessarily be consolidated into one device but may be spread across a plurality of devices in the apparatus. In the prototype apparatus, the modules were implemented in proprietary software, as indicated above.

The software running on the microprocessor 18 calculates EEG asymmetry on-line using standard power - frequency spectrum density (PSD) extraction methods (FFT and/or wavelet analyses). The asymmetry score is then taken as the difference in the natural log of the 3-8Hz sub-band (additional sub-bands will possibly yield further useful info) power between left and right mid-frontal (F3, F4, 10-20 International System) electrode sites. (Additional sites, e.g. parietal areas, will possibly yield further useful information.) This gives values in the approximate range 0.5 (extreme right) to -0.5 (extreme left).

In(Rj - In(Ij = In (R/L)

(Reference: John J. B. Allen, James A. Coan, Maria Nazarian. Issues and assumptions on the road from raw signals to metrics of frontal EEG asymmetry in emotion. Biological Psychology 67 (2004) 183-218.)

The software modules implement QRS complex (R-wave) detection algorithms which output the RR interval (interval between successive heartbeats) in milliseconds (ms). This information is then upsampled and transformed into the time domain using MatLab code (e.g. Berger transform, BioSig open source). FFT or wavelet method algorithms extract the PSD which is then analysed to output normalized quantitative power- frequency values for three different components of the emotional brain (old and new vagus, SNS). The algorithms define and interpret these existing PSD sub-bands according to the principles described above. Once done (defined and quantified) the results are easily represented individually and in relation to one another on the monitor display and the data therein are automatically processed and simultaneously represented on the display in a graphic display to reflect emotional or brain well-being. This latter processing is now described in greater detail.

The essence of the invention is to gather diverse and previously unconnected information and interpret it in terms of emotional or brain well- being, meaning approach/avoidance modes as described above. Seven core parameters for this purpose are:

1. Infant behavioural/vigilance state using the ABSS (eg. Asleep, quiet awake, crying).

2. Old vagus (very low frequency component of HRV).

3. Sympathetic nervous system or SNS (low frequency component of HRV).

4. New vagus (high frequency component of HRV).

5. Frontal EEG asymmetry.

6. Instantaneous heart rate (=1/RR interval).

7. Standard deviation of normal R-R intervals (SDNN) as a moving average over 20 heartbeats.

This information is automatically represented and interpreted as depicted in Figures 2 and 3.

The major steps carried out by the modules of the apparatus are shown in flowchart form in Figure 4. Exampie 1: Three different cries

Figure 2 shows six of the seven core parameters measured (i.e. EEG is not shown in these figures) as they are represented on the display (time moving from left to right, total time represented = the last 120 mins):

The top graph shows behavioural state according to the Anderson Behavioural State Scale which has 12 states, the lowest being 1 (quiet sleep), the highest being 12 (severe distress).

The middle graph shows instantaneous heart rate (upper trace 26) and standard deviation of instantaneous heart rate (lower trace 28).

The lowest graph shows the old vagus 30, new vagus 32 and SNS 34 on a normalised % scale.

Also shown are Place (SSC or MIS).

The infant who was monitored to derive the data of Figure 2 cried (states 10 and 11) 5 times but these 5 cries fall into 3 types recognisable as follows (the appropriate interpretations are given in parentheses in each case):

Cries 1 and 4 are hunger cries: the old vagus (nutritive) 30 is higher than the SNS (fight/flight) 34 and the new vagus (social) 32 is low. The heart rate (blue) is high and the SDNN 28 is low.

Cry 2 is a social cry: the old vagus 30 is low, the SNS 34 is low and the new vagus 32 is high. The heart rate 26 is low and the SDNN 28 is high.

Cries 3 and 5 are anger/protest cries: the old vagus 30 is lower than the SNS (fight) 34 and the new vagus (social communication) 32 is high. The heart rate 26 is high and the HRV 28 is low. Example 2: Two different breast feeds

In the same baby there are two episodes of breastfeeding depicted as ovals in the lowermost graph of Figure 2.

The first is a hunger feed: the old vagus (nutritive) 30 is high and the SNS 34 and the new vagus (here mother is a feeding object rather than a social object) 32 are low.

The second is a comfort feed: the old vagus 30 is low, the SNS 34 is low and the new vagus 32 is high.

Example 3: A sleep cycle in SSC

Figure 3 shows a full normal sleep cycle in skin-to-skin contact (SSC). During the first 2/3rds of the cycle, the old vagus is low (sleep not immobilisation), the SNS is low (no threat, hunger) and the new vagus is high (safe with mother). Heart rate is medium and HRV is medium. In the last third of the sleeping state, old vagus becomes high (hunger) and new vagus becomes low (mother is now a feeding object and no longer a social object). The SNS also rises from low to high in this phase (in response to low blood sugar) but only exceeds the old vagus once the baby wakens.

All the information in these examples comes from a pilot study involving 12 healthy full term caesarean section mother-infant dyads. Full term infants (≥ 37 weeks gestation) born by caesarean section were chosen because they are routinely nursed in a cot next to their mother's bed (i.e. MIS) as well as spending SSC and breastfeeding time with their mothers. Each infant was studied in both SSC and MIS Places for one hour in each Place. Because of possible effects of maternal medications/anaesthetic agents on infant physiology, testing took place on postnatal day two or three and not on day one. EEG asymmetry

A rolling average of the frontal EEG asymmetry score (the 7^th core parameter) will be represented in graph form with the other parameters shown above. The prevailing hypothesis in the literature is that left-sided = happy, right-sided = unhappy. In general, based on the pilot data, right- sided asymmetry (more negative scores) indicating the baby is unhappy invariably occurs in MIS and left-sided scores (more positive scores) indicating the baby is happy invariably occurs in SSC, apart from a small number of anomalous cases.

Summary table of the emotional brain patterns (EBPs) described in Examples 1 to 3:

hunger cry social cry distress cry early late hunger comfort

EBP (Cries 1 & 4) (Cry 2) (Cries 3 & 5) sleep sleep feed feed

See Crying Figure See Sleeping Figure See Crying Figure

HR High Low high medium

SDNN Low High low medium

Old vagus Higher Low lower low high high low

SNS Lower Low higher low high low low

New vagus Low High high high low low high

Output

The above emotional brain patterns (EBPs) indicating brain well-being can be represented by the software in three ways:

1. As real-time graphs of the seven key output variables as depicted above.

2. As a simple "things are good" to "things are bad" symbol on the monitor screen, i.e. a smiling or frowning baby face. The facial expression reflects the baby's motivational direction (approach or avoidance) and the intensity of the facial expression will quantify the intensity of this motivational direction (strong to weak).

3. The motivational direction will be subtyped into discrete emotional states according to the details of the EBP, e.g. hunger cry, distress cry, lonely cry, comfort-feeding, hungry-feeding, unhappy-sad, unhappy-distressed, unhappy-dissociated, unhappy-lonely, unhappy-hungry, unhappy-angry. These states can be represented graphically or in a text format (typically a predetermined text format), for example.

The diagram of Figure 5 shows in a graphic form how the method and apparatus of the invention (dark arrows) operate in measuring and generating outputs based on the relevant parameters.

Other information

The optimal interpretation of the monitor outputs at present requires an "input" regarding the behavioural state of the baby (asleep, awake) as observed. This can be automated together with various other parameters that already exist, and only needs to be added to the algorithms to make the interpretations fully automated. Examples include body movements, eye movements and regularity of respiration.

Secondary parameters (all are well-known and technically well-established as such) which may be measured and represented to enhance the operation of the invention include:

1. SNS (from Impedance cardiography (ICG) measures the average thoracic impedance (over a user-defined period, typically 10-60 sees) time-locked to the last R-wave. This ongoing value (if used) will be represented as a fourth component of the emotional brain. ICG n Impedance CardioGraphic dZ/dt data. 128 samples of n points. Dz/dt is determined by:

dZ/dt = (n x 0.0010) ohm/sec.

2. Respiration.

3. Electrooculogram (EOG).

4. Motility in 3 dimensions.

5. Audio/video with automated behavioural analysis software.

6. Temperature.

7. Skin conductance level.

8. EEG asymmetry in other scalp areas.

9. EEG coherence.

The invention allows for the interpretation of the newborn's and the infant's basic needs, which is of fundamental importance for humane care. It allows for direct monitoring of sleep and the quality of sleep, which is increasingly being recognised as fundamental for brain development. Its use may lead to a vastly improved understanding of the newborn and the developing infant. It may be useful in treating colicky infants, feeding disorders, and psychologically disturbed infants. The limited data from the pilot show the infants to be entirely different from adults, and yet the feeding and sleeping schedules imposed on infants are based on adult behaviour patterns. This invention will allow the direct observation at the level of the emotional brain of what a normal infant should do and require thereby allowing for optimal neuroaffective care of ill infants. This is specifically important in the management of premature infants and low birth weight babies.

Claims

1. A method of monitoring the emotional state of an infant, the method including:

analysing the ECG data to determine a plurality of parameters including heart rate and at least one further parameter related to the interval between R-waves in an ECG waveform of the subject, and generating second output data related thereto;

processing the first and second output data to determine in real time or near-real time of the emotional state of the subject; and

providing an indication of the emotional state of the subject.

2. A method according to claim 1 wherein the at least one further parameter determined from the ECG data includes one or more of the following:

(a) Old vagus (very low frequency component of HRV);

(b) Sympathetic nervous system or SNS (low frequency component of HRV); (c) New vagus (high frequency component of HRV);

(d) Standard deviation of normal R-R intervals (SDNN) as a moving average.

3. A method according to claim 1 or claim 2 wherein at least one of the following parameters are measured and analysed to determine the emotional state of the subject:

(e) Infant behavioural/vigilance state using the ABSS;

(f) Instantaneous heart rate (=1/RR interval);

(g) Standard deviation of normal R-R intervals (SDNN) as a moving average.

4. A method according to any one of claims 1 to 3 wherein the indication of the emotional state of the subject comprises a graphical representation suggesting at least one of at least two possible emotional states.

5. A method according to any one of claims 1 to 4 wherein the indication of the emotional state of the subject comprise quantitative values obtained from recorded data, time lines of such data over a selected measurement period, or text-format interpretations of the subject's current state.

6. Apparatus for monitoring the emotional state of an infant, the apparatus including:

an input circuit for receiving and pre-processing EEG and ECG data acquired from a subject;

a processor arranged to analyse the EEG data to determine asymmetry of the frontal lobe activity and to generate first output data related thereto; to analyse the ECG data to

^ determine a plurality of parameters including heart rate and at least one further parameter related to the interval between R-waves in an ECG waveform of the subject, and to generate second output data related thereto; and to process the first and second output data to generate display data; and

7. Apparatus according to claim 6 including at least one module for analysing the EEG and ECG data, wherein a module for analysing the ECG data is arranged to determine at least one of the following further parameters from the ECG data:

(a) Old vagus (very low frequency component of HRV);

(b) Sympathetic nervous system or SNS (low frequency component of HRV);

(c) New vagus (high frequency component of HRV);

(d) Standard deviation of normal R-R intervals (SDNN) as a moving average.

8. Apparatus according to claim 7 wherein at least one module of the of the apparatus is arranged to measure and analyse one or more of the following parameters to determine the emotional state of the subject:

(e) Infant behavioural/vigilance state using the ABSS;

(f) Instantaneous heart rate (=1/RR interval);

(g) Standard deviation of normal R-R intervals (SDNN) as a moving average.

9. Apparatus according to any one of claims 6 to 8 wherein said at least one display is capable of displaying a graphical representation suggesting at least one of at least two possible emotional states.

10. Apparatus according to any one of claims 6 to 9 wherein said at least one display is capable of displaying an indication of the emotional state of the subject comprising quantitative values obtained from recorded data, time lines of such data over a selected measurement period, or text-format interpretations of the subject's current state.