GB2574293A

GB2574293A - Heart condition monitoring

Info

Publication number: GB2574293A
Application number: GB1903970.0A
Authority: GB
Inventors: Walinjkar Amit
Original assignee: RAYTEL SECURITY SYSTEMS Ltd; Raytel Security Systems Ltd
Current assignee: RAYTEL SECURITY SYSTEMS Ltd; Raytel Security Systems Ltd
Priority date: 2018-05-18
Filing date: 2019-03-22
Publication date: 2019-12-04
Anticipated expiration: 2039-03-22
Also published as: GB2574293B; GB201903970D0; GB201808131D0

Abstract

Heart condition is monitored in real-time using a system worn by a person, the system having cable-free electrodes 12 located on the body surface, at least some of the electrodes transmitting data wirelessly to a processing unit 15. The electrodes produce electrocardiography signals in analogue form, transform the signal into digital form and include transmission means to transmit the digital signals to the processing unit. The processing unit converts the digital signals into analogue form and processes them to create compositedigitised signals. An analysis of the composite digitised signals is then carried out on the processing unit using an algorithm to detect an arrhythmia. An alarm is generated if an arrhythmia is detected. Machine learning may determine the arrhythmia type. Spectral density values may be calculated to determine if the signals have been distorted during transmission. Parameters of the ECG signal may be analysed to determine irregularity, heart rate, P-wave morphology, PR interval and QRS wave detection. The wireless transmission may be Bluetooth (RTM).

Description

-1 Heart Condition Monitoring

This invention relates to a system for monitoring the heart condition of a person in real-time or near real-time, with the system being configured to be worn by the person. More specifically, the invention involves monitoring of a person in a passive, non-invasive and non-inhibiting way to detect, in real-time or near realtime, the presence of an arrhythmia, while the person is free to go about their day to day activities.

The heart comprises a muscle called myocardium which is rhythmically driven to contract and drives the circulation of blood throughout the body. Before every normal heartbeat, a wave of electrical current passes through the entire heart and the pattern of this electrical current and its propagation spreads over the entire structure of the heart and leads to an effective flow of blood in and out of the heart. The measurable change in bio-potential difference (voltage) on the human body surface due to this activity results in a waveform signal which is known as an electrocardiogram (ECG or EKG).

Heart disease can be caused by a mix of conditions which are complex to measure and to monitor. The ECG, representing the heart condition as a waveform of bio-electric potential, is the most important tool used in cardiac monitoring. There are various factors which affect the ECG, including (but not limited to) abnormalities in cardiac muscles, metabolic abnormalities of the myocardium, posture and state of human body, noise due to other bio-electric activities inside the human body and over the surface of the human skin.

There are several systems available for the monitoring of a person’s health which can be worn by that person. Existing wearable ECG monitoring systems

-2generally require 24 - 48 hours of continuous monitoring. Such systems do not have the ability to analyse the results nor to detect or alert the person of ECG abnormalities and cardiac risks ahead of time. Furthermore, many such systems have been implemented on hardware which restricts mobility so cannot be used by persons whilst they are engaged in their day to day activities, which is when they are most likely to suffer a cardiac arrest or a heart attack.

It is a principle aim of the present invention to provide a portable, wearable system for passively monitoring the heart condition of a person in real-time, in order to detect arrhythmia and to address at least some of the above-identified problems in existing monitoring systems and methods. That system will ideally be as small and unobtrusive as possible to make it wearable continuously without impinging on a user’s normal activities. The term “real-time” as used herein is intended to mean analysis of the results whilst the system is worn, within a reasonable time, taking into consideration sampling rates and other unavoidable minor delaying factors. This is in contrast to existing systems which require later analysis of the full results after the system has been worn, typically after at least 24 - 48 hours.

According to this invention, there is provided a system for monitoring the heart condition of a person in real-time, the system being configured to be worn by the person and comprising:

at least three cable-free electrodes for location on the body surface, each electrode being configured to produce electrocardiography signals in analogue form, each electrode having signal adaption means to transform the electrocardiography signals into digital form;

-3a processing unit configured to:

i. receive digital signals from the electrodes and to convert the digital signals into analogue signals;

ii. to process the analogue signals to create composite digitised signals; and iii. to carry out an analysis of the digitised signals using an algorithm-based methodology to detect an arrhythmia;

at least two of the electrodes comprising wireless transmission means to facilitate wireless transmission of the digital signals from the at least two electrodes to the processing unit; and an alarm configured to operate if arrhythmia is detected.

The person (wearer of the system) may be an individual who has already been diagnosed with arrhythmia or who has been experiencing worrying symptoms, or even a seemingly healthy individual with no prior history of arrhythmia. The system may be used as a tool for use by medical personnel to monitor patients or for use in the sports or fitness industry.

Each of the electrodes preferably include support means to maintain them secured in position on the body surface. The support means may comprise a self-adhesive pad, adhesive tape, adhesive (e.g. collodion), an elasticated or non-elasticated strap or an elasticated or non-elasticated band. The electrodes may advantageously be provided as part of a module. Each module preferably comprises an electrocardiogram (ECG) electrode for passively electrically contacting the body of the person, along with the signal adaption means and transmission means. Preferably the modules are low weight and are compact to minimise the perception and discomfort of the person when worn. Each

-4independent component would also have a power supply/charging unit in the form of a battery or the like.

An electrode or module may be contained within a housed unit. The, or each unit may be configured to contain more than one electrode and/or module. If more than one electrode and/or module is located in a single unit the electrodes still function independently and provide different data streams to the processing unit.

The module or unit (where applicable) and/or support means may be configured to receive an electrode gel to improve electrical connection to the skin, as is well-known in the art.

A clinical or ambulatory ECG system typically uses twelve electrodes located on the person’s body. The system of the present invention must comprise a minimum of three electrodes but may comprise more. Three electrodes have shown to achieve an accuracy of 97% whereas twelve electrodes have been shown to achieve an accuracy of 99%. The greater the number of electrodes the more accurate the analysis, but the less the mobility of the wearer.

Typically, a heart attack or cardiac arrest occurs when a person is active (i.e. not in a relaxed state). The ability of the processing unit to analyse the digitised signals in real-time or near real time and to initiate an alarm if an anomaly is identified ensures that the presence of an arrhythmia can be detected during normal day to day activities almost instantaneously and often before a heart condition deteriorates. The ability of the system to work without restricting the wearer’s movements is an important feature. This not only reduces any discomfort to the wearer but also provides a true reflection of their heart condition

-5during their normal activities. Any anomalies caused by motion artefacts are corrected by algorithms within the system. Many existing technologies involve the use of electrical cables extending between the electrodes of an ECG system. These can be a hinderance to the wearer and can affect their ability to carry out their daily activities in a normal way. The electrodes of the present invention are cable-free. This means that there are no external cables linking them, although there may be wires or other electrical connections internally of each unit.

It will be appreciated that, in order to function, the system for monitoring the heart condition of a person, as discussed herein, must carrying out of various steps, as follows:

a) monitoring heart electrical activity from at least three electrodes located on the body surface to produce electrocardiography signals in analogue form, each electrode being cable-free and incorporating signal adaption means to transform the signal into digital form and including transmission means to facilitate transmission of the signal;

b) transforming the electrocardiography signals, by way of the signal adaption means, into a digital form for transfer thereof to a processing unit;

c) transmitting the digitised signals, by way of the transmission means, to the processing unit, at least two of the electrodes transmitting the signals wirelessly;

d) using the processing unit to convert the digitised signals into analogue form and to process the analogue signals to create composite digitised signals;

e) carrying out an analysis of the digitised signals on the processing unit using an algorithm-based methodology to detect an arrhythmia; and

f) generating an alarm if an arrhythmia is detected.

All of the electrodes may include wireless transmission means to facilitate wireless transmission of the signal in step c) to the processing unit. Preferably the wireless transmission means has a low power consumption and is low cost. The transmission means may comprise RF or another wireless protocol. Preferably, the transmission means is a Bluetooth or Bluetooth Low Energy (BLE) module with the digitised signals being transmitted in step b) by way of Bluetooth protocol. Bluetooth and BLE transmission protocols require a lower amount of power to operate than most other wireless protocols and are short-range.

In one arrangement of the present invention, the processing unit may be portable (or mobile). In this arrangement, the processing unit may be carried around by the wearer, for example in their pocket, or simply kept close by. Where a short-range transmission means is adopted, such as Bluetooth or BLE, the processing unit should be maintained in close proximity to the person. As an alternative arrangement, at least one of the electrodes may be physically integrated with the processing unit. In this arrangement, that integrated electrode may not require wireless transmission means as a physical connection between that electrode and the processing unit is already provided within the integrated unit, that physical connection being a cable free connection (i.e. with no external cables linking them). The remaining electrodes will however still require wireless transmission means for the transfer of the digital signals from those electrodes to the processing unit.

The method and system of the present invention preferably utilise appropriate algorithms for analysis of the signals in real time. Such algorithms

-7may be produced and used in conjunction with an open source database comprising a collection of normal and abnormal arrhythmia signal samples. The database utilised in the development of the method and system of the present invention is the Massachusetts Institute of Technology, Beth Israel Hospital Arrhythmia Database, otherwise known as the MIT-BIH arrhythmia database. The MIT-BIH database also includes an open source library, known as the MITBIH WFDB that can extract features from digitized ECG signals. It will be appreciated that this database and library are known and do not form part of the invention per se; other types of database and or library which are suitable for the purpose may equally be used. The algorithms may be configured to correct any anomalies caused by motion of the wearer.

The distinguishing of normal waveforms from abnormal waveforms using learning-based methods (such as neural networks) can be extremely useful in predicting, based on recognition, the presence of an arrhythmia. However, when analysing in real-time with a portable (mobile) system (in contrast to a large computer-based system) such a process can be problematic due to the power and data requirements in order to learn from each of the signal samples. To address this, the Applicant has discovered that abnormal waveforms can also be distinguished from normal waveforms using spectral analysis and more particularly by comparative analysis of a set of features of the waveforms (health parameters), including the power spectral density (PSD) values. In the present invention a database library (such as the MIT-BIH database library) may be ported to the system’s portable/wearable processing unit and used for feature extraction based on ECG morphology. Algorithms may also be produced in order

-βίο determine, based on the signals, a set of health parameters and the power spectral density (PSD) values. The database library and algorithms are preferably pre-processed and ported to the processing unit during manufacture or setup.

The analysis of the digital signal on the processing unit in step e) preferably comprises a determination of the values of five distinct health parameters. Those five distinct health parameters may comprise the following:

1. irregularity detection

2. heart-rate detection

3. P-Wave morphology detection

4. PR Interval detection

5. QRS wave detection

The irregularity detection can be determined by variations and standard deviations in RR interval; the RR interval is the time interval between consecutive heart beats. With heart-rate detection, the speed of the heartbeat is measured by the number of contractions of the heart per minute (bpm). Tachycardia is a fast heart rate, defined as above 10Obpm at rest. Bradycardia is a slow heart rate, defined as below 60bpm at rest. A normal, regular heart rate is between 60 100bpm. When the heart is not beating in a regular pattern, an arrhythmia is present.

The P-Wave is a waveform in an electrocardiogram tracing representing atrial depolarization. The QRS wave (also known as the QRS complex) is a series of waveforms on an electrocardiogram that represents both normal and abnormal depolarization of ventricular muscle cells. It is composed of Q, R, and

-9S waves: a Q wave is the negative deflection before the first R wave, an R wave is any positive deflection, and an S wave is the negative deflection after an R wave. The QRS wave is normally between 60 - 100ms. The PR interval is the period, measured in milliseconds, that extends from the beginning of the P wave (the onset of atrial depolarization) until the beginning of the QRS wave; it is normally between 120 and 200ms in duration. The PR interval is sometimes termed the PQ interval.

In step e) the values may be analysed by the processing unit to determine whether they reside within a threshold range, based on the pre-processed algorithms. For example, a normal waveform may produce values within the normal ranges, as follows:

Regular heart rate: between 60-100bpm

PR interval: between 120 and 200ms

QRS wave: between 60 - 100ms

Whereas an abnormal waveform representing a ventricular fibrillation may produce abnormal values, as follows:

Irregular heart rate: >140

PR interval: 0

QRS wave: 0/absent

If an anomaly is detected, the system may involve an additional step to confirm the presence and type of arrhythmia. In this case, the processing unit may then perform comparative operations on the composite digitised signals using a machine learning-based methodology in order to confirm arrhythmia and to determine the type. This additional step may also be performed as a

-10secondary check even if an anomaly is not detected by the five-point health parameter check. Preferably, the learning-based methodology comprises a comparison of the digitised signal with a set of standard arrhythmia profiles.

To achieve this, the assembly code is pre-trained utilising the MIT-BIH database to produce prediction programmes specifically designed to predict the presence of an arrhythmia. The assembly code is trained on normal and abnormal waveforms from individuals with four types of arrhythmia. The present invention preferably focusses on four types of arrhythmia which could be observed in seemingly healthy individuals for early stage detection. Preferably the processing unit contains the prediction programmes to recognise certain features of the different arrhythmias displayed in the four types of arrhythmia in order to detect and classify an arrhythmia. More types of arrhythmia could of course be observed using the system of the present invention; in such cases the assembly code may be trained on normal and abnormal waveforms from individuals with many types of arrhythmia

Preferably, the processing unit comprises a microcontroller to carry out the comparative analysis of the digitised signals. Preferably, the database library, algorithms and programmes running on the microcontroller extract the features from the test waveform in real time to analyse the signals and to determine the presence of an anomaly. The microcontroller may also then perform predictions using pre-defined prediction programmes.

Advantageously, the electrodes are designed not to inhibit the wearer and so are cable-free. As a result of such a configuration, the electrodes do not have a common ground source (or reference), and this can lead to problems related to

-11 phase distortion and common mode rejection. To circumvent these issues the signal adaption means preferably comprises a digital sampler so that in step b) the signal adaption means digitally samples the electrocardiography signals to transform the electrocardiography signals into a digital form.

The inventor has determined that a normal ECG signal and an abnormal ECG signal (tachycardic or fibrillated) in a sample vector have different spectral densities. The aim of the method and system of the present invention is to detect arrhythmia rather than simply to gather normal sinus rhythm. For this reason, the method preferably involves the passing of spectral information along with the samples. Specifically, calculations may be carried out on the electrodes in step b) to determine spectral density values of the signals and the spectral density values then transmitted by the transmission means, along with the digitised signals, in step c). The nature of wireless transmission can lead to distortions in the signal which can result in loss of transmission. Preferably, calculations are carried out by the processing unit before step d) to determine power spectral density (PSD) values of the received signals. Utilising this information, the processing unit may then compare the PSD values calculated in step b) with the PSD values of the received signals to determine whether the signal has been distorted during transmission. If the PSD values do not match and if they have excess deviation, then there has been some signal distortion and the vector frame must be discarded.

The processing unit may comprise an aggregator configured to receive the digital signals from the at least three electrodes and to convert the digital signals into analogue signals. The term aggregator should be interpreted as one or more

-12 components which serve to combine the signals. The aggregator may comprise a digital to analogue converter. The aggregator preferably also comprises wireless receiving means for the reception of the digital signal. The wireless receiving means must correspond to the wireless transmission means and may for example comprise a Bluetooth or BLE module. The aggregator may also comprise a digital signal processor (DSP) to reconstruct the signals received by the wireless receiving means. The calculations to determine the PSD values of the received signals may be carried out by the DSP. This may be achieved, prior to conversion, by interpolation in order to reconstruct the received signals. The DSP may then be configured to carry out the comparison between the PSD values calculated in step b) and the PSD values of the received signals to determine whether the signal has been distorted during transmission. The conversion process to reconstruct the signals ensures that the spectral energy contents of the signals have been preserved.

The processing unit preferably further comprises a front-end device configured to process the analogue signals into a form suitable for the microcontroller. The front-end device is preferably configured to create the composite digitised signals in step d). Preferably, the front-end device processes the signal to ensure that the digitised signal is sampled and digitized into the correct form.

The alarm may be configured to generate an audio or visual signal upon detection of an arrhythmia. Alternatively, or additionally, the alarm may comprise wireless transmission means configured to transmit an alert to an external device

-13if an arrhythmia is detected and indicating the type of arrhythmia. The external device may be a computer or device of an NHS or other healthcare system.

By way of example only, an embodiment of this invention will now be described in detail, reference being made to the accompanying drawings in which:Figure 1 is a perspective view of a person wearing electrode modules as a first configuration of a system for monitoring the heart condition of a person, in accordance with the present invention;

Figure 2 is a perspective view of the right and left wrist of a person wearing electrode units as a second configuration of a system for monitoring the heart condition of a person, in accordance with the present invention;

Figure 3 is a flow chart showing the main stages of the process of the system according to the present invention;

Figure 4 shows a system for monitoring the heart condition of a person in real-time according to the present invention; and

Figure 5 is a schematic representation showing the arrangement of the units of Figure 2, as utilised with the system of the present invention.

Referring initially to Figure 1, there is shown a first arrangement of a system 10 for monitoring the condition of their heart according to the present invention. The system 10 includes three electrocardiogram (ECG) electrodes 11 located on the body of a man 12. These electrodes 11 are shown positioned on the man’s chest, but the invention is not limited to any particular location; the electrodes 11 will be placed at the location most appropriate for obtaining the required ECG signals and may include the man’s wrists or ankles for example.

-14The electrodes 11 shown in Figure 1 are each provided as part of a module 13. Each module 13 contains a single electrode 11 to produce an electrocardiography signal and integrated signal adaption means 14 (see Figure 3) to transform the signal into digital form. The modules 13 shown in Figure 1 are identical and are connected to support means in the form of self-adhesive pads 15 which serve to ensure the modules 13 are maintained on the body. Generally the type of support means used depends on the area of the body where the modules 13 are located; if on the wrist area, for example, straps may be more suitable. The electrode modules 13 are cable-free and are lightweight. This allows the man 12 to carry on with everyday activities, for example gardening, without discomfort or system inaccuracies.

Figure 2 shows a second arrangement of a system 18 of the present invention comprising two units 19, 20, each one being connected to the respective wrist 21 of a person by way of support means in the form of a strap 22. In the arrangement shown, two electrodes modules 13 (not visible) are housed within one of the units 19 and one electrode module 13and a processing unit 25 (described in more detail below) are housed within the other unit 20. The units 19, 20 may be visually similar to that of a watch and may even comprise additional functionality so as to perform the function of a watch or other entertainment-based device. The units 19, 20 may or may not be identical.

Each electrode module 13 also comprises wireless transmission means 26 to transmit wirelessly the digital signal to a processing unit 25.

In the system 10 of Figure 1, a portable processing unit 25 is positioned in the pocket 27 of the man’s trousers and is arranged to receive the digital signals

-15transmitted from the electrode modules 13 and to analyse these signals to determine the presence (or not) of an arrhythmia.

As an alternative arrangement, as shown in the system 18 of Figures 2 and 5 (as will described in more detail below) one of the electrode modules 13 is integrated with the processing unit 25 and is housed in a single unit 20 for location on a region of the body, with the other two electrodes modules 13 being housed together within another unit 19. This arrangement would not require the wearer to remember to carry the processing unit around with them.

Referring now in more detail to Figures 3, 4 and 5, the system for monitoring the heart condition of a person will now be described together in more detail.

Each electrode module 13 comprises a silver chloride (AgCI) electrode combined with an integrated circuit chip programmed to acquire ECG waveform signals and to process these waveforms. As a first step the voltages are isolated from the electrodes, following which the signals acquired are then synchronised. The signals are then band limited for normal and abnormal waveforms and antialiasing. The signals are sampled in consecutive 10 second batches, filtered to remove noise and transmitted to the processing unit 15, along with information about the power signal densities (PSD) of the signals.

Referring to Figure 3, the system has two possible configurations. In the first configuration (referenced as (a) in Figure 3) all three electrode modules 13 are independent, with the integrated circuit chip in each configured to transmit the signals wirelessly, by way of Bluetooth to the processing unit 125. In the second configuration (referenced as (b) in Figure 3 and as shown in Figure 5) one of the

-16electrode modules 13 is physically integrated with the processing unit 25. In this second arrangement, only two of the electrode modules 13 need to transmit the signals wirelessly, by way of Bluetooth, although the integrated circuit chip on all three electrode modules 13 may still remain the same for simplicity.

The processing unit 25 comprises an aggregator 30, a front-end device 31 and a microcontroller 32. The aggregator 30 includes a Bluetooth receiver 33, an integrated microcontroller circuit chip including a digital signal processor (DSP) 34, and a digital to analogue converter (DAC) 35.

The signals are gathered by the Bluetooth receiver 33 over the three reception channels 36 (one per electrode) and these are passed to the microcontroller and DSP chip 34 to perform bandlimited interpolation in order to reconstruct the signal. The signals need to be reconstructed in order to determine the PSD of the received signals as this is the only way to detect loss in transmission. Only 3,600 samples are transmitted and received at a time. The PSDs are again calculated for the received samples and are matched with the transmitted samples to determine whether there has been any signal distortion. If the PSDs do not match the vector frame has to be discarded. The microcontroller circuit chip 34 combines the signals using multiplexing. The signals are then passed to the DAC 35 for conversion into analogue form.

The analogue signals 40 from the DAC 35 are passed to the front-end device 31 which then processes them to ensure that they are compatible with the requirements of the microcontroller 32 (as described in more detail below) by sampling and digitising. In order to eliminate aliasing effects, the signals were

-17 sampled and filtered and PSD calculated again for further analysis by the microcontroller 32.

The method and system utilise algorithms and programmes for analysis of the signals in real-time. The algorithms and programmes are produced and used in conjunction with an open source database such as that known as the MIT-BIH arrhythmia database (the Massachusetts Institute of Technology, Beth Israel Hospital Arrhythmia Database) comprising a collection of normal and abnormal arrhythmia signal samples. The MIT-BIH database also includes an open source library 41, known as the MIT-BIH WFDB that can extract features from digitized ECG signals. To produce the system and implement the method, the MIT-BIH library 41 is ported, during production, to the microcontroller 32 and used for feature extraction based on ECG morphology. Algorithms are also produced in order to determine, based on the extracted features, a set of health parameters and the power spectral density (PSD) values.

As a first step, the microcontroller 32 carries out a five-point parameter check to detect and identify arrhythmia. The microcontroller 32 uses the MIT-BIH WFDB library 41 to extract features from the test waveform in real-time. The five parameters include:

1. irregularity detection

2. heart-rate detection

3. P-Wave morphology detection

4. PR Interval detection

5. QRS wave detection

-18The values are then assessed against defined threshold values and a combination of the results are used by the microcontroller to detect a class of arrhythmia.

If the five-point check identifies an anomaly the microcontroller 32 may issue an alarm immediately and/or carry out further analysis to confirm the findings. Such further analysis is based on a learning methodology which comprises a comparison of the digitised signal with a set of standard arrhythmia profiles. To achieve this assembly code is pre-trained utilising the MIT-BIH database to produce prediction programmes specifically designed to predict the presence of an arrhythmia. The assembly code is trained on normal and abnormal waveforms from individuals with four types of arrhythmia. The present invention focusses on four types of arrhythmia which could be observed in seemingly healthy individuals for early stage detection. During production of the system, the prediction programmes are compiled on the microcontroller. The prediction programmes are designed to recognise certain features of the different arrhythmias displayed in the four types of arrhythmia in order to detect and classify an arrhythmia. The further learning-based analysis may also be carried out if the five-point check does not identify an anomaly as a confirmatory analysis.

If an anomaly is detected either during the five-point check or during the prediction program stage, the microcontroller 32 initiates an alert or a visual or audio alarm 42 to notify the person wearing the system and/or another party (such as a doctor or medical team) that an arrhythmia has been detected. An alert 43 may also be transmitted to an electronic health record database.

Claims

1. A system for monitoring the heart condition of a person in real-time, the system being configured to be worn by the person and comprising:

a processing unit configured to: receive digital signals from the electrodes and to convert the digital signals into analogue signals; to process the analogue signals to create composite digitised signals; and to carry out an analysis of the digitised signals using an algorithm-based methodology to detect an arrhythmia;

2. A system as claimed in claim 1, wherein the processing unit is configured also to perform comparative operations on the composite digitised signals using a machine learning based methodology in order to confirm arrhythmia and to determine the type.

3. A system as claimed in claim 1 or claim 2, wherein the processing unit comprises an aggregator configured to receive the digital signals from the at least three electrodes and to convert the digital signals into analogue signals.

4. A system as claimed in claim 3, wherein the aggregator comprises a digital to analogue converter.

5. A system as claimed in any of the preceding claims, wherein the processing unit further comprises a front-end device configured to process the analogue signals to create the composite digitised signals.

6. A system as claimed in any of the preceding claims, wherein the processing unit further comprises a microcontroller to carry out the analysis of the digitised signals.

7. A system as claimed in any of the preceding claims, wherein the transmission means comprises a Bluetooth module.

8. A system as claimed in any of the preceding claims, wherein the signal adaption means comprises a digital sampler.

9. A system as claimed in any of the proceeding claims wherein at least one of the electrodes is physically integrated with the processing unit.

10. A system as claimed in any of the claims 1 to 9 wherein each of the electrodes comprises transmission means to facilitate wireless transmission of the digital signals.

11. A system as claimed in any of the proceeding claims, wherein the alarm is configured to generate an audio or visual signal.

12. A system as claimed in any of the proceeding claims, wherein the alarm comprises wireless transmission means to transmit an alert to an external device.