U.S. patent application number 17/582108 was filed with the patent office on 2022-05-12 for method and device for respiratory monitoring.
The applicant listed for this patent is PMD DEVICE SOLUTIONS LIMITED. Invention is credited to Stephen CUSACK, Christopher KINSELLA, Myles MURRAY.
Application Number | 20220142504 17/582108 |
Document ID | / |
Family ID | 1000006104381 |
Filed Date | 2022-05-12 |
United States Patent
Application |
20220142504 |
Kind Code |
A1 |
MURRAY; Myles ; et
al. |
May 12, 2022 |
METHOD AND DEVICE FOR RESPIRATORY MONITORING
Abstract
A respiration monitoring system has deformation transducers on a
flexible substrate arranged to adhere to a patient's torso. A
processor receives signals in channels from the transducers and
processes them to eliminate, reduce or compensate for noise arising
from patient motion artefacts, to provide an output representative
of respiration. The transducers have a size and a mutual location
on the substrate so that a first transducer can overlie at least
part of the 10.sup.th rib and a second transducer can overlie at
least part of the 11.sup.th rib or the abdomen, and the processor
processes data from the first transducer as being primarily
representative of rib distending respiration and from the second
transducer as being primarily representative of either diaphragm
respiration or patient motion artefacts.
Inventors: |
MURRAY; Myles; (Cork,
IE) ; CUSACK; Stephen; (Cork, IE) ; KINSELLA;
Christopher; (Cork, IE) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
PMD DEVICE SOLUTIONS LIMITED |
Cork |
|
IE |
|
|
Family ID: |
1000006104381 |
Appl. No.: |
17/582108 |
Filed: |
January 24, 2022 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14769044 |
Aug 19, 2015 |
11259716 |
|
|
PCT/EP2014/053048 |
Feb 17, 2014 |
|
|
|
17582108 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
A61B 2562/0261 20130101;
A61B 5/725 20130101; A61B 5/113 20130101; A61B 5/0803 20130101;
A61B 5/6833 20130101; A61B 5/4818 20130101; A61B 5/7257 20130101;
A61B 5/6823 20130101; A61B 5/0816 20130101; A61B 2562/0219
20130101; A61B 5/0826 20130101; A61B 5/0022 20130101; A61B
2560/0425 20130101; A61B 5/7207 20130101; G16H 40/67 20180101; A61B
2560/0412 20130101; A61B 5/7282 20130101; A61B 5/7246 20130101;
A61B 2562/164 20130101; A61B 2560/045 20130101; G16H 20/40
20180101; A61B 2562/04 20130101; A61B 5/7225 20130101; A61B
2560/0285 20130101; A61B 5/721 20130101 |
International
Class: |
A61B 5/08 20060101
A61B005/08; A61B 5/00 20060101 A61B005/00; A61B 5/113 20060101
A61B005/113 |
Foreign Application Data
Date |
Code |
Application Number |
Feb 20, 2013 |
IE |
2013/0062 |
Claims
1. A respiration monitoring system comprising: a flexible
substrate, an adhesive arranged on a surface of the flexible
substrate to releasably adhere the flexible substrate to a
patient's torso, a plurality of embedded deformation transducers
fixed to said flexible substrate including at least a first
transducer and a second transducer, the first transducer and the
second transducer being located on the substrate at an angle
relative to a mutual location on the substrate, the first
transducer and the second transducer having a size and the mutual
location on the substrate so that simultaneously the first
transducer is configured to overlie at least part of a patient's
10.sup.th rib and the second transducer is configured to overlie at
least part of a patient's 11.sup.th rib or abdomen, the transducers
being positioned on the substrate to enable measuring both thoracic
and abdominal displacement in a single location, an electronic
controller releasably mounted on the substrate, the electronic
controller being positioned on a same side of both the first
transducer and the second transducer, the electronic controller
receiving signals by conductors from the first transducer and the
second transducer, and an accelerometer producing an output signal
representative of a posture of the patient's torso, wherein the
electronic controller is configured to receive signals from the
first transducer and the second transducer and to compensate for
motion noise based on the output signal from the accelerometer, and
to thereby derive an output representative of respiration based
upon the signals from the first transducer and the second
transducer.
2. The respiration monitoring system of claim 1, wherein the system
comprises a unitary sensor for adhering to a patient's skin, said
sensor including: the substrate with the deformation transducers,
and the electronic controller, wherein the electronic controller is
included in a housing on the substrate with a signal conditioning
circuit, and wherein the electronic controller housing is
releasably mounted on the substrate.
3. The respiration monitoring system of claim 1, wherein the
electronic controller is configured to trigger an artefact
detection algorithm at regular intervals in which signals which are
outside predetermined limits of measurement are removed.
4. The respiration monitoring system of claim 1, wherein the
electronic controller is configured to execute, when determining
respiration rate, a frequency domain algorithm
5. The respiration monitoring system of claim 4, wherein the
electronic controller is configured to execute the frequency domain
algorithm to take accelerometer data from the accelerometer as a
secondary input and to compensate for cyclical interference from
the subject or environment such as walking, by extracting frequency
domain information from the accelerometer data.
6. The respiration monitoring system of claim 5, wherein the
electronic controller is configured to detect and compensate for
movements using the accelerometer data.
7. The respiration monitoring system of claim 6, wherein the
electronic controller is configured to perform at least one of fall
detection, step detection and orientation monitoring using the
accelerometer data.
8. The respiration monitoring system of claim 1, wherein the
electronic controller is configured to detect inhalation and
exhalation events and monitor lung capacity.
9. The respiration monitoring system of claim 1, wherein the
electronic controller is configured to execute, when determining
respiration rate, a time domain algorithm.
10. The respiration monitoring system of claim 9, wherein the
electronic controller is configured to execute the time domain
algorithm to produce a waveform represented by a repeating pattern
of peaks and troughs at a rate indicative of the respiratory rate
of the patient, and detect a distance between the peaks and a
distance between the troughs in the waveform to derive the
respiration rate.
11. The respiration monitoring system of claim 1, wherein the
electronic controller is configured to produce a waveform
represented by a repeating pattern of peaks and troughs at a rate
indicative of the respiratory rate of the patient.
12. The respiration monitoring system of claim 11, wherein the
electronic controller is configured to produce the waveform for
diagnosis of dysfunctional breathing events in sleeping subjects by
detecting portions of the waveform indicative of a dysfunctional
breathing event.
13. The respiration monitoring system of claim 1, wherein the
electronic controller is configured to receive a unique identifier
for a use with a particular subject, and to discontinue or erase
said identifier upon removal of the substrate from the subject
and/or re-charging for a next use.
14. The respiration monitoring system of claim 1, wherein the first
transducer and the second transducer are of equal length, width,
thickness and composition.
15. A respiration monitoring system comprising: a flexible
substrate, an adhesive arranged on a surface of the flexible
substrate to releasably adhere the flexible substrate to a
patient's torso, a plurality of embedded deformation transducers
fixed to said flexible substrate including at least a first
transducer and a second transducer, an accelerometer configured to
produce an output signal representative of a posture of the
patient's torso, and an electronic controller releasably mounted on
the substrate, the electronic controller configured to receive
signals by conductors from the first transducer and the second
transducer, to compensate for motion noise based on the output
signal from the accelerometer, and to derive an output
representative of respiration based upon the signals from the first
transducer and the second transducer by executing a time domain
algorithm.
16. The respiration monitoring system of claim 15, wherein the
electronic controller is configured to execute the time domain
algorithm to produce a waveform represented by a repeating pattern
of peaks and troughs at a rate indicative of the respiratory rate
of the patient, and to detect a distance between the peaks and a
distance between the troughs in the waveform to derive the
respiration rate.
17. The respiration monitoring system of claim 15, wherein the
system comprises a unitary sensor for adhering to a patient's skin,
said sensor including: the substrate with the deformation
transducers, and the electronic controller, wherein the electronic
controller is included in a housing on the substrate with a signal
conditioning circuit, and wherein the electronic controller housing
is releasably mounted on the substrate.
18. A respiration monitoring system comprising: a flexible
substrate, an adhesive arranged on a surface of the flexible
substrate to releasably adhere the flexible substrate to a
patient's torso, a plurality of embedded deformation transducers
fixed to said flexible substrate including at least a first
transducer and a second transducer, an accelerometer configured to
produce an output signal representative of a posture of the
patient's torso, and an electronic controller releasably mounted on
the substrate, the electronic controller configured to receive
signals by conductors from the first transducer and the second
transducer, to compensate for motion noise based on the output
signal from the accelerometer, and to derive an output
representative of respiration based upon the signals from the first
transducer and the second transducer by producing a waveform
represented by a repeating pattern of peaks and troughs at a rate
indicative of the respiratory rate of the patient.
19. The respiration monitoring system of claim 18, wherein the
electronic controller is configured to produce the waveform for
diagnosis of dysfunctional breathing events in sleeping subjects by
detecting portions of the waveform indicative of a dysfunctional
breathing event.
20. The respiration monitoring system of claim 18, wherein the
system comprises a unitary sensor for adhering to a patient's skin,
said sensor including: the substrate with the deformation
transducers, and the electronic controller, wherein the electronic
controller is included in a housing on the substrate with a signal
conditioning circuit, and wherein the electronic controller housing
is releasably mounted on the substrate.
Description
INTRODUCTION
Field of the Invention
[0001] The invention relates generally to devices useful in
measuring and monitoring respiratory events in a human subject.
Prior Art Discussion
[0002] Respiratory rate is the measure of the number of breaths a
person has per minute and is a key vital sign in human subjects.
Spirometry is the measure of lung capacity or lung volume in a
human subject. Deterioration of these respiratory functions is the
decline or increase of these measures. Measurements outside, or
approaching the boundaries of the predetermined physiological
normal values are a pre-indicator to harmful and fatal emerging
ailments in human subjects.
[0003] Respiration is the process by which living organisms take in
oxygen and convert it to energy. Part of this process is the
mechanical inhalation of air which, for humans, is done via the
nose and mouth. The mechanical respiratory effort is produced by
the muscles of respiration. These muscles aid in bath inspiration
and expiration. The muscle groups which make up this collection
include the diaphragm, external intercostal, and internal
intercostal muscles. This process is known as respiratory effort,
and it is enabled by either or both of: [0004] a) The partial or
total displacing of the rig cage (hereafter referred to as rib
breathing) upwards and outwards by the external and internal
intercostal muscles, along a fixed locus, to produce a vacuum
inside the thoracic cavity, thus drawing air into the lungs to
enabling respiration to occur, [0005] b) The diaphragm pushing down
into the abdominal region (hereafter referred to as diaphragm
breathing) forcing the abdominal organs to distend outward and thus
producing a vacuum in the thoracic region by increasing the volume
in the abdominal region.
[0006] The above movements may be referred to as distending the rib
cage.
[0007] A reason for the respiratory rate or capacity of the lung of
a patient to fluctuate over a period of time, where physical
activity is not considered, can be the result of physiological
changes in the health of the patient. Infections in the body can
result in a fever and higher heart rate. An infection also produces
an increase in respiratory effort, and could be viral or bacterial,
or as a result of the environment or complications resulting from
medication or surgery. Pneumonia, chronic obstructive pulmonary
disease (COPD), and sepsis are all ailments representative of the
above and can be indicated by fluctuating respiratory function.
This may express either as an alteration of the respiratory rate of
the patient or the capacity of the patient to draw in air for
efficient respiration.
[0008] Respiratory rate is a predominant metric in a predicative
patient scoring system known as the Early Warning Score (EWS).
Chronic patients suffering from lung diseases such as COPD can be
monitored over long periods of time by measuring their lung
capacity. As lung diseases affect the normal mechanical operation
of respiratory effort, measuring the ability of patients to breathe
deeply is also a key measure of their deterioration or recovery.
The comprehensive measure of respiratory rates enables medical
staff to better assess the EWS with high accuracy and intervene
sooner.
[0009] US2012/0296221 (Philips) describes a method and apparatus
for determining a respiration signal. A single multi-axial
accelerometer is positioned on the body. WO2009/074928 (Philips)
describes use of ECG electrodes on an elastically deformable
bridge, and there is also a strain sensor and an accelerometer.
[0010] The invention is directed towards providing a system for
respiratory monitoring which is simpler and/or more robust, and/or
more reliable than the prior art.
SUMMARY OF THE INVENTION
[0011] According to the invention, there is provided a respiration
monitoring system comprising: [0012] a plurality of deformation
transducers on a flexible substrate arranged to adhere to a
patient's torso, and [0013] a processor adapted to receive signals
from said transducers and to process them to eliminate, reduce, or
compensate for noise arising from patient motion artefacts, to
provide an output representative of respiration.
[0014] In one embodiment, the deformation transducers are elongate
and are arranged on the substrate at a mutual acute angle.
Preferably, the angle is in the range of 20.degree. to 80.degree.,
preferably 25.degree. to 40.degree., and most preferably in the
region of 27.degree. to 33.degree..
[0015] In one embodiment, the transducers have a size and a mutual
location on the substrate so that a first transducer can overlie at
least part of the 10.sup.th rib and a second transducer can overlie
at least part of the 11.sup.th rib or the abdomen, and the
processor is adapted to process data from the first transducer as
being primarily representative of rib distending respiration and
from the second transducer as being primarily representative of
either diaphragm respiration or patient motion artefacts.
Preferably, the deformation transducers are positioned at an acute
angle to each other on the substrate, and the processor is adapted
to process data from the transducers on the basis that an apex
defined by said mutual position is pointed rearwardly and
downwardly with respect to a human subject.
[0016] In one embodiment, the system further comprises an
accelerometer. In one embodiment, the processor is adapted to
process an accelerometer output by correlating the degree of motion
artefacts with bodily displacement for aiding the process of
eliminating motion artefacts and detect cyclical movements.
[0017] In one embodiment, the system includes a gyroscope.
Preferably, the processor is adapted to process a gyroscope output
by enabling the posture of the body to be known to the processor,
thus enabling anomalies of the transducers to be accounted for.
[0018] In one embodiment, the system comprises a unitary sensor for
adhering to a patient's skin, said sensor including the substrate
with the deformation transducers, and the processor. In one
embodiment, the processor is included in a housing on the substrate
with a signal conditioning circuit. Preferably, the processor
housing is releasably mounted on the substrate.
[0019] n one embodiment, the processor is adapted to communicate
wirelessly via an interface to a host processor.
[0020] In one embodiment, the deformation transducers include at
least two strain transducers. In one embodiment, in the strain
transducers are piezoelectric transducers.
[0021] In one embodiment, the processor is adapted to detect
excessive displacements resulting in over-pressurisation from
invasive or non-invasive artificial ventilation machines.
[0022] In one embodiment, the processor is adapted to perform
signal conditioning by baseline subtraction against an input
voltage signal from the transducers, and to further condition the
signal using an exponential moving average filter.
[0023] In one embodiment, the processor is adapted to trigger an
artefact detection algorithm at regular intervals in which signals
which are outside the limits of measurement are removed.
[0024] In one embodiment, the processor is adapted to execute a
time domain algorithm when determining respiration rate.
[0025] In one embodiment, the processor is adapted to execute a
frequency domain algorithm when determining respiration rate.
Preferably, the time domain algorithm checks distances between
peaks and troughs in a respirator waveform and derives a
respiration rate. In one embodiment, the frequency domain algorithm
uses a fast Fourier transform to extract frequency domain
information. In one embodiment, the sensor includes an
accelerometer and the processor is adapted to execute the frequency
domain algorithm to take accelerometer data as a secondary input
and to compensate for cyclical interference from the subject or
environment such as walking, by extracting frequency domain
information from the accelerometer.
[0026] In one embodiment, the processor is adapted to detect and
compensate for large movements using the accelerometer data.
[0027] In one embodiment, the processor is adapted to assume that a
deformation waveform is represented by a repeating pattern of peaks
and troughs at a rate indicative of the respiratory rate of the
subject, and magnitude of a received transducer signal is
considered only of importance if said signal becomes so large as to
exceed an output limit of the sensor, or so small as to become
indistinguishable from noise.
[0028] In one embodiment, the processor is adapted to detect apnea
events in sleeping subjects. Preferably, the processor is adapted
to recognize missing breathing signals as representative of
apnea.
[0029] In one embodiment, the system further comprises a wireless
transceiver and the processor is adapted to transmit to an external
device data to display a respiratory rate history of a subject.
[0030] In one embodiment, the processor is adapted to receive a
unique identifier for a use with a particular subject, and to
discontinue or erase said identifier upon removal and/or
re-charging for a next use. In one embodiment, the processor is
adapted to save a scanned Medical Record Number (MRN) as a unique
identifier. In one embodiment, the processor is adapted to
automatically apply a temporary identifier upon removal or
re-charging.
[0031] In another aspect, the invention provides a method of
monitoring respiration of a human subject using a system
comprising: [0032] a plurality of deformation transducers on a
flexible substrate arranged to adhere to a patient's torso, and
[0033] a processor adapted to receive signals from said transducers
and to process them to eliminate, reduce, or compensate for noise
arising from patient motion artefacts, to provide an output
representative of respiration, [0034] the method comprising the
steps of adhering the substrate to a human subject and the
processor processing signals from the transducers to derive an
output representative of respiration of the human subject.
[0035] In one embodiment, the substrate is placed so that a first
transducer substantially overlies a 10.sup.th rib and a second
transducer overlies a floating rib or the abdomen, and the
processor monitors signals from said transducers by treating
signals arising from deformation of the first transducer as being
representative of rib distending respiration and by treating
signals arising from deformation of the second transducer as being
representative of diaphragm breathing or a non-respiration
artefact.
[0036] In one embodiment, the processor automatically decides on
what the deformation of the second transducer represents according
to a signal from an auxiliary sensing device.
[0037] In one embodiment, the auxiliary sensing device is an
accelerometer.
BRIEF DESCRIPTION OF THE DRAWINGS
[0038] The invention will be more clearly understood from the
following description of some embodiments thereof, given by way of
example only with reference to the accompanying drawings in
which:
[0039] FIG. 1 is a perspective view of a sensor of a system of the
invention,
[0040] FIG. 2 is an exploded view of a re-usable part,
[0041] FIG. 3 is a set of views of layers of the transducers,
and
[0042] FIG. 4 is an exploded perspective view of the layers of the
sensor's substrate,
[0043] FIG. 5 is a diagram showing an optimal position for sensor
on the body;
[0044] FIG. 6 is a Hock diagram of the sensor system;
[0045] FIG. 7 shows an instrumental amplifier circuit to create a
high common-mode resection and high gain to eliminate any mutual
environment interference and boost the signal for processing
respectively for a single transducer;
[0046] FIG. 8 shows low pass filtering circuitry to produce a
single output (per sensor) for digital signal processing, with high
frequency components removed, for a single transducer
embodiment;
[0047] FIG. 9 is a flowchart outlining an algorithm used to process
the sensor outputs to generate a respiratory rate;
[0048] FIG. 10 is a flowchart detailing the use of a Fast Fourier
Transform as part of the algorithm in FIG. 9;
[0049] FIG. 11 is a flowchart detailing the use of a time domain
algorithm as part of the algorithm in FIG. 9;
[0050] FIGS. 12(a) to 12(f) are plots, having a normalised
numerical vertical axis and a time horizontal axis, showing various
transducer and system signals as follows: [0051] FIG. 12(a) shows
the raw output from a single transducer over a 60 s time period
showing the peaks and troughs indicative of normal breathing, in
which one movement artefact can be seen as an increase in the
signal strength, [0052] FIG. 12(b) shows the same signal with
baseline correction and smoothing applied, [0053] FIG. 12(c) shows
the results from an artefact detection function, [0054] FIG. 12(d)
shows the same signal with the section designated as artefact
smoothly removed from the waveform. [0055] FIG. 12(e) shows the
signal in FIG. 12(d) with peaks and troughs identified by the
system's processor, and [0056] FIG. 12(f) shows the frequency
spectrum of the signal, with the most prominent signal highlighted
at approximately 0.25 Hz, or 15 breaths per minute;
[0057] FIG. 13(a) shows signals for each channel representative of
normal breathing over a 60 s period which has minimal movement
artefact while the subject is standing, and FIG. 13(b) shows the
corresponding signal when the subject is lying on their back, again
having a normalised numerical vertical axis and a time horizontal
axis; and
[0058] FIG. 14(a) shows the signals from transducers over a period
of 1 hour, collected when the subject was asleep. FIG. 14(b) shows
the signals from an accelerometer housed in the electronics housing
for the same period, again having a normalised numerical vertical
axis and a time horizontal axis.
DESCRIPTION OF THE EMBODIMENTS
[0059] Referring to FIGS. 1 and 2 a monitoring system comprises a
sensor I with a disposable substrate 2 for adhering to a patient
torso and a re-usable electronics controller 3 adhered to the
substrate 2. The substrate 2 comprises a body 4 within which are
embedded elongate transducers 5 and 6 for measuring deformation.
These are linked by conductors 7 to the controller 3. A sensor
system includes the sensor 1 and also a host processor linked by
cable or wirelessly, and this may in turn communicate with a remote
server.
[0060] The controller 3 comprises a plastics housing with a top
part 10 and a base 11, containing a circuit board 12 and a
rechargeable battery 13, and an alarm sounder 14. There is a
connector 15 for wired connection to an external device or host
system, although the circuit 12 is also Bluetooth enabled for
wireless communication with such a device or system.
[0061] The controller 3 is mechanically joined to the substrate 2
by use of an industrial grade hook and loop fastener with the hook
side on the side of the controller 3 and the loop side on the
consumable substrate 2. This construction allows for durable
attachment of the device. It further allows removal of these two
elements which is useful in a medical application where consumable
body contact sensors are desired to be for single patient single
use.
[0062] Referring in particular to FIG. 3 the transducers 5 and 6
comprise a piezoelectric film 5(a), 6(a) sandwiched between: [0063]
a coating ink pattern 5(b), 6(b) and a positive ink pattern 5(c),
6(c), on top; and [0064] a negative ink pattern 5(d), 6(d) and a
Mylar layer 5(e), 6(e) underneath.
[0065] The composition of the transducer is therefore a multi-layer
piezo stack separated by a metal foil. In this embodiment the piezo
stack is a multi-purpose, piezoelectric transducer for detecting
physical phenomena such as vibration or impact or general
deformation. The piezo film element is laminated to the sheet 5(e)
of polyester (Mylar), and produces a useable electrical signal
output when forces are applied to the sensing area.
[0066] This compositional stack is heat-laminated using a
translucent polymer. Each piezo film layer is partially extended to
form a terminal by which a clamp is fixed to. This provides a
secure electrical contact for the instrumentation amplifier
circuitry.
[0067] The substrate body 4 is shown in most detail in FIG. 4. It
comprises polypropylene clear release film 4(a), 3M.TM. medical
grade silicone adhesive 4(b), and a polyester layer 4(c). The
transducers 5 and 6 are located between the adhesive 4(b) and the
polyester 4(c) layers.
[0068] As shown in FIG. 5, the sensor may in one embodiment be
placed so that the top transducer 5 is over the 10.sup.th rib,
which is the lowermost fixed rib. This leaves the lower transducer
6 in the vicinity of the 11.sup.th rib, which is floating. Thus,
the transducer 6 is effectively over the abdomen and is not
affected by the ribs. This is described in more detail below.
[0069] Referring to FIG. 6, at a block diagram level the sensor 1
comprises the transducers 5 and 6 feeding a filtering circuit 20,
and an ADC (not shown) feeds a Bluetooth module 24. There is a DSP
processor 21 also linked with the Bluetooth module 25. A battery
management circuit 22 is linked with the battery 13, and there
battery charging terminal 23. The controller 3 houses an
accelerometer 25, and there are LED and alarm sounder output
devices 26 and 27.
[0070] The two transducers 5 and 6 are of equal length, width,
thickness, and composition. They are positioned 30.degree. apart
from one another about a single point of common placement which
ensures a preferred form factor. This preferred configuration is
not the only configuration at which this invention will be
effective. The angle between each transducer can be different and
indeed they may be parallel. However the preferred range is
25.degree. to 55.degree., and the most preferred is in the region
of 27.degree. to 33.degree.. The preferred length and width of each
transducer is in the range of 30 mm to 50 mm and 50-400 .mu.m
thick.
[0071] The transducers 5 and 6 provide the deformation information
as described below to allow the processor 21 to automatically
generate an output indicating patient respiration. However, the
accelerometer 25 allows improved effectiveness in analysing signals
arising from wearer's activity and posture. Such variables of
posture and activity have direct influence upon the effectiveness
of the system. The system can also identify how quickly the human
subject is moving, and the subject's posture and when movement
based artefacts have been induced in the strain transducer signal.
This further enables the human subject to live a normal functional
life while the device comprehensively measures the respiratory
performance without imposing
[0072] The sensor 1 may be positioned for example over the 9.sup.th
to 11.sup.th rib, with the controller 3 approximately situated
under the subject's arm. The vertical position is determined with
reference to the subject's 10.sup.th rib, with the transducer 5
being preferably situated on or just below the 10.sup.th rib and in
line with this rib. The transducer 6 would therefore be adhered to
the subject's abdomen. The transducer 6 is preferably horizontal,
but subject physiology may require the transducer 6 to be placed at
an angle. The apex of the angle should point towards the rear of
the subject. FIG. 5 shows the sensor 1 mounted on the patient's
skin at one side. However, the sensor could alternatively be on the
opposite side in a mirror-image fashion.
[0073] The transducer 5 is particularly responsive to a distending
movement of the rib cage, forwardly and laterally. This is almost
entirely due to respiration. There may also be pivoting out of the
plane of the page in FIG. 5, primarily due to motion artefacts such
a walking. Importantly, the transducer 5 is approximately equally
responsive to rib distending and motion artefacts, whereas the
transducer 6 is less responsive to rib distending and equally
responsive to the motion artefacts. When the subject changes their
posture, and/or begins breathing under a different regime (normally
chest breathing or diaphragm breathing) the signal expressed on the
transducers 5 and 6 can change greatly. Typically the transducer 5
which is resting on the rib responds with greater magnitude when
the subject is upright and/or breathing mostly using chest
movements. When the subject is lying and/or diaphragm breathing the
transducer 6, resting on the abdomen, typically responds more
strongly. In atypical cases, for instance when the subject is
breathing heavily using the ribs, the respiratory response from a
transducer can be of such small magnitude as to be
indistinguishable from background noise. In this event, the data
from this transducer or 6 is discarded, and the other transducer is
used solely to derive the respiratory rate.
[0074] Different subjects show different signals on transducers for
the same posture due tai emphasis on gut or rib breathing, and
variations in placement. It is not possible to guarantee the
patient's position with transducers. The accelerometer 25 helps to
determine the orientation of the patient, and the processor
compensates the transducer outputs according to information from
the accelerometer 25.
[0075] The system may be used for monitoring respiratory
performances in a clinical environment, or alternatively in a
non-clinical environment such as physical exercise monitoring for
sports performance enhancement.
[0076] The system may be used for the monitoring of apnea events in
sleeping subjects. Small configuration changes to the sensor will
allow for apnea monitoring. Examples of such alterations include
algorithm emphasis on detecting missing breathing signals, or
modification of the software to produce a waveform for use in
diagnosis by a medical professional.
[0077] Regarding data processing and communication, in one
configuration, the Bluetooth (BT) module 24 is replaced with a
removable hard disk. In another configuration the BT module 24
constantly streams the breathing waveforms, and processing is
carried out on a desktop PC or other computer. In instances where
healthcare professionals wish to monitor the produced signals
directly, limited algorithms can be implemented to clean up the
respiratory signal for presentation.
[0078] A Bluetooth module 24 is used to communicate with an
external device to display the respiratory rate history of the
wearer. To ensure continuity of service, on attachment, the BT
module is renamed with the patient's Medical Record Number (MRN),
for example as scanned from a patient records barcode. The renaming
is temporary and lasts for the duration of the device attachment to
the patient. Upon removal or recharging, the BT module is
automatically renamed to its default identifier. The renaming of
the BT module 24 with the MRN allows any authorised device to
interact with the sensor 1 for the duration it is attached to the
patient.
[0079] In instances where the patient can be assumed to be in a
steady position e.g. short time spent lying down, signals from a
single transducer can suffice to record respiratory rate. However,
the multi-transducer configuration covers the full spectrum of
patient postures and rib/diaphragm breathing.
[0080] In more detail, the signals from both transducers 5 and 6
are filtered and the signal is processed to extrapolate the true
wanted signal. This arrangement achieves both filtering and
analytical processing capability at the point of measurement. It
achieves this with very little restriction in patient movement.
Also, some of the components, such as the signal conditioning
circuits 20 and the processor 21 are local on the sensor 1. Such a
sensor can also be more robust in terms of its application to
different physiological parameters e.g. body mass index, body
position, location, activity and/or similar parameters. The
inclusion of the accelerometer 25 in the device allows such well
known art as fall detection, step detection and orientation
monitoring to be easily incorporated into the sensor 1. The
preferred location for an accelerometer is in the reusable
electronic circuitry unit, preferably integrated into the
processing circuit 16. The exact placement of the accelerometer is
of little importance, as the accelerometer is used to detect gross
movement of the subject's body.
[0081] The sensor 1 does not have electrical wires which might
interfere with the patient. Also, the sensor 1 has a low-profile
construction so as riot to interfere with the natural movement of
the arms of the patient, with an ergonomically efficient design.
The sensor is designed to be wearable for a period of up to 8 days.
During this period, the device continuously collects and processes
data from the transducers and when interrogated by the supervising
medical professional report on the subjects respiratory rate over
the proceeding number of hours.
[0082] FIG. 7 shows an instrumentation amplifier which amplifies
the signals arriving from a single transducer for later
processing.
[0083] FIG. 8 shows a difference amplifier followed by a 2.sup.nd
order low pass filter. The difference amplifier removes the
reference voltage from the incoming signals and amplifies the
signal by a gain of one. The low pass filter removes higher
frequency signals from the sensor signal.
[0084] The signal processing of the outputs of the movement
transducers 5 and 6 and the accelerometer 25 is explained in more
detail in FIGS. 9, 10 and 11. The plots of FIGS. 12(a) to 12(e) are
generated at the blocks in FIG. 9 as indicated. Sensor inputs are
connected into the microprocessor where all digital analytics are
calculated. At this stage digital signal conditioning is performed.
This is required supplementary to analogue filtering so as to
reduce the effect of abasing and spurious noise. Filtering in the
digital domain provides a richer and more versatile filtering
process than what can be achieved in the analogue domain.
[0085] Once acquired, the incoming signals are processed to
calculate the respiratory rate of the subject over a given time
period. Several main algorithm steps are used for the reliable
calculation of rates in the presence of movement or other
artefacts; signal conditioning, artefact detection, artefact
resolution, respiration rate derivation, as well as other
miscellaneous supporting algorithms. Rate detection algorithms were
noted to fall into two main categories; time domain analysis and
frequency domain analysis. Time domain analysis includes techniques
such as peak and trough detection, template matching and machine
learning. Frequency domain analysis includes techniques such as the
discrete Fourier transform, wavelet analysis and auto- and
cross-correlation techniques. Algorithms can include inputs from
the on-board accelerometer or gyroscope.
[0086] One implementation of an analysis algorithm is outlined in
FIG. 9. This implementation is given by way of example only, and
does not limit the invention to the use of other algorithms, or
sub-algorithms. This algorithm is optimised for low power
consumption and uses the accelerometer 25 data in addition to the
deformation transducers 5 and 6 to derive a clean, conditioned
respiratory rate. Signal conditioning is carried out by baseline
subtraction against the input voltage signal. The signal is further
conditioned using an exponential moving average filter to smooth
the signal. When the algorithm is triggered, every 25 s, an
artefact detection protocol is triggered. Artefacts detected on the
piezoelectric transducer signal (for example, signals which are
`railing`, or outside the limits of measurement) are them removed
from the signal by smoothly bringing the signal to zero in these
areas. Two separate respiratory rate algorithms are then run--one a
time domain algorithm and one a frequency domain algorithm. The
first concentrates on looking at the distance between the peaks and
troughs in the respirator waveform and deriving a rate for that.
This is outlined in FIG. 11. The second uses a fast Fourier
transform to extract frequency domain information from the
waveform, shown in FIG. 10. This algorithm also takes the
accelerometer data as a secondary input. Cyclical interference from
the subject or environment, e.g. walking, is compensated for by
extracting frequency information from the accelerometer. Large
movements are also detected and compensated for using the
accelerometer data. Once the rate calculations are made, the
extracted rates are buffered for communication via Bluetooth to an
external tablet PC.
[0087] Signals output from the sensor transducers differ greatly
from subject to subject and when changes in posture or breathing
regime occur. This includes changes in signal strength, changes in
the shape of the repeated breathing pattern, and the relative
strength of the signals from each of the strain transducers. The
implemented algorithm only assumes that the respiratory signal is
represented by a repeating pattern of peaks and troughs at a rate
indicative of the respiratory rate of the subject, as shown in
FIGS. 12 and 13 and 8. The magnitude of the sensor output is
considered only of importance if the signal becomes so large as to
exceed the output limit of the sensor, or so small as to become
indistinguishable from noise. The shape of the repeating signal is
not considered to be indicative of any breathing regime. The
magnitude and shape of the repeating pattern can change greatly
depending on posture, device positioning and subject to subject
variation. For these reasons, embedded algorithms have been
selected so the sensor does not require that the unit be calibrated
for any individual subject.
[0088] FIGS. 12(a) to 12(f) are plots of the main intermediate
calculations from an implemented algorithm to determine the
respiratory rate for a single transducer signal, as shown in FIG.
9. FIG. 12(a) shows the raw output from a single sensor over a 60
second time period showing the peaks and troughs indicative of
normal breathing. One movement artefact can be seen as an increase
in the signal strength at approximately second 35. After the motion
artefact, a short sharp downward artefact can be seen. FIG. 12(b)
shows the same signal with baseline correction and a moving average
smoothing filter applied. The sharp downward spike is removed from
the signal, but the large movement artefact remains. FIG. 12(c)
shows the results from an artefact detection function. Where the
black line is in the higher state, a large non-respiratory artefact
has been determined to have occurred. Other artefact detection
methods may be overlaid on this as required. FIG. 12(d) shows the
same signal with the section designated as artefact smoothly
removed from the waveform. Small downward troughs can be seen
either side of the removed section. This area is flagged for the
following step to ensure it does not interfere with the peak trough
detection algorithm. FIG. 12(e) shows the signal in FIG. 12(d) with
peaks and troughs identified by the system's processor. The area
around the detected artefact is removed from consideration as it is
not an accurate representation of the signal. FIG. 12(f) shows the
frequency spectrum of the signal, with the most prominent signal
highlighted at approximately 0.25 Hz, or 15 breaths per minute. The
smooth removal of the artefact has resulted in a clearly
discernable breathing frequency.
[0089] FIGS. 13(a) and 13(b) are plots showing examples of two
different breathing regimes--rib breathing and diaphragm breathing.
FIG. 13(a) shows the signals from the two transducers when the
subject is standing up and predominantly rib breathing. A short
movement artefact is visible around second 25. The diaphragm signal
is much smaller and less coherent than the rib signal. Artefact
detection will remove the diaphragm signal from consideration due
to low signal strength. FIG. 12(b) shows a signal from the same
subject when the subject is lying on their back and predominantly
diaphragm breathing. In this case the rib signal is of low
magnitude and will be rejected by the algorithm. It is important to
note that these figures show the extremes of rib and diaphragm
breathing and that normal breathing and differences from subject to
subject will have a greater or lesser effect on each sensor.
[0090] FIGS. 14(a) and 14(b) are plots showing the piezoelectric
transducer and accelerometer signals for a subject sleeping over a
period of one hour. Individual breaths are not discernable in FIG.
14(a) at this resolution. The upper and lower magnitudes of the
transducers 5 and 6 are shown as solid lines. Transducer 6 is to
the fore, and transducer 5 is to the rear and shown hatched. The
magnitude of the signals can be seen to change at periods during
the hours. Abrupt shifts in the subject's body position can be seen
in FIG. 14(b) as jumps in the accelerometer signals, here moving
from lying on the hack, to lying on the side and then back again
for the last 20 minutes. The subject can be seen to be nominally
still for periods of up to 15 minutes between these movements.
[0091] The transducers transport the change in voltage through
electrical contacts which have leads connecting the contacts of
each movement transducer to the input electric contacts of the
filter circuitry. Filtering circuitry is integrated on a printed
circuit board upon which the amplifiers and the processor unit
reside. All transportation of the signal from the filter
pre-transmission is done on the PCB.
[0092] The processor 14 and/or other devices such as GPRS and
Bluetooth radio respiratory sensor is stacked on top of the sensor
element which is on the body. This is secured mechanically and
offers easy connection and removal while ensuring a strong electric
connection between both parts.
[0093] The preferred relative positions e senor as shown in FIG. 5
on the 10.sup.th rib is preferred, although the 7.sup.th and
8.sup.th may alternatively be used. This placement is the preferred
location to utilise the mechanics of respiration. The ribcage, at
the denoted location, is more flexible and subject to the largest
deformations during respiratory effort. Where all parts of the
thoracic region undergo fixed loci of displacement, the magnitude
of displacement is relative to individual locations. The
fundamental function of these mechanics is to create a vacuum
within the thoracic cavity thus creating negative airspace to draw
air into the lungs via nose or mouth. This is undertaken through
two modes of operation which can be largely mutually exclusive.
[0094] The distending first operation triggers an involuntary
contraction of the muscles around the ribcage, causing the rib cage
to lift up. As the rib cage lifts up, it creates an increased
internal volume in the thoracic cavity. This increase in volume
also creates a vacuum. Air flows from positive pressure into
negative pressure. Thus, air flows into the mouth and nose of a
human subject and causes respiration to begin. Air is then pushed
out by the muscles around the ribcage while relaxing, thus
decreasing the internal volume of said cavity and pushing air out
of the body. This is also aided by the diaphragm as it maintains a
positive pressure upon the base of the lungs. This diaphragm is a
muscle which divides the thoracic region from the abdominal
region.
[0095] A distending second operation involves an increase in the
internal volume of the abdomen region, which causes a negative
pressure and thus draws down the diaphragm. By causing this, the
internal volume of the thoracic region increases, thus creating a
vacuum and drawing air in. Air is expelled when the volume of the
abdomen cavity is decreased and the diaphragm is again pushed up
against the lungs, decreasing the volume of the thoracic region and
expelling air out.
[0096] The effect of the two operations attributed to respiratory
effort is seen across the thoracic and abdominal region. It is
effective to measure respiration at any location using the
methodology as outlined by this invention of a plurality of sensors
in a set configuration. However the preferred location as outlined
in this invention is the most efficient area of measurement.
[0097] These two operations can act independently if negligible rib
cage movement is ignored. More often these operations occur in
parallel. Thus, to be able to measure both the thoracic and
abdominal displacement in a single location is a significant
advantage.
[0098] Further to the need to detect respiratory rate, the device
can also detect with high accuracy the moments of inhalation and
exhalation as show in FIG. 12(e), and the duration of each. Such
application is highly sought after when monitoring lung capacity to
ensure lung damage from over-pressurising the volume is not
exceeded during invasive and/or non-invasive ventilation.
[0099] In embodiments which have one or more accelerometers, these
are used to detect when movements occur and this information may be
used to smooth or remove artefacts from the strain transducer
signals. Artefact correction is applied to the strain transducer
signal, and the processor does not assume that all artefacts are
accounted for on the accelerometer--arm movements, direct contact
with sensors etc. Also, the processor may use accelerometer
orientation to weigh the relative usefulness of the two strain
transducers (e.g. weight in favour of abdomen sensor when patient
is lying down.
[0100] Some of the advantages of the invention may be summarised
as: [0101] (a) Improved accuracy by ensuring a superior method of
sensor application to the wearer which does not require the
wearer's assistance nor require the wearer to be assisted. [0102]
(b) By having both filtering and signal processing at the point of
measurement improves accuracy due to reducing anxiety of the wearer
and promoting, longer continuous use, thereby improving analytics.
[0103] (c) It also reduces any effects of external influences such
as electromagnetic interference from peripheral devices, unlike the
prior art arrangements having lengthy wires promoting noise in the
signal. [0104] (d) Eliminating the majority of unwanted motion
artefacts irrespective of placement within a preferred area of
application. [0105] (e) Reducing the effect of philological
variances such as body mass index, body position, location,
activity and condition again pre-processing to ensure high level of
accuracy. [0106] (f) Having a secure but removable fixing of the
sensor and single construction enables reduction in cross
contamination from device reuse which more efficient utilisation of
higher end electronics. [0107] (g) Having a profile and contour
promotes easier cleaning. [0108] (h) Having profile and contours
that promote patient comfort and reduction from unintentional
interference from moving limbs.
[0109] The invention is not limited to the embodiments described,
but may be varied in construction and detail. For example the
system may additionally include a gyroscope and the processor may
process the gyroscope output by enabling the posture of the body to
be known to the processor, thus enabling anomalies of the
transducers to be accounted for.
* * * * *