GB2600126A

GB2600126A - Improvements in or relating to wearable sensor apparatus

Info

Publication number: GB2600126A
Application number: GB2016726.8A
Authority: GB
Inventors: Zhou Keming; Daly Ian; Soni Aakash; Zhang Hongtao
Original assignee: August International Ltd; August Int Ltd
Current assignee: August International Ltd; August Int Ltd
Priority date: 2020-10-21
Filing date: 2020-10-21
Publication date: 2022-04-27
Also published as: GB202016726D0

Abstract

Wearable e.g. wrist-worn apparatus 2 such as a smart-watch detects physiological properties of a user and comprises e.g. a photoplethysmogram (PPG) sensor for detecting blood volume changes and providing blood volume data. This is analysed to extract data relating to: heart rate (HR); inter-beat interval(s) (IBI); pulse rate variability (PRV); peak amplitude; peak-to-peak or peak amplitude change over time, or morphology; power spectral density (PSD); low, high or very low-frequency (VLF) components of the PRV power spectrum; or energy ratios between defined frequency bands. The device may further comprise an electrocardiogram (ECG) sensor and filtering to remove motion components from the data. The apparatus may assess physiological signs of emotion (figure 5) and classify data according to arousal and/or valance levels.

Description

Improvements in or Relating to Wearable Sensor Apparatus The present invention relates to a wearable apparatus for detecting physiological properties of a user and a method for detecting physiological properties of a user of a wearable apparatus. In particular, the apparatus is a wrist-worn apparatus. The present invention also relates to a wearable apparatus for detecting physiological properties of a user and assessing physiological signs of emotion and/or detecting physiological signs of depression, and a method for detecting physiological properties of a user of a wearable apparatus and assessing physiological signs of emotion and/or detecting physiological signs of depression.

Wearable sensors are known in the art. More recently, there has been a trend towards smartwatches including a sensor for detecting physiological signals of a user so as to determine basic information on a heart-beat or blood oxygen level of the wearer of the smartwatch, typically using an optical sensor. However, the analysis of the received data is very basic and very limited, and is not used to its full potential.

Those skilled in the art will know that, to detect an electrocardiogram (ECG) signal, one must complete an electrical circuit, which does make such detection cumbersome or difficult in small wearable apparatus, especially in smart watches or the like. If an ECG signal is to be detected, at least a pair of electrodes are required and they must contact different parts of the skin of a user to make the circuit. This poses great difficulties for small wearable apparatus, where part of the attraction is its size, ease of use, and overall functionality, as the detection of ECG signals is not, typically, its primary purpose.

ECG sensors are considered by those skilled in the art as the standard for measuring heart rate owing to accuracy, as this is a measure of the electrical activity of the heart, whereas PPG sensors measure a change in blood flow during heart activity. Owing to the typical placement of a PPG sensor on users wrist, PPG data/signal can be unstable or noisy because of movement of the wrist. Further, external light artefacts may affect the quality of the data. Moreover, subject dependent physiological characteristics may also affect the data, such as the type of skin, skin absorption properties, skin structure, blood oxygen saturation, blood flow rate, skin temperatures and the measuring environment. Therefore, using just PPG data can lead to inaccurate measurements for heart rate, or make measuring heart rate unreliable or non-uniform owing to changing characteristics.

In a related field, it is observed that human emotions are psychophysiological experiences that affect all parts of everyday lives and emotions typically comprise a number of components, such as, feelings, bodily changes, cognitive reactions, behaviours, and thoughts. Although smartwatches are known in the art, no-one has previously suggested a link between such wearable sensors and any assessment of emotional state or state of depression.

The present invention is aimed at providing an improved wearable sensor apparatus, and associated method. The present invention is also aimed at providing a wearable sensor apparatus which can be used to analyse physiological signs of emotion and/or depression.

According to a first aspect, the present invention provides a wearable apparatus for detecting physiological properties of a user, the apparatus comprises: an optical sensor or photoplethysmogram (PPG) sensor contactable with the skin of said user, or arranged to illuminate the skin of said user, for detecting blood volume changes and providing blood volume data; and means for analysing the blood volume data and extracting therefrom data relating to any one or more of the group comprising: a. heart rate (HR); b. inter-beat interval(s) (IBI); c. pulse rate variability (PRV); d. peak amplitude; e. peak-to-peak change over time; f. change in the peak amplitude over time; g. morphology of the PPG peak; h. power spectral density (PSD); very low-frequency (VLF) components of the PRV power spectrum; j. low-frequency (LF) components of the PRV power spectrum; k. high-frequency (HF) components of the PRV power spectrum; and/or I. energy ratios between defined frequency bands.

Preferably, the means for analysing is configured to conduct low-pass and/or high-pass filtering, so as to remove motion components from the data.

Preferably, the apparatus further comprises: an electrocardiogram (ECG) sensor contactable with the skin of said user, for detecting electrical signals of a heart of said user and providing electrical data; means for analysing the blood volume data and electrical data; and means for calibrating the blood volume data using the electrical data. Preferably, the ECG sensor comprises first, second and third electrodes, each configured to be contactable with different parts of the skin of said user when being Most preferably, when the apparatus is wrist-worn, the ECG sensor comprises three electrodes, one negative, one positive and one earth. The positive and earth electrodes are configured to be contactable with the skin of said user when worn, by being located at or on the back of the wrist-worn apparatus, and the negative electrode is configured so as to be intermittently contactable with the skin of said user, by being located on the front or side of the wrist-worn apparatus or around the dial, where it can be touched when required.

Most preferably, the negative electrode is configured to be contactable by a digit of said user.

Most preferably, the PPG sensor is located so as to illuminate a wrist region of said user when worn, by being located at or around the back of the wrist-worn apparatus.

Preferably, the means for analysing the blood volume data and electrical data is configured to conduct: time alignment analysis of the blood volume data and electrical data; and/or: peak detection analysis of the blood volume data and electrical data. Preferably, the means for calibrating the blood volume data is configured to determine a correction coefficient from peak detection analysis and, using the correction coefficient, calculate heart rate obtained from blood volume data.

Most preferably, intermittently calibrating the blood volume data using the electrical data.

Preferably, the wearable apparatus is a smartwatch. Alternatively, the wearable apparatus is a smart wristband, chest strap or armband.

The present invention may also relate to a wearable apparatus for detecting physiological properties of a user substantially as herein disclosed, with reference to any one of Figures 1 to 5 of the accompanying drawings and/or any example described herein.

According to a second aspect, the present invention provides a method for detecting physiological properties of a user of a wearable apparatus, the method comprising: detecting blood volume data from an optical sensor or photoplethysmogram (PPG) sensor in contact with the skin of the user, or illuminating the skin of said user; and analysing the blood volume data and extracting therefrom data relating to any one or more of the group comprising: a. heart rate (HR); b. inter-beat interval(s) (1131); c. pulse rate variability (PRV); d. peak amplitude; e. peak-to-peak change over time; f. change in the peak amplitude over time; g. morphology of the PPG peak; h. power spectral density (PSD); very low-frequency (VLF) components of the PRV power spectrum; j. low-frequency (LF) components of the PRV power spectrum; k. high-frequency (HF) components of the PRV power spectrum; and/or energy ratios between defined frequency bands.

Preferably, analysing further comprises low-pass and/or high-pass filtering of the data.

Preferably, selecting one or more of the group comprising a. to I. and classifying according to arousal and/or valance levels.

Preferably, the method further comprises: detecting electrical signals of a heart of the user from an electrocardiogram (ECG) sensor to provide electrical data; analysing the blood volume data and electrical data; and calibrating the blood volume data using the electrical data.

Preferably, analysing comprises: time alignment analysis of the blood volume data and electrical data; and/or: peak detection analysis of the blood volume data and electrical data.

Preferably, calibrating comprises determining a correction coefficient from peak detection analysis and, using the correction coefficient, calculating heart rate obtained from blood volume data.

Preferably, intermittently calibrating the blood volume data using the electrical data.

The present invention may also relate to a method for detecting physiological properties of a user of a wearable apparatus substantially as herein disclosed, with reference to any one of Figures 3 to 5 of the accompanying drawings and corresponding description, or any example described herein.

According to a further aspect, the present invention provides a wearable apparatus for detecting physiological properties of a user of the wearable apparatus and assessing physiological signs of emotion and/or detecting physiological signs of depression of said user, the apparatus comprises a wearable apparatus according to the first aspect, and further comprises means for emotion recognition configured to conduct analysis of the blood volume data, any one or more of the extracted blood volume data a. to I., and/or electrical data utilising a discrete emotional model or affective dimensional model.

Preferably, the means for emotion recognition utilises a circumflex model of affect / emotion.

According to a further aspect, the present invention provides a method for detecting physiological properties of a user of a wearable apparatus and assessing physiological signs of emotion and/or detecting physiological signs of depression of the user, comprising a method according to the second aspect, and further comprising emotion recognition analysis of the blood volume data, any one or more of the extracted blood volume data a. to I., and/or electrical data utilising a discrete emotional model or affective dimensional model.

Preferably, the method utilises a circumflex model of affect / emotion.

Advantageously, a PPG sensor is easily incorporated into a wearable apparatus from where it can be used to continuously collect PPG heart rate data. An ECG sensor, for which a user must make a circuit, can be used to intermittently collect ECG heart rate data and use that ECG data so as to calibrate the PPG data. This provides the best of both PPG and ECG. The former is more easily collected and potentially collected continuously, and the latter can be used to calibrate the former so as to ensure that the PPG determined heart rate is more accurate.

Advantageously, ECG data is used to calibrate PPG data so as to enhance the accuracy of heart rate derived from PPG data.

Advantageously, it has been realised that emotion detection or recognition can be based on physiological signals alone and that, through appropriate analysis of the data available, the physiological signals can be linked to an emotional state of a user.

Advantageously, it has been realised that detection of depression can be based on physiological signals alone and that, through appropriate analysis of the data available, the physiological signals can be linked to a state of depression.

As such, emotion assessment and depression detection can be made using a smart wearable apparatus, especially a smartwatch.

To analyse emotional state or state of depression, accurate features are required for data analysis and the ECG calibration process improves the accuracy of 25 the PPG data.

In addition to determining the emotional state, or depression state, of a user, the wearable apparatus can detect sleep disturbance, blood pressure, tiredness, etc. using its additional sensors.

Advantageously, the present invention provides an improved wearable sensor apparatus, especially an improved smartwatch. Further, the present invention provides a low-cost, wearable sensor which can, effectively: assess emotions or states of depression; enable a user to monitor his/her overall health; and/or using digital technology (e.g. Apps, cloud computing, etc.), enable a healthcare practitioner to monitor the overall health of the user remotely.

As the apparatus is wearable and easily transported, it can be used in a variety of scenarios including during rest and play, such as whilst listening to music or during sports, such that emotions can be assessed in situ and in real-time.

The invention will now be disclosed, by way of example only, with reference to the following drawings, in which: Figure 1 is a system diagram showing components of a monitoring system; Figure 2 is hardware block diagram of a wearable sensor apparatus; Figure 3 is flowchart identifying data acquisition, data analysis and emotion recognition; Figure 4 is a graph comparing ECG signals and PPG signals; and Figure 5 shows an example of a Circumplex Model of Affect.

Figure 1 shows a monitoring system, identified generally by reference 1. The monitoring system includes a smartwatch 2, a phone! tablet / computer 3, a network 4, data storage 5, and a server 6. The smartwatch may communicate directly with the phone! tablet! computer 3 via communication link 7, or may communicate directly with the network 4 via communication link 8. Each of the phone / tablet! computer 3, data storage 5 and server 6 may communicate directly with the network 4 via communication links 9a to 9c, respectively. Communication links 7 to 8 and 9a to 9c provide two-way communication and may be wired or wireless communication links, but are preferably the latter.

In use, the smartwatch 2 is capable of working independently to collect and store physiological data of a wearer, and further capable of transmitting the data. The smartwatch 2 may analyse the data collected, or analysis may be conducted at the server 6. When communicating directly via communications link 8, the smartwatch 2 uses the cellular! mobile network. Whereas, when communicating via communications link 7, the smartwatch 2 uses short-range wireless technology, such as Bluetoothml, and the phone! tablet / computer 3 uses the cellular! mobile network or wi-fi for communications link 9a to provide such data to the network 4.

Once at the network 4, the data is sent to the data storage 5 and/or the server 6. The data storage 5 stores all data collected by the smartwatch 2, and supplies such data to the server 6 as and when required. The server 6 uses the data from the data storage 5 and analyses or further analyses the data of the wearer.

Referring now to Figure 2, the smartwatch 2 from Figure 1 includes a microcontroller unit (MCU) 20, with an associated battery 21, storage module 22 and flash memory 23. The battery 21 supplies power to the whole smartwatch 2. The MCU 20 may be responsible for controlling operation of the smartwatch to collect data and then transmit it for analysis at the server 6, or the MCU may be responsible for conducting initial filtering and/or analysis of the collected data before forwarding to the server 6, where additional analysis, etc. may be conducted. The storage module 22 is responsible for storing user data such as data supplied by one or more sensors and/or other inputs. The flash memory 23 provides means of expanding the system / firmware of the smartwatch 2. Figure 2 only shows the main components of the smartwatch 2.

In the embodiment shown, the smartwatch 2 includes an antenna 24, a GPS module 25 and BluetoothTM module 26, although it should be noted that the smartwatch 2 may have just one, two or all three of those means of communication.

The antenna 24 is responsible for cellular communications through mobile telephone networks such as GSM, 3G, 4G, 5G, NB-IoT (Narrow Band -Internet of Things), etc. The BluetoothTM module is responsible for short-range wireless communications. The GPS module 25 is used to track a location of the user but could, of course, be based upon other satellite location services. Further, the smartwatch 2 includes a display 27, speaker 28 and motor 29 for vibrational output.

A number of sensors or inputs to the MCU 20 are provided. A first being a PPG (photoplethysmogram) & ECG (electrocardiogram) sensor 30, although it should be noted that some embodiments of the invention only require a PPG sensor 30a, whereas other embodiments require both a PPG 30a and ECG sensor 30b.

The PPG sensor 30a and ECG sensor 30b are responsible for collecting blood volume data and electrical data, respectively.

Additional sensors are provided by an inertial measurement unit (IMU) 31, a temperature sensor 32, and a microphone 33, and input to the MCU 20 is provided by way of one or more buttons 34 and/or a touchscreen. The IMU 31 is responsible for collecting the wearer's activity data such as step count, sleep monitoring and/or posture change. The temperature sensor 32 is responsible for collecting data relating to the wearer's body temperature. The microphone 33 is responsible for collecting data relating to voice commands or other audible signals.

Input to the smartwatch from the wearer is intended to be through the one or more buttons 34, the display 27 acting as a touchscreen, and/or through the microphone 33.

Output from the smartwatch to the wearer is intended to be though the display, 27, speaker 28 and/or the motor 29. The display 27 is an LCD (liquid crystal display) / OLED (organic light emitting diode) screen, or LED (light emitting diode) lights.

In a preferred embodiment, the ECG sensor 30b of the smartwatch 2 has three electrodes, one negative, one positive and one earth. The positive and earth electrodes are located at or on the back of the smartwatch 2, and the negative electrode is located on the front or side of the smartwatch 2 or around the dial, where it can be touched when required. The PPG sensor 30a is located so as to illuminate a wrist region of said user when worn, by being located at or around the back of the smartwatch 2.

Although a smartwatch is disclosed, the wearable apparatus could be a smart wristband, chest strap or armband.

In use, blood volume data is collected by the PPG sensor 30a and, optionally, data is further collected from the IMU sensor 31, temperature sensor 32 and microphone 33. Such collected data is either analysed by the MCU 20 before being transmitted to the server 6 where it may be further analysed, or transmitted and analysed at the server 6. Data relating to inputs from the one or more buttons 34 and touchscreen display 27 may be treated in the same way, and used during analysis. Although various sources of data may be utilised, the main analysis is conducted on the blood volume data collected from the PPG sensor 30a, and that is further elaborated upon in relation to Figures 3 and 4. As a further option, electrical data is collected from the ECG sensor 30b so that such data can be used to calibrate the PPG data.

Figure 3 shows a flowchart identifying: data acquisition, reference 30; data analysis, references 32 to 35; and emotion recognition, reference 36. Signal calibration, reference 31, is typically conducted intermittently and requires ECG data.

During data acquisition 30, PPG data is collected from the PPG Sensor 30a continuously whilst the smartwatch is being worn by the user and this function is active. However, ECG data may also be collected from the ECG sensor 30b intermittently, but the latter is collected only so as to calibrate the PPG data.

During signal calibration 31, which is only conducted intermittently, calibration includes two sub-processes being time alignment and peak detection. Figure 4 is a graph showing the correlation between ECG signals and PPG signals.

Conducting time alignment involves the following: recording the PPG signal and ECG signal for over 30 seconds; down sampling the PPG signal to match a sampling rate of the ECG signal, or vice versa; using cross-correlation to find the time difference in the two signals; if the time difference is higher than zero, then cropping the first time difference seconds from the PPG signal; and if the time difference is lower than zero, then cropping the first time difference seconds from ECG signal.

Accordingly, conducting time alignment provides time aligned signals. Conducting peak detection involves the following: calculating peaks of both the ECG and PPG signals; obtaining a correction coefficient (CC) by calculating the total distance between each peak of the ECG and the PPG data and dividing that by number of peaks of the PPG signal, according to cc - t=ltdt Npeaks (where, CC is the correction coefficient, N is the number of IBI values used, td is time difference between ECG and PPG peaks, and Npeaks is the number of peaks).

calculating the heart rate by deriving an average inter-beat interval (IBI) from the PPG signal and using the correction coefficient above according to HR = E7_11Bli

CC

(where, HR is heart rate, CC is the correction coefficient, N is the number of IBI values used).

For example, if 30 seconds of data is used for calibration and a user's HR is 74 beats per minute (BPM), then 30s of data will have a HR of 37 BPM and this results in 36 IBI values and N equals 36.

Having determined CC, this is then used to calculate all future values of HR obtained by the PPG sensor 30a until the next calibration is conducted.

During Signal pre-processing 32, the acquired PPG signal is processed by low-pass and high-pass filtering to remove noise from the signal and, together with independent component analysis, the aim is to remove from the signal any motion artefact, or portion of the signal attributable to movement of the wearer and, thereby, the smartwatch 2.

During feature extraction 33, various features are extracted or isolated from the filtered signal, and the extracted features relate to, but are not limited to one or more of the following: head rate (HR) -heart-beats per minute; inter-beat intervals (IB I) -being a distance between each peak or heart-beat; pulse rate variability (PRV) -being a measure of the variation in time between each heart-beat; peak amplitude -being a difference between the mid-point of the signal and the positive peak; peak-to-peak change over time -being a measure of the second-order derivative of the peak-to-peak intervals; change in peak amplitude over time -the amplitude of the signal changes over time and variations in that indicate different emotions; morphology of the PPG peak -being analysis of the shape of the PPG signal e.g. the gradient of the pre-peak and post-peak slopes; considering spectral power in the bands about 0.1 to about 0.2 Hz, about 0.2 to about 0.3 Hz, and about 0.3 to about 0.4 Hz to define a power spectral density (PSD) of the PPG signal at given bands -which describes the power present in the signal at different frequencies; very low-frequency (VLF), being about 0.003 to about 0.04 Hz components of the PRV power spectrum; low-frequency (LF), being about 0.04 to about 0.15 Hz components of PRV power spectrum; high-frequency (HF), being about 0.15 to about 0.4 Hz components of PRV power spectrum; and/or the energy ratio between the frequency bands LF (about 0.04 to about 0.15 Hz) and HF (about 0.15 to about 0.4) Hz.

During feature selection 34, any one or more of the group of those extracted features listed above are selected for further processing.

During classification of emotions 35, a classifier is used to classify arousal and valance levels to recognise the user's emotions.

The classifier could be a fuzzy logic model or a machine learning model, such as a neural networks, decision trees, or support vector machines, etc. For example, after extracting these features, a neural network (or system) can be trained to learn these features. Once, learnt, the network (or system) can apply such learning to real-time PPG data and classify real-time emotions. A skilled person would understand that a neural network or other machine learning model is trained with labelled data collected previously.

Two approaches are widely used for bio-signal-based emotion detection, and those are a discrete emotional model (DEM) and an affective dimensional model (ADM). A circumplex model of affect is part of the ADM approach, and this invention focuses on ADM where conditions for discrete emotions are relaxed and affective states are treated as a combination of two parameters: arousal; and valance. Arousal is a measure of emotional stimulation (intensity) and it varies between low and high, whereas valance is a measure of pleasantness for the experienced emotion, ranging from very pleasant (positive) to very unpleasant (negative). The role of emotion recognition 36 is to link the detected physiological properties of the wearer to potential signs of emotion and/or detect potential signs of depression. During emotion recognition 36, a user's emotions are identified based upon the physiological signals processed and analysed, and such recognition is based on a Circumplex Model of Affect as shown in Figure 5.

In use, PPG data is acquired over a given time. This data is pre-processed in order to remove noise. Then, feature extraction takes place in order to extract a feature set desired for further evaluation. The feature set will have properties relevant to a type or types of emotion being analysed, and the values of those features vary between different emotions reflecting high and low levels of arousal and valance. During feature selection, the feature values are grouped and prepared for data analysis. Afterwards, the feature values are passed through a trained network or classifier in order to identify a user's emotion. A classifier will provide levels of arousal and valance. During the emotion recognition stage, a user's emotion is mapped according to the Circumplex Model of Affect.

Claims

Claims: 1.) A wearable apparatus for detecting physiological properties of a user, the apparatus comprises: an optical sensor or photoplethysmogram (PPG) sensor contactable with the skin of said user, or arranged to illuminate the skin of said user, for detecting blood volume changes and providing blood volume data; and means for analysing the blood volume data and extracting therefrom data relating to any one or more of the group comprising: a. heart rate (HR); b. inter-beat interval(s) (IBI); c. pulse rate variability (PRV); d. peak amplitude; e. peak-to-peak change over time; f. change in the peak amplitude over time; g. morphology of the PPG peak; h. power spectral density (PSD); very low-frequency (VLF) components of the PRV power spectrum; j. low-frequency (LF) components of the PRV power spectrum; k high-frequency (HF) components of the PRV power spectrum; and/or I. energy ratios between defined frequency bands.
2.) An apparatus as claimed in claim 1, wherein the means for analysing is configured to conduct low-pass and/or high-pass filtering, so as to remove motion components from the data.
An apparatus as claimed in claim 1 or claim 2, further comprising: an electrocardiogram (ECG) sensor contactable with the skin of said user, for detecting electrical signals of a heart of said user and providing electrical data; means for analysing the blood volume data and electrical data; and means for calibrating the blood volume data using the electrical data.
4.) An apparatus as claimed in claim 3, wherein the ECG sensor comprises first, second and third electrodes, each configured to be contactable with different parts of the skin of said user when being worn 5.) An apparatus as claimed in claim 4, wherein when the apparatus is wrist-worn, the ECG sensor comprises three electrodes, one negative, one positive and one earth, the positive and earth electrodes are configured to be contactable with the skin of said user when worn and the negative electrode is configured so as to be intermittently contactable with the skin of said user when required.6.) An apparatus as claimed in any one of claims 3 to 5, wherein the means for analysing the blood volume data and electrical data is configured to conduct: time alignment analysis of the blood volume data and electrical data; and/or: peak detection analysis of the blood volume data and electrical data.7.) An apparatus as claimed in claim 6, wherein the means for calibrating the blood volume data is configured to determine a correction coefficient from peak detection analysis and, using the correction coefficient, calculate heart rate obtained 20 from blood volume data.8.) An apparatus as claimed in any one of claims 3 to 7 comprising intermittently calibrating the blood volume data using the electrical data.9.) An apparatus as claimed in any preceding claim, wherein the wearable apparatus is a smartwatch.10.) A method for detecting physiological properties of a user of a wearable apparatus, the method comprising: detecting blood volume data from an optical sensor or photoplethysmogram (PPG) sensor in contact with the skin of the user, or illuminating the skin of said user; and analysing the blood volume data and extracting therefrom data relating to any one or more of the group comprising: a. head rate (HR); b. inter-beat interval(s) (IBI); c. pulse rate variability (PRV); d. peak amplitude; e. peak-to-peak change over time; f. change in the peak amplitude over time; g. morphology of the PPG peak; h. power spectral density (PSD); very low-frequency (VLF) components of the PRV power spectrum; j. low-frequency (LF) components of the PRV power spectrum; k. high-frequency (HF) components of the PRV power spectrum; and/or energy ratios between defined frequency bands.11.) A method as claimed in claim 10, wherein analysing further comprises low-pass and/or high-pass filtering of the data.12.) A method as claimed in claim 10 or claim 11 comprising selecting one or more of the group comprising a. to I. and classifying according to arousal and/or valance levels.13.) A method as claimed in any one of claims 10 to 12, wherein the method further comprises: detecting electrical signals of a heart of the user from an electrocardiogram (ECG) sensor to provide electrical data; analysing the blood volume data and electrical data; and calibrating the blood volume data using the electrical data.14.) A method as claimed in claim 13, wherein analysing comprises: time alignment analysis of the blood volume data and electrical data; and/or: peak detection analysis of the blood volume data and electrical data.15.) A method as claimed in claim 13 or claim 14, wherein calibrating comprises determining a correction coefficient from peak detection analysis and, using the correction coefficient, calculating heart rate obtained from blood volume data.16.) A method as claimed in any one of claims 13 to 15 comprising intermittently calibrating the blood volume data using the electrical data.17.) A wearable apparatus for detecting physiological properties of a user of the wearable apparatus and assessing physiological signs of emotion and/or detecting physiological signs of depression of said user, wherein the apparatus comprises a wearable apparatus according to any one of claims 1 to 9, and further comprises means for emotion recognition configured to conduct analysis of the blood volume data, any one or more of the extracted blood volume data a. to I., and/or electrical data utilising a discrete emotional model or affective dimensional model.18.) An apparatus as claimed in claim 17, wherein the means for emotion recognition utilises a circumflex model of affect / emotion.19.) A method for detecting physiological properties of a user of a wearable apparatus and assessing physiological signs of emotion and/or detecting physiological signs of depression of the user, comprising a method according to any one of claims 10 to 16, and further comprising emotion recognition analysis of the blood volume data, any one or more of the extracted blood volume data a. to I., and/or electrical data utilising a discrete emotional model or affective dimensional model.20.) A method as claimed in claim 19, wherein the method utilises a circumflex model of affect / emotion.