Health monitor device, system and method

Abstract

A health monitor device (1) for monitoring the health and / or one or more vital signs of a user comprises: an attachment means (2) for attaching the health monitor device (1) to a limb or the neck of the user; a sensing means (3) configured to output raw sensor data, the sensing means (3) comprising at least a plurality of electrodes (31) for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user; a processing means (4) configured to produce processed output data from the raw sensor data; and a communication means for sending or reporting the raw sensor data and / or processed output data to one or more further devices. The processed output data comprises at least a first output signal indicative of the respiration of the user.

Description

The present application relates to health monitor devices and associated systems and methods for monitoring the health and / or one or more vital signs of a user. Background Electronic devices for monitoring the health and vital signs of patients are becoming increasingly important in modern healthcare. Such devices can be used to provide health professionals with important real-time data and information, allowing important clinical decisions to be made based on up-to-date information about a patient’s condition. In recent years the size of such devices has become generally smaller, allowing health data to be collected and recorded in a variety of settings, including outside of traditional healthcare venues such as hospitals and clinics. The ability to measure and record relevant health data over extended periods of time is particularly advantageous, since this allows detailed longitudinal health studies to be carried out more easily and efficiently. Respiration rate, heart rate and heart rate variability are three of the fundamental vital signs that can be used to assess the health status of a patient. The measurement of vital signs is valuable in revealing a personal emotional state in response to stimuli and environmental conditions. Continuous measurement of vital signs is particularly advantageous, especially for at-risk subjects such as the elderly, patients in recovery after surgery, or for those whose occupation submits them to stress, such as emergency service workers. Making simultaneous measurements of respiration rate and other health indicators / vital signs, such as those related to cardiac activity or body temperature, can be useful in determining the overall health of a patient. According to the prior art, respiration can be monitored using a band applied around the chest of a user to measure chest expansion and contraction, or by covering the mouth and nose with a mask having a transducer to measure pressure or air velocity inside the mask. Such systems are often invasive and can cause discomfort, especially when used for significant periods of time. The popular technique of photoplethysmography (PPG) has limitations in its application to the human body. Factors including skin tone and obesity can affect the accuracy of measurements which employ light emitting sensors. Other techniques using radar, acoustic, and ultrasound sensors are reliant on fixed placement of measuring devices relative to anatomical features. This is usually because the signal being emitted or received requires line of sight with an artery. There exists a need for an improved way of monitoring the health and / or one or more vital signs of a user. Summary of the Invention According to an aspect of the invention there is provided a health monitor device for monitoring the health and / or one or more vital signs of a user, the health monitor device comprising: an attachment means for attaching the health monitor device to a limb or the neck of the user; a sensing means configured to output raw sensor data, the sensing means comprising at least a plurality of electrodes for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user; a processing means configured to produce processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the respiration of the user; and a communication means for sending or reporting the raw sensor data and / or processed output data to one or more further devices. Advantageously, the health monitor device can be used to continuously monitor the health and / or one or more vital signs of a user, for example over extended periods of time. Optionally the processed output data comprises at least a bioimpedance signal. Optionally the processing means comprises a filtering means. Advantageously, the filtering means can be used to filter the raw sensor data and obtain one or more signals that are indicative of the health and / or one or more vital signs of a user. Optionally the processing means comprises a filtering means configured to obtain the first output signal by filtering the raw sensor data. Optionally the first output signal is a bioimpedance signal. Optionally the filtering means comprises a high-pass filter. Optionally the filtering means comprises a high-pass filter having a cutoff frequency of at least 5 kHz. Optionally the filtering means comprises a high-pass filter having a cutoff frequency of 10 kHz or 25 kHz. Optionally the filtering means comprises a demodulator, such as a synchronous demodulator. Optionally the filtering means comprises at least one bandpass filter. Optionally the filtering means comprises a first bandpass filter and a second bandpass filter. Optionally the filtering means comprises a first bandpass filter. Advantageously, the first bandpass filter can be used to isolate a signal that is indicative of the respiration of a user. Optionally the filtering means comprises a first bandpass filter having a higher cutoff frequency of 0.5-2 Hz. Optionally the filtering means comprises a first bandpass filter having a higher cutoff frequency of 0.6 Hz. Optionally the filtering means comprises a first bandpass filter having a lower cutoff frequency of 0.01-0.4 Hz. Optionally the filtering means comprises a first bandpass filter having a lower cutoff frequency of 0.1 Hz. Optionally the filtering means comprises a first bandpass filter having a lower cutoff frequency of 0.01-0.4 Hz and a higher cutoff frequency of 0.5-2 Hz. Optionally the filtering means comprises a first bandpass filter having a lower cutoff frequency of 0.1 Hz and a higher cutoff frequency of 0.6 Hz. Optionally the processing means is further configured to obtain a second output signal indicative of the cardiac activity of the user. Optionally the filtering means is configured to obtain the second output signal by filtering the raw sensor data. Optionally the second output signal is a bioimpedance signal. Optionally the filtering means comprises a second bandpass filter. Advantageously, the second bandpass filter can be used to isolate a signal that is indicative of the cardiac activity, such as the heart rate and / or heart rate variability, of a user. Optionally the filtering means comprises a second bandpass filter having a lower cutoff frequency of 0.5-2 Hz. Optionally the filtering means comprises a second bandpass filter having a lower cutoff frequency of 0.8 Hz. Optionally the filtering means comprises a second bandpass filter having a higher cutoff frequency of 5-20 Hz. Optionally the filtering means comprises a second bandpass filter having a higher cutoff frequency of 8 Hz. Optionally the filtering means comprises a second bandpass filter having a lower cutoff frequency of 0.5-2 Hz and a higher cutoff frequency of 5-20 Hz. Optionally the filtering means comprises a second bandpass filter having a lower cutoff frequency of 0.8 Hz and a higher cutoff frequency of 8 Hz. Optionally the processing means is further configured to obtain a third output signal. Optionally the third output signal is an electrocardiogram (ECG) signal or electrodermal activity (EDA) signal. Optionally the filtering means is configured to obtain the third output signal by filtering the raw sensor data. Optionally the filtering means comprises a low pass filter. Advantageously, the low pass filter can be used to isolate a low-frequency signal, such as an ECG or EDA signal that is indicative of the cardiac activity of a user. Optionally the filtering means comprises a low-pass filter having a cutoff frequency of less than 1000 Hz. Optionally the filtering means comprises a low-pass filter having a cutoff frequency of 150 Hz or 500 Hz. Optionally the attachment means is an attachment means for attaching the health monitor device to a limb or the neck of the user. Optionally the attachment means is an attachment means for attaching the health monitor device to an arm of the user. Optionally the attachment means is an attachment means for attaching the health monitor device to an upper arm of the user. Optionally the attachment means is an attachment means for attaching the health monitor device to a neck of the user. Optionally the attachment means comprises a strap. Advantageously, the strap can be used to apply the plurality of electrodes to the skin of the user. Optionally the strap is a rigid or flexible band or collar. Optionally the attachment means comprises hook-and-loop fasteners. Optionally the attachment means comprises an elasticated band, such as an arm band or collar. Advantageously, the elasticated band may be provided in various sizes to accommodate a range of limbs / necks of various users. Optionally the attachment means comprises a patch, for example a flexible patch that may be attached or applied to the skin of a user. Optionally the attachment means comprises a flexible substrate. Advantageously, the flexible substrate can be applied to the skin of the user for long-term or shortterm usage. Optionally the plurality of electrodes are provided on the flexible substrate. Optionally the attachment means comprises an adhesive layer. Advantageously, the adhesive layer allows the flexible substrate and / or electrodes to be applied to the user. Optionally the adhesive layer comprises a hypoallergenic adhesive. Optionally the adhesive layer is provided on the flexible substrate. Optionally the plurality of electrodes are printed, coated or otherwise attached to the flexible substrate. Optionally the attachment means comprises a flexible substrate coated with an adhesive, such as a hypoallergenic adhesive, optionally with the electrodes printed or coated on the substrate. Optionally the plurality of electrodes are suitable for application to the skin of a user. Optionally the plurality of electrodes are dry electrodes. Optionally the plurality of electrodes are solid metallic electrodes, such as stainless steel electrodes or similar. Optionally the plurality of electrodes are conductive textile electrodes, for example woven into the attachment means. Optionally the plurality of electrodes are connected to the attachment means, in use. Optionally the plurality of electrodes comprises at least three electrodes. Advantageously, three electrodes can be applied to the skin of a user to apply current and measure voltage. Optionally the plurality of electrodes comprises an array of electrodes. Advantageously, the array of electrodes can be applied to the skin of a user. Further advantageously, one or more electrodes in the array may be selectable. Optionally the plurality of electrodes comprises more than four electrodes. Advantageously, a subset of the electrodes may be selected for a 3-point or a 4-point measurement. Optionally the health monitor device further comprises an electrode selection means. Optionally the electrode selection means is configured to determine and / or select a subset of the plurality of electrodes in the electrodes array. Optionally the electrode selection means is configured to determine and / or select an optimal plurality of electrodes for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user. Advantageously, the optimal plurality of electrodes may be selected to optimise the measurement of impedance and / or voltage. Optionally the electrode selection means is configured to analyse the raw sensor data and / or processed output data for a first plurality and a second plurality of electrodes. Optionally the electrode selection means is configured to determine the optimal plurality of electrodes based on the analysis of the raw sensor data and / or processed output data. Optionally the electrode selection means is configured to automatically determine and / or select the optimal plurality of electrodes. Advantageously, automatic determination and / or selection of the best electrode set from an array of electrodes greater than four removes the need for removal and repositioning of the device to produce an optimal signal(s). Optionally the sensing means further comprises one or more contact means. Optionally the or each contact means comprises one or more electrodes, such as dry electrodes. Optionally the or each contact means comprises one or more resilient electrodes. Optionally the or each resilient electrode is configured for extending through hair. Optionally the or each resilient electrode comprises an elongate resilient member. Optionally the or each elongate resilient member is impregnated with electrically conductive material. Optionally the or each resilient electrode comprises an electrically-conductive coating. Optionally the plurality of contact means are suitable for application to the skin of a user. Optionally the plurality of contact means are connected to the attachment means, in use. Optionally the plurality of contact means comprises at least three contact means. Advantageously, three contact means can be applied to the skin of a user to apply current and measure voltage. Optionally the plurality of contact means comprises an array of contact means. Advantageously, the array of contact means can be applied to the skin of a user. Further advantageously, one or more contact means in the array may be selectable. Optionally the plurality of contact means comprises more than four contact means. Advantageously, a subset of the contact means may be selected for a 3-point or a 4-point measurement. Optionally the sensing means further comprises optical sensing means. Optionally the sensing means further comprises optical sensing means for measuring optical transmittivity or reflectivity. Optionally the optical sensing means is configured for measuring optical transmittivity or reflectivity of a user’s skin and the tissue beneath the user’s skin, including the blood flowing in the vessels beneath the skin. Optionally the optical sensing means comprises one or more optical emitters for applying an optical signal to the skin of the user and to the tissue beneath the skin of a user. Optionally the optical sensing means comprises one or more optical detectors for measuring an optical response. Optionally the sensing means further comprises one or more of an accelerometer and a temperature sensor. Advantageously, further sensors can be used to improve the accuracy of the measurement of the conditions of a user. Optionally the communication means is a wired communication means. Optionally the communication means is a wireless communication means. Advantageously, the use of a wired or wireless communication means allows sensor data and / or processed output data to be transmitted from the health monitor device. Optionally the one or more further devices comprise at least a mobile communications device. Advantageously, the mobile communications device can be used to perform further data processing and / or receive user input and / or present data to a user. Optionally the one or more further devices comprise at least a server. Advantageously, the server can be used to perform further data processing, for example processing data from a plurality of health monitor devices. According to another aspect of the invention there is provided a system for monitoring the health and / or one or more vital signs of a user, the system comprising: a health monitor device; and one or more further devices in wireless or wired communication with the health monitor device, wherein the one or more further devices are configured to receive sensor data and / or processed output data from the health monitor device. Advantageously, the system allows the health and / or one or more vital signs of a user to be monitored and communicated over a network. Optionally the health monitor device and / or one or more further devices is configured to provide an alarm or notification. Advantageously, the alarm or notification may be indicative of the health and / or one or more vital signs of a user. According to a further aspect of the invention there is provided a method for processing data, the method comprising: receiving raw sensor data from a sensing means, the sensing means comprising at least a plurality of electrodes for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user; and producing processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the respiration of the user. Advantageously, the method can be used to process data indicative of the present health condition of a user. Optionally producing processed output data comprises: processing the raw sensor data received from the sensing means. Optionally producing processed output data comprises: processing raw sensor data received from an impedance sensor. Optionally producing processed output data comprises: filtering the raw sensor data to obtain the first output signal. Optionally the first output signal has a frequency content of more than 0.01 Hz, or more than 0.1 Hz, or more than 0.4 Hz. Optionally the first output signal has a frequency content of less than 2 Hz, or less than 0.6 Hz, or less than 0.5 Hz. Optionally the first output signal has a frequency content of between 0.0.1 Hz and 2 Hz. Optionally the first output signal has a frequency content of between 0.1 Hz and 0.6 Hz. Optionally producing processed output data from the raw sensor data comprises: filtering the raw sensor data to obtain a second output signal. Optionally the second output signal is indicative of the heart rate and / or heart rate variability of a user. Optionally the second output signal has a frequency content of more than 0.5 Hz, or more than 0.8 Hz, or more than 2 Hz. Optionally the second output signal has a frequency content of less than 5 Hz, or less than 8 Hz, or less than 20 Hz. Optionally the second output signal has a frequency content of between 0.5 Hz and 20 Hz. Optionally the second output signal has a frequency content of between 0.8 Hz and 8 Hz. Optionally the method further comprises selecting a first plurality of electrodes from the array of electrodes. Optionally the method further comprises applying an alternating current and measuring a voltage using the first plurality of electrodes. Optionally the method further comprises analysing the raw sensor data output from the first plurality of electrodes. Optionally the method further comprises selecting a second plurality of electrodes from the array of electrodes. Optionally the method further comprises applying an alternating current and measuring a voltage using the second plurality of electrodes. Optionally the method further comprises analysing the raw sensor data output from the second plurality of electrodes. Advantageously, the method can be used to identify an optimal plurality of electrodes for use in a measurement. Optionally the sensing means comprises an array of electrodes, and the method further comprises: selecting a first plurality of electrodes from the array of electrodes; applying an alternating current and measuring a voltage using the first plurality of electrodes; analysing the raw sensor data output from the first plurality of electrodes; and selecting a second plurality of electrodes from the array of electrodes; applying an alternating current and measuring a voltage using the second plurality of electrodes; analysing the raw sensor data output from the second plurality of electrodes. Advantageously, the method can be used to identify an optimal plurality of electrodes for use in a measurement, without necessarily having to move the electrodes between measurements. Optionally the method further comprises: selecting an optimal plurality of electrodes. Optionally the method further comprises: selecting an optimal plurality of electrodes based on the analysis of the raw sensor data output from the first plurality of electrodes and the second plurality of electrodes. Optionally receiving raw sensor data from the sensing means comprises: receiving raw sensor data from an impedance sensor. Optionally receiving raw sensor data from the sensing means comprises: receiving raw sensor data from an electrocardiogram (ECG) sensor. Optionally receiving raw sensor data from the sensing means comprises: receiving raw sensor data from an optical sensing means. Optionally receiving raw sensor data from the sensing means comprises: receiving raw sensor data from one or more of an accelerometer and a temperature sensor. Optionally the method further comprises sending or reporting the raw sensor data and / or processed output data to one or more further devices. Optionally the method further comprises sending or reporting the raw sensor data and / or processed output data to a mobile communications device or a server. According to a further aspect of the invention there is provided a health monitor device for monitoring the health and / or one or more vital signs of a user, the health monitor device comprising: an attachment means for attaching the health monitor device to an upper arm or neck of the user; a sensing means configured to output raw sensor data, the sensing means comprising at least a plurality of electrodes for applying an alternating current and measuring a voltage; a processing means configured to produce processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the health and / or one or more vital signs of a user; and a communication means for sending or reporting the raw sensor data and / or processed output data to one or more further devices. According to a further aspect of the invention there is provided a health monitor device for monitoring the health and / or one or more vital signs of a user, the health monitor device comprising: a sensing means configured to output raw sensor data; a processing means configured to produce processed output data from the raw sensor data; and a communication means for sending or reporting the raw sensor data and / or processed output data to one or more further devices. According to a further aspect of the invention there is provided a system for monitoring the health and / or one or more vital signs of a user, the system comprising: a health monitor device; and one or more further devices in wireless or wired communication with the health monitor device. According to a further aspect of the invention there is provided a method for processing data, the method comprising: receiving raw sensor data from a sensing means; and producing processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the health and / or one or more vital signs of a user. Any feature or features described in relation to any aspect, embodiment or example may be combined with any one or more features of any other aspect, embodiment or example. Brief Description of the Drawings The invention will be described by way of example only referring to the figures, in which: Figure 1 shows a schematic view of a health monitor according to an aspect of the invention. Figure 2A shows a side view of a health monitor according to an aspect of the invention. Figure 2B shows a front view of the health monitor of figure 2A attached to the arm of a user. Figure 3 shows a perspective view of a health monitor according to an aspect of the invention. Figure 4 shows a bioimpedance measurement setup. Figure 5 shows the frequency dependence of typical ECG and bioimpedance measurements. Figure 6 shows a bioimpedance and ECG measurement setup. Figure 7 shows a sensor arrangement on an arm of a user. Figure 8 shows a further sensor arrangement on an arm of a user. Figure 9 shows an attachment arrangement having a plurality of sensors thereon, for arrangement on a limb as per figure 7. Figure 10 shows a further attachment arrangement having a plurality of sensors thereon. Figure 11 shows a further attachment arrangement having a plurality of sensors thereon, for arrangement on a limb as per figure 8. Figure 12 shows a further attachment arrangement having a plurality of sensors thereon. Figure 13A shows an attachment arrangement having a plurality of sensors thereon. Figure 13B shows an alternative configuration of the attachment arrangement and plurality of sensors of figure 13A. Figure 14A shows a further attachment arrangement having a plurality of sensors thereon. Figure 14B shows an alternative configuration of the attachment arrangement and plurality of sensors of figure 14A. Figure 15A shows a further attachment arrangement having a plurality of sensors thereon. Figure 15B shows an alternative configuration of the attachment arrangement and plurality of sensors of figure 15A. Figure 15C shows a further alternative configuration of the attachment arrangement and plurality of sensors of figure 15A. Figure 16 shows a system according to an aspect of the invention, and a plurality of users. Figure 17 shows a method according to an aspect of the invention. Figure 18 shows example raw impedance sensor data. Figure 19 shows example raw and filtered impedance sensor data. Figure 20 shows a further attachment arrangement having a plurality of sensors thereon. Figure 21 shows a further attachment arrangement having a plurality of sensors thereon. Figure 22A shows a front view of a health monitor according to an aspect of the invention. Figure 22B shows a rear view of the health monitor of figure 22A. Figure 22C shows the health monitor of figure 22A applied to the neck of a user. Figure 23 shows a schematic view of a health monitor according to an aspect of the invention. Figure 24 shows front and side views of an example contact arrangement for use with aspects of the invention. Figure 25 shows a sensor arrangement on the neck of a user. Figure 26 shows an alternative sensor arrangement on the neck of a user. Figure 27 shows a further alternative sensor arrangement on the neck of a user. Figure 28 shows a bioimpedance measurement setup. Figure 29 shows example raw and filtered impedance sensor data. Figure 30 shows example frequency-dependent data. Figure 31A shows a front view of a health monitor according to an aspect of the invention. Figure 31B shows a rear view of the health monitor of figure 31 A. Figure 31C shows the health monitor of figure 31A applied to the neck of a user. Figure 32 shows example raw and filtered impedance sensor data. Figure 33 shows example frequency-dependent data. Detailed Description In figure 1 there is shown a health monitor device 1 according to an aspect of the invention. The health monitor device 1 comprises: an attachment arrangement 2 for attaching the health monitor device to a limb (such as an upper arm) or the neck of the user; a sensing arrangement 3 configured to output raw sensor data, for example from a plurality of sensors; a processing unit 4 configured to produce processed output data from the raw sensor data; and a communication module 5 for sending or reporting the raw sensor data and / or processed output data to one or more further devices. The health monitor device 1 further comprises: a source of electrical power 6, such as a battery; a memory 7 for storing computer instructions, raw sensor data and / or processed output data; and an electrode selection unit 8. The health monitor device 1 is suitable for monitoring the health and / or one or more vital signs of a user. In examples the processed output data generated by the health monitor device 1 comprises at least a first output signal indicative of the respiration of a user. Additionally, the processed output data generated by the health monitor device 1 can further comprise a second output signal indicative of the cardiovascular activity of the user. Furthermore, the processed output data generated by the health monitor device 1 can include processed output data and signals from other sensors, such as electrocardiogram (ECG) sensors, electrodermal activity (EDA) sensors, photoplethysmography (PPG) sensors, electrochemical impedance spectroscopy (EIS) sensors, accelerometers and / or temperature sensors. The processing unit 4 comprises a filtering arrangement 41. The filtering arrangement 41 can be used to analyse the raw sensor data and obtain one or more signals that are indicative of the health and / or one or more vital signs of a user. The filtering arrangement 41 is configured to obtain output signals by filtering and / or demodulating raw sensor data received from the sensing arrangement 3. The filtering arrangement 41 comprises a first bandpass filter 42, a second bandpass filter 43, a high-pass filter 44, a low-pass filter 45 and a synchronised demodulator 46. The various cutoff frequencies of the filters 42, 43, 44 and 45 may be adjusted to be suitable for the particular user of the health monitor device 1. The filtering arrangement 41 is configured to obtain the first output signal by filtering the raw sensor data (i.e. the output of the sensing arrangement 3) using the high-pass filter 44, demodulating the output of the high-pass filter 44 using the synchronous demodulator 46, and filtering the resultant demodulated signal using the first bandpass filter 42. Advantageously, the first bandpass filter 42 can be used to isolate a signal that may be analysed to determine the respiration of a user. In examples (see e.g. figure 6), the high-pass filter 44 has a cutoff frequency fc of 5-50 kHz (e.g. 10 kHz or 25 kHz), and the first bandpass filter 41 has a lower cutoff frequency / i. of 0.01-0.4 Hz (e.g. 0.1 Hz) and a higher cutoff frequency fa of 0.5-2 Hz (e.g. 0.6 Hz). These cutoff frequencies may be adjusted to be suitable for the particular user of the health monitor device 1. The processing unit 4 is further configured to obtain a second output signal indicative of the cardiac activity of the user. The filtering arrangement 41 is configured to obtain the second output signal by filtering the raw sensor data (i.e. the output of the sensing arrangement 3) using the high-pass filter, demodulating the output of the high-pass filter 44 using the synchronous demodulator 46, and filtering the resultant demodulated signal using the second bandpass filter 43. The filtering arrangement 41 comprises a second bandpass filter 43 that can be used to isolate a signal that may be analysed to determine the cardiac activity, such as the heart rate and / or heart rate variability, of a user. In examples (see e.g. figure 6), the high-pass filter 44 has a cutoff frequency fc of 5-50 kHz (e.g. 10 kHz or 25 kHz), and the second bandpass filter 43 has a lower cutoff frequency / i. of 0.5-2 Hz (e.g. 0.8 Hz) and a higher cutoff frequency of 5-20 Hz (e.g. 8 Hz). These cutoff frequencies may be adjusted to be suitable for the particular user of the health monitor device 1. The processing unit 4 may be further configured to obtain a third output signal, also indicative of the cardiac activity of the user. The filtering arrangement 41 is configured to obtain the third output signal by filtering the raw sensor data (i.e. the output of the sensing arrangement 3) using the low-pass filter 45. The output signal of the low-pass filter 45 includes an ECG component and may be analysed to determine the cardiac activity, such as the heart rate and / or heart rate variability, of a user. In examples (see e.g. figure 6), the low-pass filter 45 has a cutoff frequency fc of 50-1000 Hz (e.g. 150 Hz or 500 Hz). The cutoff frequencies may be adjusted to be suitable for the particular user of the health monitor device 1. The processing unit 4 comprises an analysis unit 49. The analysis unit 49 is configured to process, analyse and interpret the filtered waveforms i.e. the output(s) of the filtering arrangement 41. For example, the analysis unit 49 may be used to determine respiration rate, heart rate and / or heart rate variability of a user based on the outputs of the first and / or second bandpass filters 42,43. Furthermore the analysis unit 49 may be used to monitor changes in the output of the low-pass filter 45, the first bandpass filter 42 and / or second bandpass filter 43 over time. As will be appreciated, the analysis unit 49 may be part of the processing unit 4 and local to the health monitor device 1. In optional embodiments, the analysis unit 49 may be remote from the health monitor device 1 and may be located for example in a remote server, such as an RF-linked server. In use, the health monitor device 1 can be attached to e.g. the upper arm or neck of a user for measuring the health and / or one or more vital signs of the user. In particular, the attachment arrangement 2 is configured for attachment to the upper arm or neck of a user such that at least a part of the sensing arrangement 3 is applied to the skin of the user. Figures 2A and 2B provide views of the example health monitor 1 according to an aspect of the invention. In this example, the attachment arrangement 2 comprises an elasticated strap 21. The strap 21 is a flexible band comprising hook-and-loop fasteners. For example, the hook-and-loop fasteners may be provided on part of the strap 21, for example at one end of the strap 21. The strap 21 can be wrapped around a limb of the user 10, such as their upper arm 11 (figure 2B), and held securely in place by the hook-and-loop fasteners. By ‘upper arm’ it is meant that portion of the arm between the shoulder and elbow. The strap 21 may be an elasticated band that can be provided in various sizes to accommodate a range of limbs / necks of various users. In this example the sensing arrangement 3 comprises a plurality of dry electrodes 31 attached to the strap 2. In other words, the dry electrodes 31 are in a fixed location on the strap 21. In examples the plurality of electrodes are solid metallic electrodes, such as stainless steel electrodes or similar. Alternatively, the electrodes may be conductive textile electrodes, for example woven into the strap 21. Alternatively the electrodes may be printed onto the band using conductive ink, or be moulded using conductive rubber or plastic material. In use, the strap 21 applies / holds the plurality of electrodes to the skin 12 of the user 10, and pushes the electrodes into contact with the skin of the user to allow long-term monitoring of the user 10. The processing unit 4, communication module 5, battery 6, and memory 7 are provided in a single case, unit or box which is also attached to the strap 21 (see figure 2A). The health monitor device 1 is adapted to be applied to the upper arm of the user, which is a comfortable position to allow long-term monitoring to take place. As will be appreciated, the strap 21 also allows the health monitor device 1 to be attached to other parts of the user, for example, the neck, forearm or leg of the user, for measuring the health and / or one or more vital signs of the user. Furthermore, the strap 21 may allow the health monitor device 1 to be attached to the limbs or necks of other animals, for example the leg or arm of other mammals, for measuring the health and / or one or more vital signs of said animals. Figure 3 shows an alternative embodiment of the health monitor device 1a comprising a rigid attachment arrangement 2a. In this example, the sensing arrangement 3 comprises a plurality of electrodes 31 which are located on an inner surface of the rigid attachment arrangement 2a. The processing unit 4, communication module 5, battery 6, and memory 7 are attached to an outer surface of the rigid attachment arrangement 2a. The sensing arrangement 3 / electrodes 31, processing unit 4, communication module 5, battery 6, memory 7 and electrode selection unit 8 are the same as those of the previous embodiment. Returning to figure 1, the health monitor device 1 comprises a sensing arrangement 3 configured to output raw sensor data. The sensing arrangement 3 comprises a plurality of n dry electrodes 31 (i.e. 31 a to 31 n) for applying a current to the skin of a user and for measuring a voltage across the skin of said user. As will be appreciated, in embodiments the number electrodes n may be any suitable number (e.g. 4, 6, 10, 20, 30, >30, etc) and the sensing arrangement 3 may comprise e.g. more than four electrodes, more than ten electrodes, more than thirty electrodes, etc. The sensing arrangement 3 further comprises a current source 32 and a voltage detector 33. The sensing arrangement may further comprise an amplifier unit 34 for amplifying the output of the voltage detector 33 and / or electrodes 31. The sensing arrangement 3 can be used for measuring bioimpedance of the skin of a user. In particular, the sensing arrangement 3 can be used for 3-point or 4-point measurements of bioimpedance. The electrode selection unit 8 is configured to determine and / or select a subset of the plurality of electrodes 31. The electrode selection unit 8 is configured to determine and / or select an optimal plurality of electrodes 31 for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user. Advantageously, the optimal plurality of electrodes 31 may be selected to optimise the measurement of impedance and / or voltage. In examples, an optimal plurality of electrodes 31 for measuring ECG signals may be identified. In an example, the electrode selection unit 8 is configured to analyse the raw sensor data and / or processed output data for at least first plurality of electrodes and a second plurality of electrodes. Additional pluralities of electrodes, for example a third plurality of electrodes, may be included in the analysis. The electrode selection unit 8 is configured to determine the optimal plurality of electrodes 31 based on the analysis of the raw sensor data and / or processed output data. While the electrode selection unit 8 is shown as separated from the processing unit 4 for clarity, in some embodiments the electrode selection unit 8 and the processing unit 4 may be combined into a single unit. The principle of measuring bioimpedance will now briefly be explained with reference to figures 4-8. Figure 4 shows a current source 32 (current drive) connected to two driving electrodes 31 a,31 d, and a voltage detector 33 connected to two measuring electrodes 31 b,31c. Each of the electrodes 31 is in contact with the skin 12 of a user 10. The driving electrodes 31 a,31 d are used to apply an alternating current (e.g. a 0.1 mA pk-pk, 50 kHz signal) to the skin 12 of the user 10, while the voltage detector 33 is configured to measure an associated potential difference between the two measuring electrodes 31b,31c, which are also applied to the skin 12 of the user 10. Assuming that the pk-pk amplitude of the driving current is constant, the voltage measured by the voltage detector 33 is proportional to the impedance of underlying body tissue. As is explained in further detail below, bioimpedance varies with respiratory and cardiac activity, and can be used to infer e.g. the breathing rate, heart rate and / or heart rate variability of the user 10. As will be appreciated, figure 4 shows an example of a four-point measurement. Three-point measurements can be achieved by combining two of the electrodes, so that one electrode (e.g. 31a) is both a driving electrode (i.e. connected to the current source 32) and a measuring electrode (i.e. connected to the voltage detector 33 also), while two other electrodes (e.g. 31c,31d) are connected to the other terminals of the current source 32 and voltage detector 33, respectfully. Three-point measurements are dominated by skin impedance under the combined current / voltage electrode, which makes detection of variations due to respiration or cardiac activity difficult. However, three-point measurements can be used to enhance stress estimation using the change in skin impedance. The signal measured by the voltage detector 33 comprises a plurality of frequency components, including both bioimpedance signals as well as e.g. ECG and EDA signals. Thus a plurality of distinct signals can be received at the voltage detector 33 from the two voltage measurement electrodes 31b,31c. The typical frequencies of such bioimpedance and ECG signals are shown in the example of figure 5. The low-frequency / ECG signals can be isolated using a low-pass filter below e.g. 150 Hz or 500 Hz, while bioimpedance signals can be isolated using a high-pass filter above e.g. 10 kHz or 25 kHz. In this way, the same pair of electrodes may be used for both ECG measurements and the voltage pickup for bioimpedance. However, if multiple electrodes are available, it may be advantageous to optimally select a different pair of electrodes for ECG measurements, instead of the pair selected for voltage monitoring for bioimpedance. Figure 6 shows in detail an arrangement for measuring bioimpedance which facilitates the measurement of an ECG signal also. Figure 6 shows a current source 32 (current drive) connected to two driving electrodes 31 a,31 d, and an amplifier 34 connected to two measuring electrodes 31b,31c. Each of the electrodes 31 is in contact with the skin 12 of a user 10. The driving electrodes 31 a,31 d are used to apply an alternating current (e.g. a 0.1 mA pk-pk, 50 kHz signal) to the skin 12 of the user 10, while the amplifier 34 is configured to output a signal which is proportional to the potential difference between the two measuring electrodes 31b,31c, which are also applied to the skin 12 of the user 10. The output of the amplifier 34 is provided to the processing unit 4. Figure 5 shows only the filtering arrangement 41 of the processing unit 4; the other components of the processing unit 4 have been omitted from this figure, for clarity. As shown in figure 6, the signal received by the processing unit 4 from the amplifier 34 is provided in parallel to the low-pass filter 45 and the high-pass filter 44. This arrangement allows the separation of the low-frequency signals (including ECG signal) from the higher-frequency bioimpedance signals. The output of the low-pass filter 45 (including ECG signal) is provided as an output of the filtering arrangement 41 / processing unit 4, and can be provided to e.g. the analysis unit 49 for further processing. The output of the high-pass filter 44 is input to a synchronous demodulator 46, which is synchronised to the current source 32. The synchronous demodulator 46 is used to extract the amplitude and phase of the changing BioZ measurement from the filtered output of the high-pass filter 44. The output of the synchronous demodulator 46 is provided in parallel to the first bandpass filter 42 and the second bandpass filter 43. The outputs of the first and second bandpass filters 42, 43 may be provided as outputs from the filtering arrangement 41 / processing unit 4, and optionally may be provided to e.g. the analysis unit 49 for further processing. In embodiments, the analysis unit 49 of the health monitor 1 can be used to perform Fourier transforms on output of the filtering arrangement 41 / processing unit 4 and thereby produce the frequency-domain data. The analysis unit 49 may be used to derive arrhythmias from the raw bioimpedance signal, to detect abnormal or meaningful changes in respiration, heart rate or heart rate variability, and / or to estimate stress using respiration, heart rate, and / or heart rate variability in the manner described below. In use, the filtering arrangement 41 allows a plurality of signals, each of which may be indicative of the health and / or vital signs of a user, to be measured using the sensing arrangement 3. In the example of figure 6, the filtering arrangement 41 provides a first signal (“Respiration Signal (BioZ)”) that is indicative of the respiration of a user, a second signal (“Heart Signal (BioZ)”) that is indicative of the cardiac activity, such as the heart rate and / or heart rate variability, of a user, and a third signal (“ECG Signal”) which is also indicative of the cardiac activity of a user Figures 7 and 8 show illustrative examples of electrodes that have been attached to the upper arm 11 of a user 10. Electrodes can be applied in a variety of orientations, including transverse orientations (i.e. perpendicular or substantially perpendicular to the axis of the upper arm, as shown in figure 7) and longitudinal orientations (i.e. parallel or substantially parallel to the axis of the upper arm, as shown in figure 8), among others. Varying the position and orientation of the electrodes on the upper arm 11 of the user 10 will result in the electrodes 31 being closer to or further from e.g. the brachial artery 701 of the user 10, which will in turn affect the signal-to-noise ratio of the measurements. In the setups shown in figures 7 and 8, the quality of the measured signal will change depending on the locations of the sensors. Finding the optimal arrangement of sensors typically involves individually moving the electrodes 31 to different locations on the user and repeatedly measuring the corresponding output signal, to compare with previous results. As will be appreciated, such a scheme can be laborious and time consuming. To solve such problems the health monitor 1 comprises a sensing arrangement 3 comprising an array of dry electrodes 31 attached to the attachment arrangement 2. The electrodes 31 can be easily applied to the skin of a user in a range of positions, and can be moved to different positions by moving the attachment arrangement 2. Figures 9 to 15 show illustrative examples of sensing arrangements 3 (in particular electrode arrays) that may be employed in the health monitor device 1. Each sensing arrangement 3 comprises a plurality of dry electrodes 31 that may be applied to the skin of a user to apply an alternating current and measure a voltage response signal. In figures 9 to 15, example arrays of electrodes 31 are shown in relation to their positions on the attachment arrangement 2, to which the electrodes 31 are attached, in use. For clarity, other features of the health monitor device 1 are not shown in these figures. Figure 9 shows an example electrode array 900 comprising four dry electrodes Sia-Sid. The four electrodes of this embodiment are aligned and spaced along an axis (i.e. the longest axis) of the attachment arrangement 2. In use, the attachment arrangement 2 is wrapped around the upper arm of a user such that the electrodes are in contact with the skin 12 of the upper arm 11 of the user 10. In use, the electrodes 31 will be spaced circumferentially around the upper arm 11 of the user 10, in a situation analogous to that shown in figure 7. To find the optimal position for applying current / measuring voltage (i.e. the position where the measured signal is strongest and / or the signal-to-noise ratio is largest), the electrode array 900 can be moved around the arm by adjusting the position of the health monitor device 1 / attachment arrangement 2 on the upper arm 11 of the user 10. For example, the attachment arrangement 2 can be moved closer to or further from the elbow of the user, or can be rotated around the upper arm 11 of the user 10. Figure 10 shows a further example electrode array 1000 comprising four dry electrodes 31a-d. The electrode array 1000 of figure 10 is similar to the example electrode array 900 shown in figure 9. The electrode array of figure 10 is distinguished from the example of figure 9 by the size of the electrodes 31: in the case of figure 10, the electrodes 31a-d are larger so that the electrodes 31 are in contact with a greater surface area of the skin of the user 10. In particular, extending the length of the electrodes makes longitudinal placement of the array 1000 on the arm less critical, as the greater area is more easily applied over the brachial artery 701, and provides an improvement in signal strength relative to noise. Figure 11 shows a further example electrode array 1100 comprising four dry electrodes 31a-d. The four electrodes 31 of this embodiment are spaced along a direction that is perpendicular to the axis (i.e. the longest axis) of the attachment arrangement 2. In use, the attachment arrangement 2 is wrapped around the upper arm 11 of a user 10 such that the electrodes 31 are in contact with the skin 12 of the upper arm 11 of the user 10. The electrodes 31 will be spaced longitudinally along a direction parallel to the axis of the upper arm of the user, in a situation analogous to that shown in figure 8. To find the optimal position for applying current / measuring voltage, the electrode array 1100 can be moved on the upper arm 11 of a user 10 in a similar manner as explained above in relation to the example of figure 9. Figure 12 shows a further example electrode array 1200 comprising four electrodes. The electrode array of figure 12 is similar to the example electrode array 1100 shown in figure 11. The electrode array of figure 12 is distinguished from the example of figure 11 by the size of the electrodes: in the case of figure 12, the electrodes 31 are larger so that the electrodes 31 are in contact with a greater surface area of the skin of the user 10. The variation of impedance is stronger along the brachial artery which is aligned longitudinally along the arm; the longer electrodes make circumferential alignment less critical. Figures 13A and 13B show two measurement configurations of an example electrode array 1300 comprising more than four electrodes 31. In particular, the electrode array 1300 comprises eight electrodes 31a-31h. The eight electrodes Sia-31 h of this embodiment are spaced along the longest axis of the attachment arrangement 2. In use when the attachment arrangement 2 is wrapped around the upper arm 11 of a user 10 such that the electrodes are in contact with the skin 12 of the upper arm 11 of the user 10, the electrodes will be spaced circumferentially around the upper arm 11 of the user 10, in a situation analogous to that shown in figure 7. The electrodes in the electrode array 1300 are individually selectable so that, in use, current can be applied to any pair of electrodes in the array and voltage can be measured between any other pair of electrodes in the array. For example, the electrode selection unit 8 can initially select one set of electrodes according to a first measurement configuration (Figure 13A) and subsequently the electrode selection unit 8 can select a second set of electrodes according to a second measurement configuration (Figure 13B). This allows a comparison to be made between the signal measured using one set of electrodes vs the signal measured using another set of electrodes, to find an optimal configuration which provides the strongest signal or greatest signal-to-noise ratio. As will be appreciated, having more than four electrodes 31 in the array 1300 allows such a comparison to be made without necessarily having to move the position of the health monitor device 1 on the arm of the user. To find the optimal position for applying current / measuring voltage, the electrode array 1300 can also be moved on the upper arm 11 of a user 10 in a similar manner as explained above in relation to the example of figure 9. Figures 14A and 14B show an example electrode array 1400 comprising more than four electrodes. In particular, the electrode array 1400 comprises six electrodes 31a- 31 f. The six electrodes 31 of this embodiment are spaced along the shorter axis of the attachment arrangement 2 so that, in use when the attachment arrangement 2 is wrapped around the upper arm 11 of a user 10 such that the electrodes 31 are in contact with the skin 12 of the upper arm 11 of the user 10, the electrodes 31 will be spaced longitudinally along the upper arm 11 of the user 10, in a situation analogous to that shown in figure 8. The electrodes 31a-31f in the electrode array 1400 are individually selectable so that, in use, current can be applied to any pair of electrodes in the array and voltage can be measured between any other pair of electrodes in the array, in a similar manner to that described above in relation to figures 13A and 13B. For example, the electrode selection unit 8 can initially select one set of electrodes according to a first measurement configuration (Figure 14A) and subsequently the electrode selection unit 8 can select a second set of electrodes according to a second measurement configuration (Figure 14B). This allows a comparison to be made between the signal measured using one set of electrodes vs the signal measured using another set of electrodes, to find which configuration provides the strongest signal or greatest signal-to-noise ratio. As will be appreciated, such a comparison can be made without necessarily having to move the position of the health monitor on the arm of the user. To find the optimal position for applying current / measuring voltage, the electrode array 1400 can also be moved longitudinally on the upper arm 11 of a user 10 in a similar manner as explained above in relation to the example of figure 9. Figures 15A, 15B and 15C show an example electrode array 1500 comprising more than four electrodes. In particular, the electrode array 1500 comprises sixteen electrodes 31. The sixteen electrodes 31 of this embodiment are arranged in a matrix on the attachment arrangement 2 so that, in use when the attachment arrangement 2 is wrapped around the upper arm 11 of a user 10 such that the electrodes 31 are in contact with the skin 12 of the upper arm 11 of the user 10, the electrodes will be spaced longitudinally and circumferentially over an area of the upper arm 11 of the user 10. The electrodes 31 in the electrode array 1500 are individually selectable so that, in use, the current can be applied to any pair of electrodes in the array and voltage can be measured between any other pair of electrodes in the array, in a similar manner to that described above in relation to figures 13A and 13B. For example, the electrode selection unit 8 can initially select one set of electrodes according to a first measurement configuration (Figure 15A) and subsequently the electrode selection unit 8 can select a second set of electrodes according to a second measurement configuration (e.g. configurations shown in either of Figure 15B or Figure 15C). This allows a comparison to be made between the signal(s) measured using one set of electrodes vs the signal(s) measured using another set of electrodes, to find which configuration provides the strongest signal or greatest signal-to-noise ratio. As will be appreciated, such a comparison can be made without necessarily having to move the position of the health monitor on the arm of the user. To find the optimal position for applying current / measuring voltage, the electrode array 1500 can also be moved on the upper arm 11 of a user 10 in a similar manner as explained above in relation to the example of figure 9. The electrodes 31 in the electrode arrays 900,1000,1100,1200,1300,1400 and 1500 may be used for measuring ECG signals in addition to bioimpedance. For example, the electrode array 1500 of figure 15B shows two electrodes 31 labelled ‘G’. These electrodes are used for ECG pickup. Where multiple electrodes are available, it may be advantageous to optimally select a different pair of electrodes for ECG pickup, instead of the pair selected for voltage monitoring for BioZ. Dependent on the physiology of the subject, optimum electrode combinations may not be in a longitudinal line or transverse column for both or either of bioimpedance or ECG acquisition; the electrode selection mechanism may allow for independent selection of any suitable combination of the relevant electrodes. The electrode selection mechanism may for example use an array of switches that can selectively connect the electrodes to the current source 32 and voltage detector 33 or amplifier 34. Figure 15C provides an example selection of a plurality of electrodes that are not in a longitudinal line or transverse column for either of bioimpedance measurements (see electrodes 31 labelled “C” and “V”) or ECG acquisition (see electrodes 31 labelled “G”). Figure 16 shows a system 1600 for monitoring the health and / or one or more vital signs of a user 10 according to an aspect of the invention. The system 1600 comprises: a health monitor device 1; and one or more further devices 1602 in wireless or wired communication with the health monitor device 1. The one or more further devices 1602 are configured to receive sensor data and / or processed output data from the health monitor device 1. The system 1600 allows the health and / or one or more vital signs of one or more users to be monitored and communicated over a network. As shown in figure 16, the one or more further devices 1602 are in wireless or wired communication with the health monitor device 1 b of a further user 10b, wherein the health monitor device 1 b of the further user 10b is similar to the health monitor device 1 of the user 10. In the example of figure 16 the communication arrangement of the health monitor device 1 is a wireless communication arrangement, for example via Bluetooth or WiFi (e.g. for short-range communication) or LoRa (e.g. for longer-range communication), allowing raw sensor data and / or processed output data from the health monitor device 1 to be transferred wirelessly to the one or more further devices 1602. In alternative embodiments, the communication arrangement of the health monitor device 1 may be a wired communication arrangement, wherein one or more communication wires extend from the health monitor device 1 to the one or more further devices 1602 for data communication via a wire. The one or more further devices 1602 comprise at least a mobile communications device and / or a server in wireless communication with the health monitor device 1. The mobile communications device or server can be used to perform further data processing, for example processing data from a plurality of health monitor devices 1. In some examples, the health monitor device 1 and / or one or more further devices 1602 can be configured to provide an alarm or notification which may be indicative of the health and / or one or more vital signs of the user 10. Figure 17 discloses a method 1700 for processing data according to an aspect of the invention. The method 1700 can be carried out by the health monitor device 1. The method 1700 comprises: receiving raw sensor data from a sensing arrangement (step 1704); and producing processed output data from the raw sensor data (step 1706). Optionally, the method further comprises: selecting a sensing arrangement configuration (step 1702) and transmitting the raw or processed output data (step 1708). Advantageously, the method 1700 can be used to process data indicative of the present health condition of a user, and is suitable for long-term health monitoring. At step 1702, the method 1700 comprises selecting a sensing arrangement configuration. The sensing arrangement 3 of the health monitor 1 comprises at least a plurality of electrodes 31 for applying an alternating current to the skin of a user and measuring a voltage across the skin of the user. Step 1702 may be carried out by the electrode selection module 8 of the health monitor device 1. Where the sensing arrangement 3 comprises exactly four electrodes, step 1702 involves selecting those four electrodes by the electrode selection unit 8. Considering the example electrode array 900 of figure 9, the outer two electrodes 31 a,31 d in the array 900 may be selected as the current electrodes (denoted ‘C’) and the inner two electrodes 31b,31c selected as the voltage measurement electrodes (denoted ‘V’). Alternatively three electrodes may be chosen if a three-point measurement scheme is to be used. Where the sensing arrangement comprises more than four electrodes, step 1702 involves selecting a subset of the plurality of electrodes by the electrode selection unit 8. Considering the example electrode array 1400 of figure 14A, the outer two electrodes 31 a,31 f in the array 1400 may be selected as the current electrodes (denoted ‘C’) and the next two inner electrodes 31b,31e selected as the voltage measurement electrodes (denoted ‘V’). Alternatively three electrodes may be chosen if a three-point measurement scheme is to be used. The electrode selection unit 8 can be configured to determine and / or select, at step 1702, an appropriate subset of the plurality of electrodes in the electrode array for applying an alternating current to the skin of the user and for measuring a voltage across the skin of the user. An optimal plurality of electrodes can be selected to optimise the measurement of impedance and / or voltage. The electrode selection unit 8 can be configured to determine and / or select, at step 1702, a set of four electrodes that will provide raw sensor data having the strongest signal and / or largest signal-to- noise ratio during a continuous bioimpedance measurement. The electrode selection unit 8 can be configured to automatically determine and / or select the optimal plurality of electrodes. Advantageously, automatic determination and / or selection of the best electrode set from an array of electrodes greater than four removes the need for removal and repositioning of the health monitor device 1 to produce an optimal signal(s). For example, as part of step 1702, the method 1700 may include selecting a first plurality of electrodes from the array of electrodes. For example, the method may involve selecting the first plurality of electrodes marked ‘C’ and ‘V’ in figure 13A (or figure 14A, or figure 15A). An alternating current may be applied between the electrodes marked ‘C’ and a voltage may be measured between the electrodes marked ‘V’. The raw sensor data output from the first plurality of electrodes may then be analysed. A second plurality of electrodes from the array of electrodes may then be selected. For example, the method may involve selecting the second plurality of electrodes marked ‘C’ and ‘V’ in figure 13B (or figure 14B, or figure 15B). An alternating current may then be applied to the electrodes marked ‘C’ and a voltage may be measured between the electrodes marked ‘V’. The raw sensor data output from the second plurality of electrodes may then be analysed. By comparing the raw sensor data output from the first plurality of electrodes with the raw sensor data output from the second plurality of electrodes, an optimal plurality of electrodes can be identified, without necessarily having to move the position of the health monitor device on the arm of the user. For example, if the raw sensor data measured using the first plurality of electrodes has a larger signal or larger signal-to-noise ratio than the raw sensor data measured using the second plurality of electrodes, then the first plurality of electrodes can be determined as the optimal set of electrodes. The first and second pluralities of electrodes may be chosen by the electrode selection unit 8 at random, or may be chosen in a specific order. For example, a specific order may be saved in the memory 7. The electrode selection unit 8, when identifying the optimal set of electrodes, may carry out measurements using further (e.g. third, fourth, fifth) pluralities of electrodes, and may carry out measurements for all possible combinations of electrodes in the electrode array. The combinations tested may be limited to those in which the electrodes selected for measuring voltage are between the electrodes selected for applying current. As will be appreciated, the electrode array 1500 shown in figures 15A, 15B and 15C is particularly advantageous, since it allows a wide variety of configurations to be tested, and allows comparisons to be made between e.g. longitudinally-spaced arrangements of sensors (figure 15A), circumferentially spaced arrangements of sensors (figure 15B) and other arrangements (e.g. figure 15C). The refinement shown in Figure 15C provides an example configuration where the selected electrodes are not in a longitudinal line or transverse column for either of bioimpedance measurements (see electrodes 31 labelled “C” and “V”) or ECG acquisition (see electrodes 31 labelled “G”). As will be appreciated, an optimal electrode set may be identified from any combination of the two-dimensional electrode array. The optimal set of electrodes can be selected at step 1702, and the raw sensor data received at step 1704 may correspond to raw sensor data collected using the optimal set of electrodes. At step 1704, the method 1700 comprises receiving raw sensor data from the sensing arrangement. Considering the example of figure 6, the raw sensor data received at step 1704 may correspond to e.g. the direct output of the measuring electrodes 31b,31c, the output of the amplifier 34, and / or the filtered and demodulated output of the synchronised demodulator 46. Example raw sensor data 1800 is shown in figure 18. The raw sensor data 1800 is a time-varying impedance signal measured using the sensing arrangement of the health monitor 1. The raw sensor data 1800 is raw impedance sensor data, i.e. it is an example of the output of the demodulator 46 shown in figure 6. The example impedance signal 1800 is a measurement of the bioimpedance of the upper arm of the user. The example impedance signal 1800 comprises a plurality of frequency components. The impedance signal 1800 includes at least a lower-frequency component which is indicative of respiratory activity, and a high-frequency component indicative of cardiac activity. At step 1706, the method 1700 comprises producing processed output data from the raw sensor data. In particular, step 1706 comprises filtering the raw sensor data to provide a plurality of signals indicative of the condition of a user. Step 1706 may be carried out by a processing unit, for example the processing unit 4 of the health monitor device 1. Alternatively, a further device may receive and process the raw sensor data from the sensing arrangement 3 of the health monitor device 1. Figure 19 provides an illustrative example of producing processed output data from raw sensor data. Panel A of figure 19 shows example raw impedance sensor data 1900 received from the sensing arrangement 3 / processing unit 4. In this example, the raw impedance sensor data 1900 corresponds to the filtered and demodulated output of the synchronised demodulator 46 shown in figure 6. Panel B of figure 19 shows a first signal 1902 that has been isolated from the raw sensor data 1900 by a first bandpass filter 42. In this example, the first bandpass filter 42 has a lower cutoff frequency of 0.1 Hz and a higher cutoff frequency of 0.6 Hz. The resultant signal, i.e. a first output signal 1902, has a frequency component that is the same as the frequency of the respiration of the user. As shown in panel B, the signal 1902 output from the first bandpass filter 42 has an amplitude of 70 mQ pk-pk and a frequency of 16.7 breaths per minute. Panel C of figure 19 shows a second, higher frequency, signal 1904 that has been isolated from the raw sensor data 1900 by the second bandpass filter 43. In this example, the second bandpass filter 43 has a lower cutoff frequency of 0.8 Hz and a higher cutoff frequency of 8 Hz. The resultant signal, i.e. the second output signal 1904, has a frequency component that is the same as the frequency of the heart rate of the user. As shown in panel C, the signal output from the second bandpass filter has an amplitude of 22 mQ pk-pk and a frequency of approximately 72 beats per minute. Returning to figure 17, at step 1708 the method 1700 comprises transmitting the raw or processed output data, for example to a further device of the one or more further devices 1602 shown in figure 16. The one or more further devices may include a mobile communications device or a server. Step 1708 may be carried out by a communication module, for example the communication module 5 of the health monitor device 1. At step 1708, the method may additionally or alternatively comprise transmitting an alarm or notification to a further device, for example one or more of the further devices 1602 shown in figure 16. The processing unit 4 may be used to monitor the first output signal 1902 or second output signal 1904. For example, the processing unit 4 may compare the first output signal 1902 and / or second output signal 1904 with one or more thresholds. When the frequency of the first output signal 1902 and / or frequency of the second output signal 1904 falls below a predetermined threshold, the processing unit 4 may generate an alarm or notification. In an illustrative example, the processing unit 4 may compare the frequency of the first output signal 1902 with a threshold frequency. When the frequency of the first output signal 1902 falls below the threshold frequency, this may indicate that a user has stopped breathing, or that their breathing rate has fallen below a safe level. In this example, the processing unit 4 may generate an alarm or notification indicating that the user’s breathing is irregular and should be checked. As will be appreciated, the analysis and generation of a notification or alarm may be carried out at a further device, for example one or more of the further devices 1602 shown in figure 16. Figure 20 shows an example of a sensing arrangement 2000 comprising electrodes 31 and an optical sensing arrangement 35. The electrodes 31 are similar to those described above with respect to figure 9, and are used in a similar way. The optical sensing arrangement 35 can be used for measuring the optical transmittivity and / or reflectivity of the user’s skin and the tissue beneath the user’s skin, including the blood flowing in the vessels beneath the skin. The optical sensing arrangement 35 comprises an optical emitter (“E”) 35a for applying an optical signal to the skin of the user and to the tissue beneath the skin of a user, and an optical detector (“D”) 35b for measuring an optical response. Figure 21 shows an example of a sensing arrangement 2100 comprising electrodes 31 and a plurality of optical sensing arrangements 35. The electrodes 31 are similar to those described above with respect to figures 15A-15C, and are used in a similar way. The optical sensing arrangements 35 are similar to the optical sensing arrangement 35 described above with respect to figure 20 and are used in a similar way. As will be appreciated, the sensing arrangement 2100 includes multiple optical emitters 35a and multiple optical emitters 35b (i.e. four emitter-detector pairs 35a,35b), allowing e.g. an optimal emitter-detector pair 35a,35b to be identified and chosen for measurements. The optimal emitter-detector pair may be identified by: comparing the measurements made using each emitter-detector pair 35a,35b; and identifying the emitter-detector pair 35a,35b for which the measured signal is strongest and / or the signal-to-noise ratio is largest as the optimal emitter-detector pair. The optimal emitter and optimal detector may be chosen separately, and the optimal emitter-detector pair may not correspond to a neighbouring pair. Figures 22A to 22C provide views of another example health monitor 101 according to an aspect of the invention. The health monitor 101 is shown schematically in figure 23, and is generally similar to the previous health monitor 1 and can be used in a similar way. Similar reference numerals (e.g. 8, 108) denote similar features. As shown in figure 23, the health monitor 101 includes an attachment arrangement 102, a sensing arrangement 103, a processing unit 104, communication module 105, battery 106, memory 107 and electrode selection unit 108. The processing unit 104, communication module 105, battery 106, memory 107 and electrode selection unit 108 are provided in a single case, unit or box which is attached to the strap 121 (see figure 22A). As shown in figure 23, the single case also includes a display screen which is in communication with e.g. the processing unit 104. The communication module 105, battery 106, memory 107 and electrode selection unit 108 are similar to those of the previous embodiment. The processing unit 104 includes an alternative filtering arrangement 141 (shown in detail in figured 28). In alternative embodiments, the processing unit 104 may include, additionally or alternatively, the filtering arrangement 41 shown in figure 6. In the example illustrated in figures 22A to 23, the attachment arrangement 102 comprises an elasticated strap 121. The strap 121 is a flexible band comprising hook-and-loop fasteners. For example, the hook-and-loop fasteners may be provided on part of the strap 121, for example at one end of the strap 121. The strap 121 can be wrapped around a limb of the user 10, such as their neck 13 (figure 22C), and held securely in place by the hook-and-loop fasteners. By ‘neck’ it is meant that portion of the body between the head and upper chest / shoulders. The strap 121 may be a collar and / or an elasticated band that can be provided in various sizes to accommodate a range of limbs / necks of various users. In the present example the sensing arrangement 103 comprises a plurality of n contact arrangements 131 (i.e. 131 a to 131 n) for applying a current to the skin of a user and for measuring a voltage across the skin of said user. As will be appreciated, in embodiments the number of contact arrangements n may be any suitable number (e.g. 4, 6, 10, 20, 30, >30, etc) and the sensing arrangement 103 may comprise e.g. more than four contact arrangements, more than ten contact arrangements, more than thirty contact arrangements, etc. The plurality of contact arrangements 131a-131n are attached to the attachment arrangement 102. In the present example, the contact arrangements 131 are provided in two groups of four, each contact arrangement 131 in a single group being attached to a retaining member 122 which can be moved along the strap 121 of the attachment arrangement 102. The strap 121 passes through appropriately sized holes in each retaining member 122. This arrangement allows measurements to be taken using sensors on e.g. either side of the neck of the user. In other examples, the contact arrangements 131a-131n may be provided in a fixed location on the strap 121. In use, the strap 121 applies / holds the plurality of contact arrangements 131 to the skin of the user 10, and pushes at least a part of the contact arrangements 131 into contact with the skin of the user to allow monitoring of the user. Figure 24 shows an example contact arrangement 131 according to an aspect of the invention. Each contact arrangement 131a-131n shown in figure 22B may be similar to the example contact arrangement 131 shown in figure 24. The example contact arrangement 131 of figure 24 comprises a base or housing 135 and a plurality of resilient dry electrodes 136. The contact arrangement 131 may comprise more or fewer resilient dry electrodes 136, for example one resilient dry electrode 136. The base or housing 135 is configured to retain the plurality of dry electrodes 136. Each resilient dry electrode 136 comprises an elongate resilient member 137. Each elongate resilient member 137 is made from a resilient material, such as silicone or rubber. In optional embodiments, each elongate resilient member 137 is impregnated with electrically conductive material. For example, each elongate resilient member 137 may be moulded using a conductive rubber or plastic material. Additionally or alternatively, each elongate resilient member 137 comprises an electrically-conductive coating 138. For example, each elongate resilient member 137 may be at least partially coated in silver paint. Each resilient dry electrode 136 is configured for being used to make electrical contact with the skin of a user, and is configured for extending through hair, such as facial hair or other hair which may be present on or around the neck of a user. In particular, the or each elongate resilient member 137 is sized to extend through hair (e.g. beard hair) so that at least a part of the elongate resilient member 137 (e.g. the tip 137a) makes contact with the skin of a user, in use. The shape and configuration of the resilient dry electrodes 136 allows electrical contact to be made with the skin of a user, allowing e.g. bioimpedance measurements to be made. As will be appreciated, the resilient and flexible nature of the dry electrodes 136 ensures that a comfortable level of pressure is applied to the skin of the user, in use. The example contact arrangement 131 further comprises an electrical lead or connector 139 which is electrically connected to each of the resilient dry electrodes 136. The electrical lead or connector 139 can be used to electrically connect the resilient dry electrodes 136 to e.g. the other components of the sensing arrangement 103, such as a current source and a voltage detector, for example via suitable switches. In other examples the contact arrangements 131a-131n used in the health monitor 101 may be solid metallic electrodes, such as stainless steel electrodes or similar. Alternatively, the electrodes may be conductive textile electrodes, for example woven into the strap 121. Alternatively, the electrodes may be printed onto the band using conductive ink, or be moulded using conductive rubber or plastic material. In use, the strap 121 applies / holds the plurality of electrodes to the skin 12 of the user 10, and pushes the electrodes into contact with the skin of the user to allow longterm monitoring of the user 10. The health monitor device 101 is adapted to be applied to the neck of the user, which is a comfortable position to allow long-term monitoring to take place. As will be appreciated, the strap 121 also allows the health monitor device 101 to be attached to other parts of the user, for example, the upper arm, forearm or leg of the user, for measuring the health and / or one or more vital signs of the user. Furthermore, the strap 121 may allow the health monitor device 101 to be attached to the limbs or necks of other animals, for example the leg or arm of other mammals, for measuring the health and / or one or more vital signs of said animals. Figures 25 to 27 show illustrative examples of contact arrangements 131 a-131 d applied to the neck 13 of a user 10. Contact arrangements 131 a-131 d (each of which includes at least one electrode, such as a resilient dry electrode) can be applied in a variety of orientations, including transverse orientations (i.e. perpendicular or substantially perpendicular to the axis of the neck, as shown in figures 25 and 26) and longitudinal orientations (i.e. parallel or substantially parallel to the axis of the neck, as shown in figure 27), among others. Varying the position and orientation of the electrodes on the neck 13 of the user 10 will result in the electrodes being closer to or further from e.g. the carotid artery of the user 10, which will in turn affect the signal-to-noise ratio of the measurements. In the measurement arrangements shown in figures 25 to 27, the quality of the measured signal will change depending on the locations of the sensors. Finding the optimal arrangement of sensors typically involves individually moving the contact arrangements 131 (or groups thereof) to different locations on the user and repeatedly measuring the corresponding output signal, to compare with previous results. As will be appreciated, such a scheme can be laborious and time consuming. To solve such problems the health monitor 101 comprises a sensing arrangement 103 comprising an array of contact arrangements 131 a-131 n attached to the attachment arrangement 102. The contact arrangements 131 a-131 n can be easily applied to the skin of a user in a range of positions, and can be moved to different positions by moving the attachment arrangement 102 and / or by moving the contact arrangements 131a-131n on the attachment arrangement 102. Figure 28 shows in detail an arrangement for measuring bioimpedance using the sensing arrangement 103 and the processing unit 104. The sensing arrangement 103 can be used for measuring bioimpedance of the skin of a user. In particular, the sensing arrangement 103 can be used for 3-point or 4-point measurements of bioimpedance. Figure 28 shows only the filtering arrangement 141 of the processing unit 104; the other components of the processing unit 104 have been omitted from this figure, for clarity. The sensing arrangement 103 comprises a current source 132 (current drive) connected to two driving contact arrangements 131a,131d, and an amplifier 134 connected to two measuring contact arrangements 131b, 131c. Each of the contact arrangements 131 is in contact with the skin 12 of a user 10. The driving contact arrangements 131a,131d are used to apply an alternating current (e.g. a 0.1 mA pk-pk, 50 kHz signal; up to 1 MHz may be used) to the skin 12 of the user 10, while the amplifier 134 is configured to output a signal which is proportional to the potential difference between the two measuring contact arrangements 131b, 131c, which are also applied to the skin 12 of the user 10. The output of the amplifier 134 is provided to the processing unit 104. In optional embodiments, four-electrode raw bioimpedance measurements may be made using a driving frequency of up to 1 MHz using two current injection electrodes and two voltage sensing electrodes, to minimise the effect of individual electrode / skin impedance. As shown in figure 28, the signal received by the processing unit 104 from the amplifier 134 is provided to a high-pass filter 144. The output of the high-pass filter 144 is input to a synchronous demodulator 146, which is synchronised to the current source 132. The synchronous demodulator 146 is used to extract the amplitude and phase of the changing BioZ measurement from the filtered output of the high-pass filter 144. The output of the synchronous demodulator 146 is provided in parallel to the first bandpass filter 142 and the second bandpass filter 143. The outputs of the first and second bandpass filters 142, 143 may be provided as outputs from the filtering arrangement 141 / processing unit 104, and optionally may be provided to e.g. the analysis unit 149 for further processing. In use, the filtering arrangement 141 allows a plurality of signals, each of which may be indicative of the health and / or vital signs of a user, to be measured using the sensing arrangement 103. In the example of figure 28, the filtering arrangement 141 provides a first signal (“Respiration Signal (BioZ)”) that is indicative of the respiration of a user and a second signal (“Heart Signal (BioZ)”) that is indicative of the cardiac activity, such as the heart rate and / or heart rate variability, of a user. The alternative filtering arrangement 141 of figure 28 differs from the filtering arrangement 41 shown in figure 6 in that the alternative filtering arrangement does not include a low-pass filter and cannot be used to measure an ECG signal. In other examples, the alternative filtering arrangement 141 of figure 28 may include a low pass-filter, similar to low-pass filter 45 shown in figure 6, for isolating ECG signals from the output of the amplifier 134. As shown in e.g. figure 22B, the sensing arrangement 103 of the health monitor 101 comprises more than four contact arrangements 131. In particular, the sensing arrangement 103 comprises eight contact arrangements 131. The eight contact arrangements 131 of this embodiment are arranged in two groups of four on the attachment arrangement 102. In use, when the attachment arrangement 102 is wrapped around the neck 13 of a user 10 such that the electrodes of the contact arrangements 131 are in contact with the skin 12 of the neck 13 of the user 10, the electrodes will be spaced longitudinally and circumferentially over an area of the neck 13 of the user 10. The contact arrangements 131 in the sensing arrangement 103 are individually selectable so that, in use, current can be applied to any pair of contact arrangements in the array and voltage can be measured between any other pair of contact arrangements in the array, in a similar manner to that described above in relation to e.g. figures 15Ato 15C. For example, the electrode selection unit 108 can initially select one set of contact arrangements 131 according to a first measurement configuration and subsequently the electrode selection unit 108 can select a second set of contact arrangements 131 according to a second measurement configuration. This allows a comparison to be made between the signal(s) measured using one set of contact arrangements 131 vs the signal(s) measured using another set of contact arrangements 131, to find which configuration provides the strongest signal or greatest signal-to-noise ratio. As will be appreciated, such a comparison can be made without necessarily having to move the position of the health monitor 101 on the neck 13 of the user 10. To find the optimal position for applying current / measuring voltage, the sensing arrangement 103 can also be moved on the neck 13 of a user 10. Figure 29 provides an illustrative example of producing processed output data from raw sensor data. Panel A of figure 29 shows example raw impedance sensor data 2900 received from the sensing arrangement 103 / processing unit 104 of the health monitor 101. In this example, the raw impedance sensor data 2900 corresponds to the filtered and demodulated output of the synchronised demodulator 146 shown in figure 28. Figure 29, panel B, shows a first signal 2902 that has been isolated from the raw sensor data 2900 by the first bandpass filter 142. In this example, the first bandpass filter 142 has a lower cutoff frequency of 0.167 Hz and a higher cutoff frequency of 0.5 Hz. The resultant signal, i.e. the first output signal 2902, has an amplitude of 150 mQ pk-pk and a frequency component that is the same as the frequency of the respiration of the user. Figure 30, panel A, shows frequency-domain data 3002 produced from the first signal 2902 output from the first bandpass filter 142 (c.f. time-domain data shown in figure 29, panel B). This Fourier transformed data 3002 shown in figure 30, panel A, includes a peak at 0.2997 Hz, corresponding to a breathing rate of approximately 17 breaths per minute. Returning to figure 29, panel C shows a second, higher frequency, signal 2904 that has been isolated from the raw sensor data 2900 by the second bandpass filter 143. In this example, the second bandpass filter 143 has a lower cutoff frequency of 0.7 Hz and a higher cutoff frequency of 3 Hz. The resultant signal, i.e. the second output signal 2904, has an amplitude of 170 mQ pk-pk and a frequency component that is the same as the frequency of the heart rate of the user. Figure 30, panel B, shows frequency-domain data 3004 produced from the second signal 2904 output from the second bandpass filter 143 (c.f. time-domain data shown in figure 29, panel C). This Fourier transformed data 3004 shown in figure 30, panel B, includes a peak at 0.99901 Hz, corresponding to a heart rate of approximately 60 beats per minute. In embodiments, the analysis unit 149 of the health monitor 101 (see figure 23) can be used to perform the Fourier transforms on the first signal 2902 and the second signal 2904 and thereby produce the frequency-domain data 3002, 3004 shown in figure 30. The analysis unit 149 may be used to derive arrhythmias from the raw bioimpedance signal on the human neck, to detect abnormal or meaningful changes in respiration, heart rate or heart rate variability, and / or to estimate stress using respiration, heart rate, and heart rate variability. In examples, the second signal 2904 can be used to determine heart rate variability for the user. Heart rate variability provides information about fluctuations in the time between heartbeats. For example, panel C of figure 29 shows multiple heartbeat cycles (e.g. 2905a-2905d), there being variations in the length of each cycle. In a given time period, the heart rate variability is the difference between the longest cycle and the shortest cycle. From the measured heart rate variability, it is possible to estimate stress of the user. A low heart rate variability (e.g. heart rate variability of less than 10 ms) may indicate that the user is in a stressed state. A high heart rate variability (e.g. heart rate variability of at least 10 ms) may indicate that the user is not in a stressed state. Furthermore, heart rate variability can be used to detect arrythmias. An arrythmia is an irregular heart rhythm. The heart may beat too quickly, too slowly, or with irregular rhythm. An arrythmia can be inferred from an irregularly large variation in beat-to-beat interval from an extracted heart rate bioimpedance signal. In an example, we consider the first four cycles 2905a-2905d shown in panel C of figure 29. Based on peak-peak measurements, the length of a first example cycle 2905a, which is the shortest cycle in this period, is 554 ms and the length of a fourth example cycle 2905d, which is the longest cycle in this period, is 562 ms. The heart rate variability in the time period covering the example cycles 2905a-2905d is 562554 = 8 ms. The processing unit 104 of the health monitor 101, particularly the analysis unit 149, may be configured to determine heart rate variability from the second signal 2904. The processing unit 104 / analysis unit 149 may classify a user as being in a stressed state when their heart rate variability, measured over a predetermined time period (e.g. the last 1 minute) is lower than a threshold (e.g. 10 ms). When the processing unit 104 / analysis unit 149 classifies a user as being in a stressed state, an alarm or notification may be provided. The processing unit 104 of the health monitor 101, particularly the analysis unit 149, may be configured to determine recovery time from e.g. the first signal 2902 and / or the second signal 2904. By recovery time it is meant the time taken for a measured health and / or vital sign to return to a normal level. For example, where a user has a heart rate which is normally within the range of 40-60 beats per minute, the recovery time corresponds to the amount of time it takes for the measured heart rate of the user to change from a peak rate (e.g. 120 beats per minute) to being within the normal range (i.e. 60 beats per minute or below). Figures 31A to 31C provide views of another example health monitor 201 according to an aspect of the invention. The health monitor 201 is generally similar to the previous health monitor 101, and can be used in a similar way, with the same numerals denoting similar features. The present example health monitor 201 differs from the previous health monitor 101 in the manner in which the plurality of contact arrangements 131a-131n are attached to the attachment arrangement 102. In the example shown in figures 31A to 31C, the contact arrangements 131 are provided in a single group of eight contact arrangements 131 attached to a single retaining member 222. The contact arrangements 131, which are individually selectable, are provided in two rows of four. This arrangement allows measurements to be taken along the axis of the neck, in a situation analogous to that shown in figure 27. The retaining member 222 can be moved along the strap 121 of the attachment arrangement 102 (the strap 121 passes through appropriately sized holes in each retaining member 122). Figure 32 provides a further illustrative example of producing processed output data from raw sensor data. Panel A of figure 32 shows example raw impedance sensor data 3200 received from the sensing arrangement 103 / processing unit 104 of the health monitor 201. In this example, the raw impedance sensor data 3200 corresponds to the filtered and demodulated output of the synchronised demodulator 146 shown in figure 28. Figure 32, panel B, shows a first signal 3202 that has been isolated from the raw sensor data 3200 by the first bandpass filter 142. In this example, the first bandpass filter 142 has a lower cutoff frequency of 0.167 Hz and a higher cutoff frequency of 0.5 Hz. The resultant signal, i.e. the first output signal 3202, has an amplitude of 30 Q pk-pk and a frequency component that is the same as the frequency of the respiration of the user. Figure 33, panel A, shows frequency-domain data 3302 produced from the first signal 3202 output from the first bandpass filter 142. This Fourier transformed data 3302 shown in figure 33, panel A, includes a peak at 0.2997 Hz, corresponding to a breathing rate of approximately 17 breaths per minute. Returning to figure 32, panel C shows a second, higher frequency, signal 3204 that has been isolated from the raw sensor data 3200 by the second bandpass filter 143. In this example, the second bandpass filter 143 has a lower cutoff frequency of 0.7 Hz and a higher cutoff frequency of 3 Hz. The resultant signal, i.e. the second output signal 3204, has an amplitude of 10 Q pk-pk and a frequency component that is the same as the frequency of the heart rate of the user. Figure 33, panel B, shows frequency-domain data 3304 produced from the second signal 3204 output from the second bandpass filter 143. This Fourier transformed data 3304 shown in figure 33, panel B, includes a peak at 1.34865 Hz, corresponding to a heart rate of approximately 80 beats per minute. In embodiments, the analysis unit 149 of the health monitor 201 can be used to perform a Fourier transform on the first signal 3202 and the second signal 3204 and thereby produce the frequency-domain data 3302, 3304 shown in figure 33. The analysis unit 149 may be used to derive arrhythmias from the raw bioimpedance signal on the human neck, to detect abnormal or meaningful changes in respiration, heart rate or heart rate variability, and / or to estimate stress using respiration, heart rate, and / or heart rate variability. As will be understood by the skilled person, the example embodiments presented above can be modified in a number of ways without departing from the scope of the invention. For example, any of the health monitors 1,101,201 and sensing arrangements 900, 1000, 1100, 1200, 1300, 1400, 1500 may further comprise an optical sensing arrangement for measuring the optical transmittivity and / or reflectivity of the user’s skin and the tissue beneath the user’s skin, including the blood flowing in the vessels beneath the skin. The optical sensing arrangement may comprise one or more optical emitters for applying an optical signal to the skin of the user and to the tissue beneath the skin of a user, and one or more optical detectors for measuring an optical response. Data measured by the optical sensing arrangement may be combined with other data measured by the health monitor device. Any of the sensing arrangements 900, 1000, 1100, 1200, 1300, 1400, 1500, 2000, 2100 may further comprise one or more of an accelerometer and a temperature sensor. For example, the accelerometer may be a multi-axis accelerometer, to indicate the level of exercise or movement of the user or subject, and the temperature sensor may be used to measure the skin temperature of the user or subject. Data measured by the accelerometer and / or temperature sensor may be combined with other data measured by the health monitor device. Advantageously, further sensors can be used to improve the accuracy of the measurement of the conditions of a user. Either or both of the health monitors 1,1b shown in figure 16 can be replaced with another example health monitor 101,201. Any of the health monitors 1,101,201 can be used to implement the method 1700. The filtering arrangements 41,141 may be embodied in hardware and / or software. A further modification of this invention could be to provide an attachment arrangement 2 in the form of a patch comprising a flexible substrate, such as a cloth or flexible plastic substrate, coated with a suitable adhesive, such as a hypoallergenic adhesive. In such examples, the electrodes may be wet or gel type electrodes for patch applications. Alternatively, the electrodes used in such patch applications may be dry electrodes, such as the example dry electrodes described above. The electrodes 31 can be printed, coated or otherwise attached to the flexible substrate, along with the battery and electronic components. This flexible substrate / adhesive could be applied to the skin of the user for long-term or shortterm monitoring of the user. Advantageously, automatic determination and / or selection of the best electrode set from an array of electrodes greater than four could remove the need for removal and repositioning of the device 1 to produce an optimal signal(s), this being especially useful when the health monitor device 1 is applied to a user in a fixed position using a patch comprising a flexible substrate and an adhesive. In the methods outlined above, receiving raw sensor data from the sensing arrangement may further comprise: receiving raw sensor data from an optical sensing arrangement and / or receiving raw sensor data from one or more of an accelerometer and a temperature sensor. In addition, further interpretation may be applied to one or more signals to determine further aspects of the subject’s condition, for example stress or physical exercise. The features disclosed in the foregoing description or the following drawings, expressed in their specific forms or in terms of an arrangement for performing a disclosed function, or a method or a process of attaining the disclosed result, as appropriate, may separately, or in any combination of such features be utilised for realising the invention in diverse forms thereof.

Claims

1. A health monitor device for monitoring the health and / or one or more vital signs of a user, the health monitor device comprising:an attachment means for attaching the health monitor device to a limb or the neck of the user;a sensing means configured to output raw sensor data, the sensing means comprising at least a plurality of electrodes for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user;a processing means configured to produce processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the respiration of the user; anda communication means for sending or reporting the raw sensor data and / or processed output data to one or more further devices.

2. A health monitor device according to claim 1, wherein the processed output data comprises at least a bioimpedance signal.

3. A health monitor device according to claim 1 or claim 2, wherein the processing means comprises a filtering means configured to obtain the first output signal by filtering the raw sensor data.

4. A health monitor device according to claim 3, wherein the filtering means comprises a first bandpass filter having a lower cutoff frequency of 0.01-0.4 Hz, for example 0.1 Hz, and a higher cutoff frequency of 0.5-2 Hz, for example 0.6 Hz.

5. A health monitor device according to any preceding claim, wherein the processing means is further configured to obtain a second output signal indicative of the cardiac activity of the user, and wherein the filtering means is configured to obtain the second output signal by filtering the raw sensor data.

6. A health monitor device according to claim 5, wherein the filtering means comprises a second bandpass filter having a lower cutoff frequency of 0.5-2 Hz, for example 0.8 Hz, and a higher cutoff frequency of 5-20 Hz, for example 8 Hz.

7. A health monitor device according to any preceding claim wherein the processing means is further configured to obtain a further output signal indicative of the cardiac activity of the user, for example an ECG signal, and wherein the filtering means is configured to obtain the further output signal by filtering the raw sensor data.

8. A health monitor device according to claim 7, wherein the filtering means comprises a low pass filter having a cutoff frequency of less than 1000 Hz, for example 150 Hz.

9. A health monitor device according to any preceding claim, wherein the plurality of electrodes comprises at least four electrodes, for example wherein the plurality of electrodes comprises an array of more than four electrodes.

10. A health monitor device according to any preceding claim, wherein the health monitor device further comprises an electrode selection means, wherein the electrode selection means is configured to determine and / or select an optimal plurality of electrodes for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user.

11. A health monitor device according to claim 10, wherein the electrode selection means is configured to analyse the raw sensor data and / or processed output data for a first plurality and a second plurality of electrodes, and to determine the optimal plurality of electrodes based on the analysis of the raw sensor data and / or processed output data.

12. A health monitor device according to any preceding claim, wherein the sensing means further comprises optical sensing means for measuring theoptical transmittivity or reflectivity of the skin of a user and optionally one or more of an accelerometer and a temperature sensor.

13. A health monitor device according to claim 12, wherein the optical sensing means comprises one or more optical emitters for applying an optical signal to the skin of the user and one or more optical detectors for measuring an optical response.

14. A health monitor device according to any preceding claim, wherein the attachment means comprises a strap, and wherein the strap is a rigid or flexible band or collar.

15. A health monitor device according to claim 14, wherein the plurality of electrodes are attached to the strap.

16. A health monitor device according to any preceding claim, wherein the attachment means comprises a flexible substrate and an adhesive layer, and wherein the plurality of electrodes are printed, coated or otherwise attached to the flexible substrate.

17. A health monitor device according to any preceding claim, wherein one or more of the electrodes are provided as part of a contact arrangement.

18. A health monitor device according to claim 17, wherein the or each contact arrangement comprises one or more resilient electrodes configured for extending through the hair of a user.

19. A health monitor device according to claim 17 or claim 18, wherein the or each resilient electrode comprises an elongate resilient member.

20. A health monitor device according to claim 19, wherein the or each elongate resilient member is impregnated with electrically conductive material.21 .A health monitor device according any one of claims 18 to 20, wherein the or each resilient electrode comprises an electrically-conductive coating.

22. A health monitor device according to any preceding claim, wherein the communication means is a wired or wireless communication means, and wherein the one or more further devices comprise at least a mobile communications device or a server.

23. A health monitor device for monitoring the health and / or one or more vital signs of a user, the health monitor device comprising:an attachment means for attaching the health monitor device to a limb or the neck of the user;a sensing means configured to output raw sensor data; anda processing means configured to produce processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the health and / or one or more vital signs of the user.

24. A system for monitoring the health and / or one or more vital signs of a user, the system comprising:a health monitor device according to any one of claims 1 to 23; and one or more further devices in wireless or wired communication with the health monitor device, wherein the one or more further devices are configured to receive sensor data and / or processed output data from the health monitor device.

25. A system according to claim 24, wherein the health monitor device and / or one or more further devices is configured to provide an alarm or notification.

26. A method for processing data, the method comprising:receiving raw sensor data from a sensing means, the sensing means comprising at least a plurality of electrodes for applying an alternating current to the skin of the user and measuring a voltage across the skin of the user; andproducing processed output data from the raw sensor data, wherein the processed output data comprises at least a first output signal indicative of the respiration of the user.

27. A method according to claim 26, wherein the processed output data comprises at least a bioimpedance signal.

28. A method according to claim 26 or claim 27, wherein producing processed output data from the raw sensor data comprises:filtering the raw sensor data to obtain the first output signal, wherein the first output signal has a frequency of e.g. between 0.1 Hz and 0.6 Hz.

29. A method according to claim 28, wherein producing processed output data from the raw sensor data comprises:filtering the raw sensor data to obtain a second output signal indicative of the heart rate and / or heart rate variability of a user, wherein the second output signal has a frequency of e.g. between 0.8 Hz and 8 Hz.

30. A method according to any one of claims 26 to 29, wherein the sensing means comprises an array of electrodes, and wherein the method further comprises:selecting a first plurality of electrodes from the array of electrodes;applying an alternating current and measuring a voltage using the first plurality of electrodes;analysing the raw sensor data output from the first plurality of electrodes; andselecting a second plurality of electrodes from the array of electrodes;applying an alternating current and measuring a voltage using the second plurality of electrodes;analysing the raw sensor data output from the second plurality of electrodes.31 .A method according to claim 30, wherein the method further comprises:selecting an optimal plurality of electrodes based on the analysis of the raw sensor data output from the first plurality of electrodes and the second plurality of electrodes.5      32. A method according to any one of claims 26 to 31, wherein the sensingmeans further comprises optical sensing means for measuring the optical transmittivity or reflectivity of the body tissue and / or flowing blood of a user, wherein receiving raw sensor data from the sensing means further comprises: receiving raw sensor data from the optical sensing means.1033. A method according to any one of claims 26 to 32, wherein the method further comprises sending or reporting the raw sensor data and / or processed output data to one or more further devices.15

Images