APPARATUS AND METHODS FOR DETECTING STROKE IN A PATIENT

Abstract

A stroke detection apparatus comprises a data processing device comprising a processor and at least one wearable sensor configured to generate movement data of at least a portion of the user's body. The data processing device is configured to process first movement data for a first movement and second movement data for a second movement received from the at least one wearable sensor. Wherein the data processing device is configured to determine asymmetry of user's movement based on the first and second movement data and generate a stroke detection signal in dependence on the determined asymmetry.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is the National Phase under 35 U.S.C. § 371 of PCT International Application No. PCT/SE2019/051319, filed on Dec. 19, 2019, which claims priority to Great Britain Patent Application No. 1820892.6, filed Dec. 20, 2018, and Swedish Patent Application No. 1930370-0, filed Nov. 12, 2019, the entire contents of each of which are incorporated herein by reference.

BACKGROUND

Technical Field

The present invention relates to apparatus and methods for detecting stroke in a patient.

Description of Related Art

A stroke is a medical condition in which poor blood flow to the brain results in cell death. There are two main causes for stroke. An ischemic stroke is typically caused by a lack of blood flow to parts of the brain resulting from a blockage in an artery that supplies blood to the brain. The blood normally delivers oxygen and nutrients to the brain. Once the oxygen and nutrients are cut off by the blockage, the brain cells cannot make enough energy and will eventually stop working. If the blockage is not cleared, the brain cells will eventually die. A haemorrhagic stroke is cause by bleeding in the brain. The bleeding is typically caused by a damaged blood vessel leaking blood. A haemorrhagic stroke may also be caused by a burst brain aneurism. In both cases, the blood spreads into the surrounding brain tissue causing increased pressure, limiting the operating of the brain cells and eventually damaging the brain tissue.

In both of the above types of stroke, the resultant effect is a change in the function of the brain, as brain cells cease to function correctly. This change can be observed through physical symptoms such as an inability to move or feel on one side of the body, problems communicating, and loss of vision. These physical symptoms often appear more or less immediately after the stroke has begun.

Historically, a stroke was very difficult to treat. Although the patient's symptoms might be recognised and diagnosed as a stroke, limited treatment was available. Where the stroke was identified as an ischemic stroke, a clot-dissolving drug such as a tissue plasminogen activator was given to the patient intravenously, in the hope that the drug would reach the clot and dissolve it sufficiently to allow blood flow to resume through the affected artery. Such a treatment would need to be successfully applied within just a few hours of the start of the stroke to ensure that the damage to the brain tissue was limited. Some studies suggest that the time window for getting the best results from clot-dissolving drug is three hours from the first signs of the stroke.

Where a stroke was not successfully treated, damage to brain tissue became inevitable and the only recourse was to provide the patient with care and rehabilitation training.

More recently, ischemic strokes have been successfully treated via an endovascular procedure called ‘mechanical thrombectomy’ in which the blood clot is removed by sending a clot retrieval device to the site of the blocked blood vessel in the brain. The device secures the clot and pulls the clot back out the blood vessel as the device is removed. Similarly, haemorrhagic strokes may also be treated via an endovascular procedure by delivering a metal clip to the damaged blood vessel or ruptured aneurysm. The clip is secured to restrict the blood flood and prevent further blood from leaking into the surrounding blood tissue. As with the clot-dissolving drug, significant damage mitigation can be achieved if the procedure is performed within a few hours of the first signs of the stroke.

Therefore, given improved treatment options for ischemic strokes, the importance of recognising and characterising the symptoms of a stroke has increased. For patients with a high risk of suffering a stroke, a method is needed of providing constant monitoring with rapid diagnosis of a stroke and escalation to a health care provider.

There are known techniques for early detection of stroke in patients employing the use of sensors worn on the patient's body. Villar. J R. “A hybrid intelligent recognition system for the early detection of strokes” describes the use of a wearable device that generates warning alarms and automatically connects to e-health services when a stroke is detected. The described approach employs two wearable devices to monitor movement data and employs genetic fuzzy finite-state machines and Time Series (TS) analysis to determine a stroke.

US patent application 2018153477 discloses a device for monitoring patients for a stroke via several sensors for determining ‘physiological signals’. The physiological signals may comprise a heart rate signal, an atrial rate signal, a heart rate variability signal, a blood pressure signal, a blood pressure variability signal, a heart sound signal, etc.

U.S. Pat. No. 7,981,058 discloses a device for monitoring patients using low cost biaxial motion sensors. The first sensor captures objective acceleration data, and the second biaxial sensor captures subjective acceleration data relative to at least the first accelerometer. Acceleration data is then used to determine nonlinear parameters and to generate at least two levels of motor function information.

US2017281054 is a US application disclosing a device for monitoring patients for a stroke via a plurality of motion sensors located at one or more anatomical locations on a patient's body and configured to detect a plurality of motion parameters corresponding to a motion of a portion of the patient's body. The motion of the portion of the patient's body is then based on a plurality of predetermined motion sentence features.

US2015157252 is a US application disclosing a device for monitoring patients via several sensors. A technique is described for switching between sensors to determine which sensor is providing the best indication of a physiological signal status of the patient.

US20160213318 is a US application disclosing a system and method for detecting a stroke in a sleeping individual. The system comprises a sensor which is worn on the hand for detecting the absence of electrical or muscular activity in the hand.

WO2018110925 is an international application which discloses an apparatus with sensors on left and right hands to measure the movement of the left side and right side for detecting a stroke during deep sleep.

However, problems associated with the above systems include problems related to producing a reliable signal indicative of a patient stroke without trigger too many false positives. What is needed is a system capable of generating a stroke detection signal with minimal false positives, and where false positives do occur, the system can gracefully handle them without too much inconvenience to the user. Furthermore, a stroke condition may occur when a patient is awake instead of when they are sleeping.

SUMMARY

Examples of the present invention aim to address the aforementioned problems.

According to an aspect of the present invention there is a stroke detection apparatus comprising: a data processing device comprising a processor; at least one wearable sensor configured to generate movement data of at least a portion of the user's body;

the data processing device configured to process first movement data for a first movement and second movement data for a second movement received from the at least one wearable sensor; wherein the data processing device is configured to determine asymmetry of user's movement based on the first and second movement data and generate a stroke detection signal in dependence on the determined asymmetry.

Optionally the data processing device is configured to determine the asymmetry of the user's movement based on the first and second movement data based on the user performing at least one predetermined body gesture.

Optionally the data processing device is configured to prompt the user to perform the at least one predetermined body gesture.

Optionally the data processing device generates a stroke detection signal when the data processing device determines that the asymmetry of the user's movement exceeds a predetermined threshold.

Optionally the data processing device is configured to determine the predetermined threshold based on the user's historical movement data for the user's body.

Optionally the predetermined threshold is based on one or more pre-sets relating to user characteristics.

Optionally an automated emergency services request is generated in dependence on the stroke escalation signal.

Optionally the at least one wearable sensor is a wearable on the arms and/or legs.

Optionally a first wearable sensor is worn on one of the user's wrists and a second wearable sensor is worn on the other of the user's wrists.

Optionally the at least one wearable sensor comprises a first sensor configured to measure movement on a first side of a plane of symmetry the user's body and a second wearable sensor configured to measure movement on a second side of the plane of symmetry of the user's body.

Optionally the plane of symmetry of the user's body is one or more of a sagittal plane, a frontal plane and/or a transverse plane.

Optionally the at least one wearable sensor is configured to measure the first movement data and the second movement data are measured at the same time.

Optionally the at least one wearable sensor is configured to measure the first movement data and the second movement data at different times.

Optionally the at least one wearable sensor is configured to transmit the generated movement data to the data processing device.

Optionally the data processing device is configured to prompt the user to perform additional predetermined body gestures when the data processing device determines asymmetry of user's movement based on the first and second movement data.

According to an aspect of the present invention there is a method of generating a stroke detection signal comprising: generating movement data of at least a portion of the user's body with at least one wearable sensor;

processing first movement data for a first movement and second movement data for a second movement received from the at least one wearable sensor; determining asymmetry of user's movement based on the first and second movement data; and generating a stroke detection signal in dependence on the determined asymmetry.

BRIEF DESCRIPTION OF THE DRAWINGS

Various other aspects and further examples are also described in the following detailed description and in the attached claims with reference to the accompanying drawings, in which:

FIG. 1 is a schematic diagram showing a patient wearing a multi-sensor stroke detection apparatus;

FIG. 2 shows a perspective view of stroke detection apparatus according to an example of the present application;

FIG. 3 is a schematic diagram of a stroke detection apparatus according to an example of the present application;

FIG. 4a and FIG. 4b are process flow diagrams showing execution flow of the wearable sensor and control device respectively;

FIG. 5a and FIG. 5b are sequence diagrams showing data-transmission order between a first and second wearable sensor device and a control device;

FIG. 6 is a process flow diagram showing execution flow of the control device 300;

FIG. 7 is an alternative process flow diagram showing execution flow of the control device 300;

FIGS. 8 shows a schematic view of the patient's body;

FIGS. 9a to 9h show a series of gestures to be performed by the patient according to an example of the present application;

FIGS. 10a to 10e show a series of gestures to be performed by the patient according to an example of the present application;

FIGS. 11a to 11e show a series of gestures to be performed by the patient according to an example of the present application;

FIGS. 12a to 12i show a series of gestures to be performed by the patient according to an example of the present application;

FIGS. 13a to 13c show a series of gestures to be performed by the patient according to an example of the present application;

FIGS. 14a to 14f show a series of gestures to be performed by the patient according to an example of the present application; and

FIGS. 15a to 15b show a series of gestures to be performed by the patient according to an example of the present application.

DETAILED DESCRIPTION

FIG. 1 is schematic diagram showing a user 100 with a stroke detection apparatus 102 comprising a plurality of wearable sensors 200a, 200b. In some examples, the user 100 is a patient 100. In some examples, the patient 100 may be at particular risk of transient ischemic attack (TIA) or other types of stroke such as ischemic stroke, haemorrhagic stroke, aneurysms, arteriovenous malformations (AVM), cryptogenic stroke, and/or brain stem stroke. Hereinafter, the term “stroke” will be used. In other examples, the user is not a patient, but uses the stroke detection apparatus 102 as a precaution. The term “user” and “patient” may be used interchangeably however for the purposes of clarity, the term “patient” will be used hereinafter.

In FIG. 1, wearable sensors 200a, 200b are respectively attached to the patient's 100 right wrist 104a and left wrist 104b. The reference number 200 generally refers to the wearable sensor, but for the purposes of clarity the reference numbers 200a, 200b refer to the wearable sensors 200a, 200b on the patient's 100 right wrist 104a and left wrist 104b.

In other examples, wearable sensors may be worn on the right and left ankles, 106a, 106b either as an alternative to the wrists or in combination with the wrists. Other positions in which the wearable sensors may be warn include footwear, headwear, as well as attached to clothing such as trousers, and upper body garments. In other examples, one or more sensors 200a, 200b can be mounted on any suitable part of the patient's body 110.

FIG. 2 shows perspective view of example of a wearable sensor 200 shown in FIG. 1. In one example, the wearable sensor 200 comprises a strap 202 configured to secure the wearable sensor 200 to the patient 100, sensor body 204 housing processing board 206. Processing board 206 may comprise power source 208, data processing device 210, and sensor package 212.

Sensor package 212 may include any suitable sensor component or plurality of sensor components configured to measure an inclination, a position, an orientation, and/or an acceleration of the part of the patient's 100 body 110 to which the wearable sensor 200 is attached. Sensor package 212 may comprise a piezoelectric, piezoresistive and/or capacitive component to convert the mechanical motion into an electrical signal. In other examples, any suitable sensor configured to detect motion of one or more portions of the patient's body is used. A piezoceramic (e.g. lead zirconate titanate) or single crystal (e.g. quartz, tourmaline) sensor may be used. In some examples, capacitive accelerometers are employed due to their superior performance in the low frequency range.

The data processing device 210 may be implemented by special-purpose software (or firmware) run on one or more general-purpose or special-purpose computing devices, such as hardware processor(s). Each “element” or “means” of such a computing device refers to a conceptual equivalent of a method step; there is not always a one-to-one correspondence between elements/means and particular pieces of hardware or software routines. One piece of hardware sometimes comprises different means/elements. For example, a processing unit serves as one element/means when executing one instruction, but serves as another element/means when executing another instruction. In addition, one element/means may be implemented by one instruction in some cases, but by a plurality of instructions in some other cases. Such a software-controlled computing device may include one or more processing units, e.g. a CPU (“Central Processing Unit”), a DSP (“Digital Signal Processor”), an ASIC (“Application-Specific Integrated Circuit”), discrete analog and/or digital components, or some other programmable logical device, such as an FPGA (“Field Programmable Gate Array”). The data processing device 210 may further include a system memory and a system bus that couples various system components including the system memory to the processing unit. The system bus may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures. The system memory may include computer storage media in the form of volatile and/or non-volatile memory such as read only memory (ROM), random access memory (RAM) and flash memory. The special-purpose software may be stored in the system memory, or on other removable/non-removable volatile/non-volatile computer storage media which is included in or accessible to the computing device, such as magnetic media, optical media, flash memory cards, digital tape, solid state RAM, solid state ROM, etc. The special-purpose software may be provided to the data processing device 210 on any suitable computer-readable medium, including a record medium and a read-only memory.

The data processing device 210 includes one or more communication interfaces, such as a serial interface, a USB interface, a wireless networking interface, etc, as well as one or more data acquisition devices, such as an analogue to digital (A/D) converter. In one example, the data processing device 210 may include a transmitter component configured to send sensor data received from the sensor package 212 and processed by the data processing device 210 and/or A/D converter (not shown) over the one or more communication interfaces. In one example, a communication interface is provided via a Bluetooth® or Wi-Fi transceiver and the processed sensor data is sent to a control device 300 (described below with reference to FIG. 3). The processed sensor data may alternatively be sent to one or more remote devices via a GSM, LTE, or any other similar licensed or unlicensed mobile communications interface.

Power source 208 may comprise a battery, kinetic energy source, or other power source suitable for a wearable device. The power source 208 is arranged to provide an energy source for powering the data processing device 210 and sensor package 212 of the processing board 206.

Wearable sensor 200 may further comprise a fastening component 214 configured to be secured with a counterpart component 216, to allow the wearable sensor 200 to be secured to a limb of the patient 100.

In one example, the fastening component 214 comprises a sensor (not shown) configured to determine whether the strap 202 of wearable sensor 200 is in an ‘open’ configuration or a ‘secured’ configuration. An example of an ‘open’ configuration of the strap 202 is shown in FIG. 2 where the wearable sensor 200 is not secured to anything. An example of the ‘closed’ configuration is shown in FIG. 1 where the wearable sensor 200 is secured to the patient 100. Similarly, an example of the ‘closed’ configuration is shown in FIG. 3, where the fastening component 214 has been fastened to counterpart component 216 to arrange the strap of wearable sensor 200 in a secured loop. In one example, the sensor of fastening component 214 is electrically connected to processing board 206 such that data processing device 210 may determine the configuration of the strap 202 of wearable sensor 200. Accordingly, the data processing device 210 can determine if the wearable sensor 200 is being worn by the patient 100 and the data processing device 210 can generate a “wearing” or “not wearing” status information for the wearable sensor 200. The data processing device 210 can use the “wearing” or “not wearing” status information for the wearable sensor 200 when determining information relating to the patient 100. For example, if the data processing device 210 determines that the patient 100 is not wearing the wearable sensor 200 from “not wearing” status information, then the data processing device 210 can determine that any alerts associated with the patient 100 may be a false alarm.

FIG. 3 shows an example wireless network 302 according to an example. In FIG. 3, two wearable sensors shown as 200a and 200b (previously discussed in reference to FIGS. 1 and 2), are worn by the patient 100 (not shown) and the respective strap 202 of each wearable sensor is in the ‘closed’ configuration. Each wearable sensors 200a, 200b collects sensor data from the patient 100, process said data, and transmit the data to the control device 300 via a wireless networking interface 304.

In some examples, the wearable sensors 200a, 200b collects sensor data from the patient 100, and transmits the data to the control device 300 without processing the sensor data via wireless networking interface 304. This is discussed in further detail in reference to FIG. 5b. In other examples, one of the wearable sensors 200a is in wireless communication with the other wearable sensor 200b. For example, the wearable sensors 200a, 200b can communicate over Bluetooth® or low energy Bluetooth®. One of the wearable sensors 200a is in wireless communication with the control device 300 via the wireless networking interface 304. The other wearable sensors 200b is not in in wireless communication with the control device 300.

In some examples, one of the wearable sensors 200a processes the collected sensor data from some or all of the wearable sensors 200a, 200b. In this way, one of the wearable sensors 200a, is a master wearable sensor 200a and the other of the wearable sensors 200b is a slave wearable sensor 200b. This is discussed in further detail in reference to FIG. 5a.

Whilst the examples as shown in the Figures only show two wearable sensors 200a, 200b, in an alternative example, additional wearable sensors can be worn by the patient 100. For example the patient 100 can wear a wearable sensor 200a, 200b on each limb. This may be desirable for patients 100 at particular risk and increased sensor data collection is required.

In some examples, the plurality of wearable sensors 200a, 200b may establish a personal area network (PAN) or a body area network (BAN) for the wearable sensors 200a, 200b to communicate with each other. In some examples, the plurality of wearable sensors 200a, 200b can establish a mesh network between each other. This is described in further detail with reference to FIG. 5a.

In a less preferred example, the wearable sensors 200a, 200b can be connected to the control device 300 via a wired connection. However, a wired connection between the wearable sensors 200a, 200b and the control device 300 may interfere with the patient's 100 movement of their body.

As mentioned above, a communication interface between the wearable sensors 200a, 200b and the control device 300 is provided via a Bluetooth® transceiver, Wi-Fi transceiver, GSM transceiver, LTE transceiver, or any other similar licensed or unlicensed mobile communications interface.

This process is described in greater detail below and with reference to FIG. 4a. The control device 300 may be any mobile or remote processing or any other suitable control device 300. In some examples, the control device 300 is a mobile phone device such an Apple™ iPhone™, iPad™, Apple Watch™, Android™ device, Wear OS™ device, Laptop device, or similar. The control device 300 receives the data from the wearable sensors 200a, 200b, processes the received data, determines a patient condition, and executes an escalation process where appropriate. This process is described in greater detail below and with reference to FIG. 4b.

In some alternative examples, the control device 300 comprises the wearable sensor 200a. In this way, the control device 300 can be a smartphone which comprises one or more accelerometers (e.g. a 6-axis accelerometer). The smartphone is then mounted in a strap and worn on the user's arm. The smartphone mounted on one arm will function both as a wearable sensor 200 and as the control device 300 according to the present application. In addition, another wearable sensor 200b e.g. a smartwatch can be worn on the other arm.

FIG. 4a shows an example of the process flow for the data processing device 210. The process flow shown in FIG. 4a may be executed in a continuous loop or periodically. Where the process is executed periodically, an energy saving mode may be employed between executions to minimise battery usage.

In step 400, sensor output is received by data processing device 210 from sensor package 212. Where the sensor data is generated by e.g. a piezoelectric component, the output from the piezoelectric component may require an optional pre-processing step 410 to ensure a desired analogue signal is produced. One example of a pre-processing step may be to reduce high frequency noise generated by the piezoelectric component from the analogue signal. Pre-processing step 410 may occur at the sensor package 212 or on the data processing device 210.

Subsequent to the pre-processing step 410, a conversion of the signal from an analogue signal to a digital signal may be performed in step 420. This analogue to digital conversion may occur at the sensor package 212 or on the data processing device 210.

The digital signal is then processed in step 430 to reduce noise and to time stamp the sensor readings.

Step 430 may optionally comprise converting the acceleration vector generated by the accelerometer into a norm of the acceleration vector. i.e. The acceleration vector is converted to a strictly positive length in a single direction. This provides several advantages, including a reduced storage space for storing the vector data and an invariance to accelerometer orientation. Other filters are envisaged, in combination with the above or independently, steps to ensure that the filtered acceleration vector signal is invariant to gravity or orientation of the sensor package 212. In a preferred example, any of the preceding filtering steps are performed locally to the wearable sensors 200a, 200b, e.g. by data processing device 210.

Step 430 may further optionally comprise, in combination with the above or independently, applying a high pass filter to remove the acceleration vector resulting from the gravitational force on the accelerometer. This may be achieved by removing slow or unchanging acceleration vectors from a differential of the acceleration vector. This advantageously allows the removal of the noise resulting from gravitational forces.

Step 430 may further optionally comprise filtering signals resulting from movements of the human body that are not a direct consequence of the signals from the brain. In other words the wearable sensors 200a, 200b do not detect indirect outcomes of the electrical signals reaching muscles. Examples of movements that may be filtered include:

Movements that are only an indirect consequence of signals from the nervous system. e.g. When walking along, electrical signals may stimulate the arms to swing forwards and backwards to ensure balance. However, the downward swing of the arms during walking may be the consequence of gravity and the mechanics of the body, rather than the stimulation of any muscles. Consequently, where possible, these movements should be identified and removed from the acceleration vector generated by the accelerometer(s).

Therefore, in some examples the data processing device 210 filters signals in signal processing step 430 to exclude one or more movements not related to muscle movement due to the central nervous system. In some examples, the data processing device 210 filters the signal to exclude on or more of: Passive limb mechanics, Other biological signals, such as heart function, tremors, or other involuntary muscular movements, environmental noise (e.g. a bus or vehicle engine). In this way, the data processing device 210 filters signals and the movement of the patient 100 can be analysed when the patient 100 is awake and out of bed.

The sensor data is then optionally formatted according to a defined data structure in step 440 for transmission to control device 300. Finally, in step 450, wearable sensor 200a, 200b transmits the formatted sensor data to control device 300 using wireless networking interface 304.

In some examples, the steps as shown in FIG. 4a are carried out by the data processing device 210. In some other examples, the steps as shown in FIG. 4a are partly or completely carried out by a remote processing device, for example the control device 300.

FIG. 4b shows an example of the process flow for the control device 300. In some examples, the control device 300 is a mobile phone 300, but can be any suitable mobile device. The process flow shown in FIG. 4b may be executed in a continuous loop or periodically. Where the process is executed periodically, an energy saving mode may be employed between executions to minimise battery usage.

In step 460, the control device 300 receives the formatted sensor data from wearable sensor 200a, 200b. The received sensor data is then optionally consolidated, in step 470, with existing data previously received from wearable sensor 200a, 200b as well as any other wearable sensors transmitting sensor data to control device 300. In one example, the data is stored in a local database stored on control device 300. In one example, the system implements ‘chunking’ of the formatted sensor data, which comprises break the data into chunks, each chunk comprising a header which indicates some parameters (e.g. the time stamp for the recorded signal data, size etc.) This allows each data chunk to resynchronise the control device 300 to the clocks of wearable sensors.

In step 475, data analysis is performed on the sensor data by control device 300. A determination of a patient condition, such as an on-going stroke condition e.g. a transient ischemic attack, is then made in dependence on the data analysis 475. This comprises the determination that a first movement of the patient's body and a second movement of the patient are asymmetric as shown in step 480. A patient condition may be determined to be present based on the sensor data exceeding a predetermined asymmetry threshold.

The process of steps of data analysis 475, determination of patient event 480 and escalation process 490 are described in more detail below and in reference to FIGS. 6, 7 and FIGS. 9a to 9h, 10a to 10e, 11 a to 11e, 12a to 12i, 13a to 13c, 14a to 14f and 15a to 15b below.

Once escalation process 490 is complete and the patient has been unable to cancel the escalation of the patient condition, control device 300 makes a transmission to the remote network as shown in step 495. For example, the control device 300 contacts network point 308 via network interface 306 in step 495 to request emergency service for handling of the patient condition. In one example, mobile device may communicate at least one of a patient condition, a GPS location of the mobile device, a patient ID, a patient medical history, a recent sensor data report, etc.

It should be noted that step 475 and subsequent steps may be executed as a directly subsequent step to step 460. Alternatively, step 475 and subsequent steps may be executed in an independent loop that is triggered independently by e.g. a periodic timing interrupt.

In some examples, the steps as shown in FIG. 4b are carried out by the control device 300. In some other examples, the steps as shown in FIG. 4b are partly or completely carried out by the data processing device 210. In this way, the data processing device 210 is capable of carrying out all the steps in FIGS. 4a and 4b and can request emergency service for handling of the patient condition as shown in step 495.

In the present description, multiple wearable sensors are used to collect data. FIG. 5a shows a flow for collecting the data from more than one wearable sensor 200a, 200b at control device 300. In steps 510 and 520, sensor data from a first wearable sensor 200a positioned on a left side of the body and second wearable sensor 200b position on a right-hand side of the body is collected.

In step 530, sensor data from the first wearable sensor 200a is transmitted to the second wearable sensor 200b via wireless networking interface 304. Once the sensor data from first wearable sensor 200a is received at the second wearable sensor 200b, the sensor data from the first wearable sensor 200a is combined with the sensor data from the second wearable sensor 200a (collected previously in step 520) and transmitted to control device 300 in step 540.

In an alternative example to that shown in FIG. 5a, FIG. 5b shows an alternative flow for collecting the data from more than one wearable sensor at control device 300. In steps 510 and 520, sensor data from a first wearable sensor 200a positioned on a left side of the body and a second wearable sensor 200b positioned on a right-hand side of the body is collected.

In step 550, control device 300 instructs the first wearable sensor 200a via wireless network interface 304 to send sensor data collected by first wearable sensor 200a to control device 300. In step 560, the first wearable sensor 200a sends the collected data to control device 300 via wireless network interface 304. In step 570, control device 300 instructs the second wearable sensor 200b via wireless network interface 304 to send sensor data collected by the second wearable sensor 200b to control device 300. In step 580, the second wearable sensor 200b sends the collected data to control device 300 via wireless network interface 304.

Turning to FIG. 6, the process of steps of data analysis 475, determination of patient event 480 and escalation process 490 are described in more detail. FIG. 6 shows an example of the steps of data analysis 475, determination of patient event 480 and escalation process 490 as shown in FIG. 4b.

One of the symptoms of a patient having a stroke is sudden numbness or weakness of face, arm, or leg, especially on one side of the body. This means that the patient during a stroke event a patient is susceptible to asymmetric body movements.

Reference to the axes of symmetry of the body will now be discussed in reference to FIG. 8. FIG. 8 shows a perspective view of a patient's body 110 with a first axis 800. The first axis is also known as the vertical or longitudinal axis of the body 110. The body also has a second axis 802 and a third axis 804. The second axis 802 is also known as a transverse axis of the body 110. The third axis 804 is also known as a sagittal axis of the body 110.

FIG. 8 also shows the body 110 having a first plane 810, also known as a sagittal plane 810 in which the first axis 800 and the third axis 804 lie. The body 110 comprises a second plane 808, also known as a transverse plane 808, in which the second axis 802, and the third axis 804 lie. The body 110 also comprises a third plane 806, also known as the frontal plane, in which the first axis 800 and the second axis 802 lie.

In this way, symmetry of body movements relates to similarity of a first movement and a second movement either side of one or more planes 806, 808, 810 of the body 110. Accordingly, asymmetry of body movements relates to the dissimilarity of a first movement and a second movement either side of one or more planes 806, 808, 810 of the body 110. In this way, the planes 806, 808, 810 of the body 110 are also planes of symmetry for movement of the body 110.

For the purposes of clarity, reference will only be made to the first plane of symmetry 810. Advantageously, using the first plane of symmetry 810 for stroke-detection purposes reflects the fundamental plane of symmetry expressed in the hemispheres of the user's brain. However, reference to symmetry of movement of the body 110 can be in respect of any of the first, second and/or third planes of symmetry 806, 808, 810. The first plane of symmetry 810 divides the body 110 into a left-hand side 812 and a right-hand side 814. One symptom of a stroke may be is sudden numbness or weakness of face, arm, or leg, especially on one side of the body about the sagittal plane 810 of symmetry.

In an example, the patient 100 is prompted to perform one or more predetermined gestures as shown in step 600 of FIG. 6. The control device 300 may prompt the patient 100 for carrying out the predetermined gestures once a system condition has been met. In one example the system condition is that a threshold for the probability of a patient condition being present has been exceeded. The determination of the probability of a patient condition being present is discussed in GB1820892.6 which is incorporated herein in its entirety by reference.

In other examples, the system condition is that the control device 300 may prompt the patient 100 based on a timer, an external request or a request from the patient 100. For example, a remote medical practitioner may request the control device 300 to prompt the patient 100 to carry out the predetermined gestures. Alternatively, the patient 100 may not be feeling well and may wish to check whether they are suffering from a stroke.

The control device 300 then instructs the patient 100 to carry out one or more gestures. The patient 100 carries out a first gesture as shown in step 602. The patient 100 then carries a second gesture as shown in step 604. In some examples, the first and second gestures are carried out at the same time. In this way, the first gesture can be made by a first part of the patient's body 110 and the second gesture can be made by a second part of the patient's body 110. For example, the patient's right wrist 104a can make the first gesture and the patient's left wrist 104b can make the second gesture.

One such example of the first and second movements being carried out at different times is shown in FIGS. 14a to 14e whereby the patient is moving a leg. FIGS. 14a to 14f show a series of leg gestures to be performed by the patient 100 according to an example of the present application. In FIGS. 14a to 14e, a first leg 1400 is moved and then the same movement is repeated with the other leg 1402. The movement of the leg 1400 can be made at the hip joint backwards and forwards as shown in FIGS. 14a, 14b, 14c. Additionally or alternatively the movement of the leg can be made at the knee joint as shown in FIG. 14e or at the ankle joint as shown in FIG. 14f.

In some examples, the control device 300 displays the first and second gesture for the patient 100 to perform in steps 602, 604 on a screen. The control device 300 display an animation of the predetermined gestures so that the patient 100 may follow and repeat the same movements. Alternatively, the control device 300 prompts the patient 100 to carry out symmetrical movements with both arms or legs and leaves the patient 100 to decide which symmetrical movements they should perform.

Optionally in another example, the control device 300 does not prompt the patient 100 to perform the first and second gestures. In this case, the patient 100 performs the first and second gestures without a prompt from the control device 300. For example, the patent 100 may have been instructed to perform the first and second gestures periodically by a medical practitioner and the control device 300 then analysis the movement data in step 475 as previously discussed.

In one example, the first wearable sensor 200a is mounted on the patient's right wrist 104a and the second wearable sensor 200b is mounted on the patient's left wrist 104b. This means that the first gesture and the second gesture are made on opposite sides of a plane of symmetry (e.g. the sagittal plane) 810 of the patient's body 110.

The first and second wearable sensors 200a, 200b measure the movement made by a first part of the patient's body 110 and the movement made by a second part of the patient's body 110 whilst the gestures are being performed. The first and second wearable sensors 200a, 200b measure one or more of acceleration, velocity, timing, distance, range of movement of the right wrist 104a and the left wrist 104b with respect to the plane 810 of symmetry. The first and second wearable sensors 200a, 200b can also measure whether the first and second movements are jerky or smooth in nature.

In this way, the control device 300 analysis the movement data form the first and second wearable sensors 200a, 200b in step 475 and determines if there is a significant difference between the movement of the first gesture and the second gesture.

For example, the control device 300 may determine that the velocity, speed or acceleration of the first gesture and the second gesture may be different. For example, the control device 300 may determine that the range of movement of the first gesture and the second gesture may be different. For example, the control device 300 may determine that the direction of movement of the first gesture and the second gesture may be different. For example, the control device 300 may determine that the range of rotation of the first gesture and the second gesture may be different. For example, the control device 300 may determine that the timing of the first gesture and the second gesture may be different.

The control device 300 then determines whether the first and second gestures are symmetric as shown in step 606.

In some examples, the first and second gestures require the patient 100 moving their right wrist 104a and left wrist 104b together to perform a simple hand clap. Various other gestures can be performed additionally or alternatively. FIGS. 9a to 9h, 10a to 10e, 11a to 11e, 12a to 12i, 13a to 13c, 14a to 14f and 15a to 15b show other possible gestures for the patient 100 to perform during steps 602, 604 and/or 612 which are discussed in further detail below.

The control device 300 determines a differential for movement data of the first and second gestures. In some examples, the control device 300 determines that the first and second gestures are asymmetric when the differential is above a predetermined threshold. In some examples, the control device 300 determines that there is asymmetrical movement when the differential of the movement data between the first and second gesture is more than a difference of 5%, 10%, 20%, 30%, 40% or 50%. For example, the control device 300 determines that the patient's right wrist 104a moves 10% quicker and through a 10% greater range than the patient's left wrist 104b. In some other examples, the control device 300 determines an activity measurement and/or an energy expenditure associated with the movement data of the first and second gestures. Accordingly, the control device 300 determines that the first and second gestures are asymmetric when the differential of the activity measurement and/or an energy expenditure is above a predetermined threshold. In some examples, the control device 300 determines that there is asymmetrical movement when the differential of the activity measurement and/or an energy expenditure associated with movement data between the first and second gesture is more than a difference of 5%, 10%, 20%, 30%, 40% or 50%.

In some examples, the threshold of the differential for determining asymmetric movement of the patient's body 110 can be dynamic. In some examples, the threshold of the differential can be modified based on the patient's characteristics (e.g. health, age, sex, or any other suitable parameter that may affect the movement of the patient's body 110).

In some examples, the control device 300 is configured to determine the predetermined threshold based on the patient's historical movement data for the patient's body 110. This means that the control device 300 can determine whether the first and second gestures are significantly asymmetric with respect to patient historical movement data. For example, the patient 100 may have limited movement in a part of their body due to an old injury which may result in the appearance of asymmetric movement of the patient 100. In this way, the control device 300 can ignore asymmetric behaviour due non-stroke conditions.

In some examples, the control device 300 is configured to determine the predetermined threshold based on one or more pre-sets relating to a characteristic of the patient 100. For example, a present may describe movement relating to a young patient or an old patient. Additionally, the pre-sets relate to other characteristics of the patient 100 such as inactive, active etc.

When the control device 300 determines that the first and second gestures are asymmetric as shown in step 606, the control device 300 can perform the escalation process step 490. In some examples, the control device 300 can escalate and contact the medical services as shown in step 608.

If the control device 300 determines that the first and second gestures are symmetric, then the control device 300 can cancel the alert as shown in step 610. After the alert has been cancelled, the control device 300 continues to receive and analysis the data received from the wearable sensors 200a, 200b as discussed in reference to FIGS. 4a, 4b.

The steps 602, 604 of the patient 100 performing the first and second gestures are optionally carried out at the same time. For example, if the patient 100 is prompted to clap their hands together, then the first and second gestures of the right wrist 104a and the left wrist 104b moving towards each other will be measured by the wearable sensors 200a, 200b at the same time.

Additionally or alternatively, the steps 602, 604 can be carried out at different times. For example, the first gesture and the second gesture may be performed by the patient 100 at different times. For example, the patient 100 can perform a first gesture with the wearable sensor 200a mounted on the right wrist 104a and then subsequently perform a second gesture with the wearable sensor 200b mounted on the left wrist 104b. Optionally the determination of the patient event in step 480 can be carried out with a single wearable sensor 200a whereby the patient swaps the wearable sensor 200a from the right wrist 104a to the left wrist 104a after making the first gesture but before making the second gesture.

Optionally, on determination that there is asymmetric movement between the first and second gestures 606, the control device 300 may prompt the patient 100 to carry out additional movements as shown in step 612. The control device 300 may prompt the patient 100 to carry out similar or more complex predetermined gestures. The additional gestures may be used by the control device 300 to determine the presence of asymmetrical body movement for a borderline case. Alternatively, the patient 100 or a remote medical practitioner may request the additional movements to be performed by the patient 100 as a double check.

The control device 300 then determines in step 614 whether the additional gestures performed in step 612 are asymmetric. The determining step of 614 is similar to step 606. If the control device 300 determines that the additional gestures are asymmetric, then the control 300 escalates to the medical services in step 608 as before. Similarly, if the control device 300 determines that the additional gestures are actually symmetrical, then the control device cancels the alert in step 610 as previously discussed.

Optionally, in one example, step 612 or step 606 may additionally comprise further tests for the patient 100 to determine whether the medical services should be alerted. In some examples, the further tests may be required to be performed before the patient performs the gestures in steps 606 or 612. Alternatively, the further tests may be required to be performed after the patient performs the gestures in steps 606 or 612.

For example, the control device 300 may display a countdown alert 700 presented to the patient 100 on the display of control device 300. The countdown alert 700 may comprise a simple countdown alert showing a countdown in seconds before which control device 300 will move to step 608 to alert the medical services. The patient 100 has the option of cancelling the countdown at any time. If the patient 100 fails to respond to the countdown alert, control device 300 will move straight to step 608 to alert medical service. Where the patient 100 cancels the countdown by performing the required cancellation task (e.g. pressing a CANCEL button), the patient 100 is presented with the option to reset the escalation process. Alternatively, the patient 100 may be presented with the option of moving straight to step 608 to alert medical service if the patient 100 feels that something is still wrong.

In one optional example, the patient 100 cannot cancel the countdown alert until all of the user tests are successfully completed. In one example, a two-part countdown is used. Countdown 1 is a short-term countdown (e.g. less than 60 seconds) and may be cancelled by the patient 100. Countdown 2 is a longer-term countdown (e.g. longer than 60 seconds) that can only be cancelled by successfully completing all of the patient 100 tests.

In another optional example, the patient 100 is additionally required to perform simple cognitive tests such as performing simple mental arithmetic as shown in 710. If the control device 300 determines that the patient 100 has failed the further tests 700, 710, then the control device 300 can escalate and alert the medical services in step 608.

Other gestures for determining asymmetry of a patient's movement will now be described in reference to FIGS. 9a to 9h, 10a to 10e, 11a to 11e, 12a to 12i, 13a to 13c, 14a to 14f and 15a to 15b.

FIGS. 9a to 9h, 10a to 10e, 11a to 11e, 12a to 12i, 13a to 13c, 14a to 14f and 15a to 15b show a series of gestures to be performed by the patient 100 according to an example of the present application.

FIGS. 9a to 9d show the patient 100 moving their right arm 900 from a downwardly pointing direction to an over the head upwardly pointing direction. FIGS. 9e to 9h show the patient moving their left arm 902 from a downwardly pointing direction to an over the head upwardly pointing direction. The gestures in FIGS. 9a to 9d mirror the gestures in FIGS. 9e to 9h. The gestures are mirrored about the sagittal plane 810 of symmetry. In this way the stroke detection apparatus 102 can determine whether there is asymmetry between the movements as previously described in reference to FIGS. 4a, 4b above. The gestures as shown in FIGS. 9a to 9h can be carried out at the same time or at different times. The patient 100 is moving the arms 900, 902 in a plane parallel with the frontal plane 806.

FIGS. 10a to 10e show a series of arm gestures whereby the patient 100 moves their arms 900, 902 from a first position resting on their hips in FIG. 10a to a position where their arms 900, 902 are pointing upwardly over their head in FIG. 10e. The gestures in FIGS. 10a to 10e are similar to the gestures as shown in FIGS. 9a to 9h except that both arms 900, 902 are moved at the same time. The gestures as shown in 10a to 10e can be used to determine asymmetry of the patient movement about the sagittal plane 810 of symmetry. The patient 100 is moving the arms 900, 902 in a plane parallel with the frontal plane 806.

FIGS. 11 a to 11e again show a series of arm gestures similar to FIGS. 10a to 10e. However, in FIGS. 11a to 11e, the patient's right arm 900 is resting on the hips and the patient's left arm 902 is raised over their head. The patient then proceeds to raise and lower the right and left arms 900, 902 respectively until in the position shown in FIG. 11e. The patient 100 can then repeat the process to return to the original position in FIG. 11a. The gestures in FIGS. 11a to 11e may be useful for a patient 100 to carry out because it requires an element of cognitive ability to coordinate the arms 900, 902. An inability of the patient to coordinate their arms 900, 902 may be another indicator of a stroke condition.

FIGS. 12a to 12i again show a series of arm gestures similar to FIGS. 10a to 10e. However, the patient 100 swings both their arms 900, 902 in an anticlockwise direction until the left arm 902 is above the shoulder in FIG. 12c. The patient 100 then swings both their arms 900, 902 down and up in a clockwise direction until the right arm 900 is above the shoulder as shown in FIG. 12i. Again, the gestures in FIGS. 12a to 12i may be useful for a patient 100 to carry out because it requires an element of cognitive ability to coordinate the arms 900, 902.

FIGS. 13a to 13c show a series of arm gestures for the patient 100. The patient 100 moves the arms 900, 902 in a plane parallel with the sagittal plane 810. The patient 100 moves their arms through the transverse plane 808 and rotates them from a first position above their head shown in FIG. 13a down to a second position by their sides as shown in FIG. 13c.

FIGS. 14a to 14f are a series of leg gestures and have been previously discussed above.

FIGS. 15a and 15b show another series of arm gestures. The patient 100 moves their arms 900, 902 from a first position where the patient 100 has their arms 900, 902 pointing out from their sides at shoulder height as shown in FIG. 15a to a second position whereby the patient 100 has their arms 900, 902 together pointing out in front of them. The patient 100 moves their arms 900, 902 in a plane parallel with the transverse plane 808.

In other examples, the patient 100 can be prompted by the control device 300 to make other movements. For example, the patient 100 is required to move their hands in circles in front of them in a plane parallel with the frontal plane 806 with each hand moving in the same or opposite directions. Additionally or alternatively the patient 100 is required to move their hands in a linear motion up and down in front of them in a plane parallel with the sagittal plane 810 with each hand moving in the same or opposite directions.

In other examples, the patient 100 can be prompted to make diadochokinetic movements. That is, rapid small movements such as moving the finger to the nose, alternating hand clapping or toe tapping movements.

In this way, the gestures are complex enough such that the control device 300 is can separate it from involuntary movements such as tremor and twitches and jerks that may be seen in epileptic seizures. The gestures can be performed in any place and at any time of the day. This means that the patient 100 can perform the gestures when prompted by the control device 300 even when travel or away from home.

Advantageously, the control device 300 can determine the movement of the gestures with only the wearable sensors 200a, 200b. This means that not further equipment is necessary to determine the movement of the patient 100. In this way, the gestures can be used demonstrate the symmetrical movement of one or more particular muscle groups or multiple muscles groups in a clear and simple way. This means that the gestures can be easy for the patient 100 to carry out whilst being sufficiently complex to determine asymmetric movement indicative of a stroke condition.

In another example two or more examples are combined. Features of one example can be combined with features of other examples.

Examples of the present invention have been discussed with particular reference to the examples illustrated. However it will be appreciated that variations and modifications may be made to the examples described within the scope of the invention.

Claims

1. A stroke detection apparatus comprising:

at least one wearable sensor configured to generate movement data of at least a portion of a user's body; and

a data processing device including a processor, data processing device configured to process first movement data for a first movement and second movement data for a second movement received from the at least one wearable sensor, determine asymmetry of the user's movement based on the first movement data and the second movement data, and generate a stroke detection signal based on the determined asymmetry.

2. The stroke detection apparatus according to claim 1, wherein the data processing device is configured to determine the asymmetry of the user's movement based on the first movement data and second movement data, which is based on the user performing at least one body gesture.

3. The stroke detection apparatus according to claim 2, wherein the data processing device is configured to prompt the user to perform the at least one body gesture.

4. The stroke detection apparatus according to claim 1, wherein the data processing device is configured to generate the stroke detection signal in response to determining that the asymmetry of the user's movement exceeds a threshold.

5. The stroke detection apparatus according claim 4, wherein the data processing device is configured to determine the threshold based on historical movement data for the user's body.

6. The stroke detection apparatus according to claim 4, wherein the threshold is based on one or more pre-sets relating to user characteristics.

7. The stroke detection apparatus according to claim 1, wherein the data processing device is configured to generate an automated emergency services request in response to the stroke detection signal.

8. The stroke detection apparatus according to claim 1, wherein the at least one wearable sensor is configured to be worn on at least one of an arm or a leg of the user.

9. The stroke detection apparatus according to claim 8, wherein

the at least one wearable sensor includes a first wearable sensor and a second wearable sensor,

first wearable sensor is configured to be worn on one of the user's wrists, and

the second wearable sensor is configured to be worn on the other of the user's wrists.

10. The stroke detection apparatus according to claim 1, wherein the at least one wearable sensor comprises:

a first sensor configured to measure movement on a first side of a plane of symmetry the user's body; and

a second wearable sensor configured to measure movement on a second side of the plane of symmetry of the user's body.

11. A stroke detection apparatus according to claim 10, wherein the plane of symmetry of the user's body is one or more of a sagittal plane, a frontal plane or a transverse plane.

12. The stroke detection apparatus according to claim 1, wherein the at least one wearable sensor is configured to measure the first movement data and the second movement data at the same time.

13. The stroke detection apparatus according to claim 1, wherein the at least one wearable sensor is configured to measure the first movement data and the second movement data at different times.

14. The stroke detection apparatus according to claim 1, wherein the at least one wearable sensor is configured to transmit the first movement data and the second movement data to the data processing device.

15. The stroke detection apparatus according to claim 1, wherein the data processing device is configured to prompt the user to perform additional body gestures in response to determining the asymmetry of the user's movement based on the first movement data and the second movement data.

16. A method of generating a stroke detection signal, the method comprising:

generating first movement data and second movement data with at least one wearable sensor, the first movement data being for first movement of a first portion of the user's body and the second movement data being for a second movement of a second portion of the user's body;

processing the first movement data and the second movement data;

determining asymmetry of the user's movement based on the first movement data and the second movement data; and

generating a stroke detection signal based on the determined asymmetry.

17. A stroke detection apparatus comprising:

at least one wearable sensor configured to generate movement data based on movement of a portion of a user's body; and

a data processing device including a processor, the data processing device configured to detect first asymmetry of the user's movement based on first movement data and second movement data from the at least one wearable sensor, the first movement data being for a first movement of a first portion of the user's body and the second movement data being for a second movement of a second portion of the user's body, prompt the user to perform at least one body gesture in response to detecting the first asymmetry of the user's movement; determine whether second asymmetry of the user's movement is present based on third movement data and fourth movement data from the at least one wearable sensor when performing the at least one body gesture, the third movement data being for a third movement of the first portion of the user's body and the fourth movement data being for a fourth movement of the second portion of the user's body; and generate a stroke detection signal in response to determining that the second asymmetry of the user's movement is present.

18. The stroke detection apparatus according to claim 17, wherein the data processing device is configured to detect the first asymmetry of the user's movement based on whether asymmetry between the first movement and the second movement exceeds an asymmetry threshold.

19. The stroke detection apparatus according to claim 18, wherein the data processing device is configured to determine whether the second asymmetry is present based on whether asymmetry between the third movement and the fourth movement exceeds the asymmetry threshold.

20. The stroke detection apparatus according claim 19, wherein the data processing device is configured to determine the asymmetry threshold based on at least one of historical movement data for the user's body or one or more pre-sets relating to characteristics of the user.