Artifact detection method for hemodynamic parameter measurement

Abstract

A device for detecting and classifying artifact-causing events associated with hemodynamic parameter measurements based on analysis of tissue pressure signals indicative of pressure between a surface of a body part of a user and a cuff of a measuring device and actuator operating signals indicative of operating parameters of a pneumatic actuator associated with the cuff .

Description

Technical Field

[0001] The present invention relates to artifact detection in the field of measurement of blood pressure and other hemodynamic parameters. Background Art

[0002] US2021121073A1 relates to a device for use with a wearable cuff in determining blood pressure and / or pulse rate and a method of operating the device.

[0003] Physiological signals provide important input to the clinical decisions of healthcare professionals. However, the accuracy of the information collected depends on the quality of the data recorded.

[0004] Unreliable signals or poor signal quality can lead to incorrect measurements, false alarms, and / or inappropriate clinical decisions.

[0005] One source of measurement error is signal artifacts. Some artifacts may be extremely short-lived, while others may be persistent.

[0006] As a non-limiting example, artifacts within the context of cuff-based blood pressure measurement (and / or other hemodynamic parameters) may have different causes, such as: patient movement; cuff movement during measurement; and measurement technique artifacts due to cuff cracking, where a sudden loosening of the cuff fastener causes a sudden increase in cuff volume and a subsequent decrease in the measured pressure signal.

[0007] In some cases, artifacts may not be noticed by healthcare providers during surgery, which can lead to inaccurate measurements, false data, and potentially inappropriate clinical evaluations. Inappropriate treatment (or inappropriate lack of treatment) may be followed.

[0008] In some other cases, artifacts may only be noticed after the measurement is completed. In these cases, the solution depends on each specific situation. For example, repeated measurements may be required, which adds additional workflow steps and additional time. Additionally, if the actual cause of the artifact is unknown, repeated measurements may result in the same artifact recurring.

[0009] Furthermore, by repeating measurements, the patient's comfort is compromised because hemodynamic measurements take an average of 90 seconds and may be uncomfortable for the patient.

[0010] There is a need for a device for detecting artifacts and preferably for taking or recommending actions to remedy the artifacts. Summary of the Invention

[0011] The invention is defined by the independent claims. The dependent claims define advantageous embodiments.

[0012] According to an example of one aspect of the present invention, there is provided a method for use during measurement of blood pressure and / or other hemodynamic parameters by a hemodynamic parameter measurement device. A hemodynamic parameter measurement device that can be used with the method may include a cuff for wrapping around a part of a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure exerted by the cuff on the body part, and the hemodynamic parameter measurement device may further include a tissue pressure sensor arranged to sense the pressure between the surface of the user's body part and the cuff, and an operating parameter sensing device for sensing operating parameters of the pneumatic actuator. For example, the operating parameter sensor may include an actuation pressure sensor for sensing the pressure within the inflatable bladder. For example, the cuff may be configured such that at least the pressure-sensitive part of the tissue pressure sensor is arranged between the body part and the cuff during use.

[0013] The method itself includes receiving a set of sensor signals from the hemodynamic parameter measurement device, wherein the set of sensor signals includes: a tissue pressure signal indicating the pressure between the surface of the user's body part and the cuff; and an actuator operation signal indicating operating parameters of the pneumatic actuator. The method further includes processing these two signals using an artifact detection module to detect the occurrence of any one of a predefined set of different types of artifact events based on the combination of the signals, each artifact event type corresponding to a physical event that affects the sensor readings. The method may further include generating an output based on the detected (one or more) artifact events. For example, the method may include generating a report indicating the type of any one or more detected artifact events, and preferably exporting the report as a data item.

[0014] Accordingly, embodiments of the present invention are based on the concept of detecting physical events (such as physical interference) associated with the measurement device or the patient that cause signal artifacts based on an analysis of at least one operating parameter signal applied to the tissue pressure signal and the pneumatic actuator. The use of these two signals for artifact detection has not been proposed before. Through experiments and simulations, the inventors have found that certain physical interference events associated with the device or the patient result in detectable features within the waveforms of the two previously mentioned signals, and thus provide a way to detect artifacts and possible actions to improve the measurement or at least annotate the measurement with warning markers where artifacts may be present.

[0015] The measurement device can be used to measure blood pressure and / or one or more other hemodynamic parameters, such as cardiac output and / or stroke volume.

[0016] For example, the above-described tissue pressure sensor may include a sensor pad that is arranged to be applied to a body part by a cuff when the cuff is worn. The sensor pad may be a fluid-filled sensor pad. The fluid pad may be in fluid communication with a pressure transducer of the tissue pressure sensor, wherein the pressure transducer is adapted to convert a change in fluid pressure within the fluid pad into an electronic pressure reading. The pressure transducer may be located outside the cuff, and the pad may be arranged as part of the cuff.

[0017] For example, the above-described physical event corresponding to an artifact event may be a physical disturbance of the device (such as a pneumatic supply cuff or tubing, or a cuff or fluid tube connecting the fluid-filled sensor pad included in the tissue pressure sensor to the pressure transducer) or a physical disturbance of the patient (e.g., movement of the patient).

[0018] In some embodiments, the method is performed in real time during the measurement of hemodynamic parameters such as blood pressure measurement. This allows for remedial measures to correct artifacts (such as repeating the measurement) to be taken while the patient is still present.

[0019] In some embodiments, artifact detection is performed based on one or more characteristic features in the waveform of one or more received sensor signals. This can be done using classical signal processing techniques. Additionally or alternatively, machine learning algorithms may also be used to detect the occurrence of one or more artifact events based on an input including the set of sensor signals. This can be achieved by appropriately training the algorithm using training data that includes sample sensor signal waveforms annotated according to whether the signal waveform is associated with a particular one or more artifacts.

[0020] In some embodiments, artifact detection includes detecting motion artifacts based on: high-pass filtering an actuator operation signal; and detecting power in the high-pass filtered signal that exceeds a predefined threshold. "Power" may mean, for example, the intensity of the AC component of the high-pass filtered signal. For example, this can be calculated as the average of the squares of the high-pass filtered signal samples over a time window. Alternatively, it can be calculated as the average of the absolute values of the high-pass filtered samples over a time window. In either case, the resulting metric indicates the intensity of the AC component of the signal, i.e., the power of the signal.

[0021] In some embodiments, artifact detection includes: detecting motion artifacts based on calculating an autocorrelation metric of a tissue pressure signal; and detecting that the autocorrelation metric drops below a threshold.

[0022] The autocorrelation measure indicates the degree to which the tissue pressure signal is correlated with itself over time.

[0023] In some embodiments, the detection includes detection of motion artifacts based on: a first detection that includes high-pass filtering an actuator operation signal and detecting that the power at any point in time in the high-pass filtered signal exceeds a predefined threshold; and a second detection that includes calculating an autocorrelation metric of the tissue pressure signal and detecting that the autocorrelation metric drops below a threshold at any point in time. Motion artifacts can be detected in the event that the time points corresponding to the first detection and the second detection are within a predefined proximity of each other. The time points can be the same or within a threshold time range of each other.

[0024] In some embodiments, the set of sensor signals further includes position and / or motion sensor signals indicative of the patient's posture and / or motion.

[0025] The method can include, for example, using the position and / or motion sensor signals indicative of the patient's posture and / or motion to detect the occurrence of an artifact event that includes a change in the patient's posture or the patient's motion.

[0026] In some embodiments, the tissue pressure sensor includes a fluid-filled sensor pad that is arranged such that the pressure between the surface of a body part of the user and the cuff is applied to the fluid-filled pad. The fluid-filled sensor pad can be arranged between the actuator of the cuff and the body part of the user in use. The fluid-filled pad can be fluidly connected to a pressure transducer via a fluid line, and wherein the pressure applied to the fluid pad is electronically converted by the pressure transducer into a pressure reading.

[0027] In some embodiments, artifact detection includes detecting an artifact event corresponding to the user touching the fluid line or tube that connects the fluid-filled sensor pad included in the tissue pressure sensor to the pressure transducer. Here, the detection can include: calculating an autocorrelation metric of the tissue pressure signal over a time period; and calculating a noise metric of the actuator operation signal, the noise metric indicating the degree of change of the signal relative to a smooth ramp function over the time period. Touching of the fluid line can be detected in response to the autocorrelation of the tissue pressure signal being below a predefined threshold and the noise metric of the actuator operation signal being below a predefined threshold.

[0028] To explain, if the fluid-filled tube of the tissue pressure measurement system is accidentally touched, this will result in a "ringing pattern" in the tissue pressure signal, although it is not expected to affect the actuator operation parameter signal. The "ringing pattern" can be identified in the tissue pressure signal based on a reduction in the correlation signal, which shows the degree to which the tissue pressure signal correlates with itself over time.

[0029] Alternatively, since the "ringing pattern" depends on the tube length and fluid type, templates or features of an expected "ringing pattern" (e.g., resonant frequency and damping coefficient) can be defined and used to identify such "ringing patterns" in the tissue pressure signal. In the case where such a "ringing pattern" is detected in the tissue pressure signal, an indication that the tube filled with fluid has been touched is obtained while the actuator operation parameter signal remains unaffected.

[0030] Absence of artifacts in the actuator operation signal can be detected by evaluating whether the actuator operation signal takes the form of a ramp signal (i.e., smoothly increasing pressure) that increases over time without any (oscillatory) signal components or inflection points.

[0031] Thus, in some embodiments, artifact detection can include detecting an artifact event corresponding to a user touching a fluid line that connects a sensor pad filled with fluid included in a tissue pressure sensor to a pressure transducer included in the tissue pressure sensor, and wherein the detection includes detecting one or more predefined characteristic waveform patterns within the waveform of the tissue pressure signal.

[0032] In some examples, the detection can also depend on detecting the absence of such waveform patterns in the actuator operation signal and / or detecting the absence of any oscillatory components or inflection points in the actuator operation signal (or the baseline of the operation signal).

[0033] In some embodiments, the cuff includes a plurality of at least partially overlapping layers or layer portions, e.g., slidable relative to each other to change the diameter of the cuff. In an embodiment, artifact detection includes detecting an artifact event corresponding to a stick-slip event associated with the overlapping layers. For example, in some embodiments, the cuff can include a housing structure, wherein the housing structure is arranged to be located between the pneumatic actuator and the body part when the cuff is worn on the body part and is arranged to surround the body part when the cuff is worn. The housing structure can include at least partially overlapping portions, and wherein the at least partially overlapping portions of the housing structure can move or slide relative to each other so as to reduce the diameter of the housing structure in response to an increase in the pressure applied by the pneumatic actuator. Artifact detection can include detecting an artifact event corresponding to a stick-slip event associated with the at least partially overlapping portions.

[0034] In some embodiments, the housing structure can include a single sheet or layer element rolled into a tubular shape, and wherein the circumferential ends of the rolled sheet or layer overlap to form at least partially overlapping (one or more) portions.

[0035] Stick-slip events can be detected based on: detecting an upward inflection point in the tissue pressure signal that exceeds a first predefined gradient threshold, and detecting a downward inflection point in the actuator operation signal that exceeds a second predefined gradient threshold. In particular, the inflection point can be an inflection point in the baseline of the corresponding signal.

[0036] During stick-slip artifacts, there is friction within the housing that prevents the circumference of the housing from gradually decreasing over time. At regular intervals, this friction is released / removed, and the following chain reaction occurs: the circumference of the housing suddenly decreases; the actuator operation parameter (e.g., actuator pressure) suddenly decreases; the tissue pressure suddenly increases.

[0037] In some embodiments, artifact detection includes detecting an artifact event corresponding to slippage or loosening of the cuff fastening device. Detection here can be based on: detecting a downward inflection point in the actuator operation signal that exceeds a first predefined gradient threshold; and detecting a downward inflection point in the tissue pressure signal that exceeds a second predefined gradient threshold. In particular, the inflection point can be an inflection point in the baseline of the corresponding signal.

[0038] In particular, at certain times, the Velcro of the cuff balloon may become loose, so the housing will acquire a (slightly) larger circumference. This is known in the art as the cuff "cracking". Due to the larger circumference, it is expected that: the tissue pressure signal will suddenly decrease; and the actuator operation parameter (such as actuator pressure) will suddenly decrease.

[0039] In some embodiments, the detection further includes detecting a downward inflection point in each of the tissue pressure and actuator operation signals, followed by a signal portion in each corresponding signal that represents a negative offset relative to the corresponding signal before the downward inflection point; and wherein the detection further includes determining whether the duration of the signal portion exceeds a predefined threshold.

[0040] After the artifact, both the tissue pressure and the actuator operation parameter (e.g., actuator pressure) show an additional persistent (negative) offset.

[0041] In some embodiments, the actuator operation parameter sensing device includes an actuation pressure sensor for sensing the pressure within the inflatable balloon. In some embodiments, the actuator operation signal includes an actuator pressure signal indicating the pressure within the inflatable balloon.

[0042] In some embodiments, the actuator operation signal includes an actuator activity signal indicating the pumping power level or pumping rate of the pump of the pneumatic actuator.

[0043] As a more general statement, the actuator operation signal can be a direct signal or a signal indicating the actuation state of the actuator, such as a signal indicating the physical position or inflation state of the actuator.

[0044] In some embodiments, in response to the detection of an artifact, the method includes generating a control signal for output to a hemodynamic parameter measurement device to control the device to abort a current hemodynamic parameter measurement.

[0045] In some embodiments, the method further includes determining a recommended response action for the user to take based on applying a recommendation module to the output of the artifact detection module. In some embodiments, the generated report also indicates the recommended response action for each detected artifact.

[0046] In some embodiments, the recommended response action is to repeat the hemodynamic parameter measurement. Additionally or alternatively, in some embodiments, the recommended response action is to re - adapt (e.g., re - position or re - wrap) the cuff of the hemodynamic parameter measurement device to the user and repeat the hemodynamic parameter measurement. Additionally or alternatively, in some embodiments, the recommended response action is to replace the cuff of the hemodynamic parameter measurement device with a new cuff and repeat the hemodynamic parameter measurement.

[0047] In some embodiments, a report is generated and exported at the end of a complete measurement cycle of the hemodynamic parameter measurement device, or immediately when a given artifact event is detected during the hemodynamic parameter measurement cycle.

[0048] Another aspect of the present invention is a computer program product, the computer program product including code units configured to run on a processor operably coupled to a hemodynamic parameter measurement device, the hemodynamic parameter measurement device including a cuff for partially wrapping around a portion of a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure applied by the cuff to the body portion, and the hemodynamic parameter measurement device further includes a tissue pressure sensor and an operating parameter sensing device, the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body portion and the cuff, the operating parameter sensing device for sensing operating parameters of the pneumatic actuator. The code units are configured to cause the processor to perform the method according to any of the embodiments or examples described in this document or according to any of the claims of this application when running. In some embodiments, the operating parameter sensing device may be an actuation pressure sensor for sensing the pressure within the inflatable bladder. In some embodiments, the actuator operation signal includes an actuator activity signal indicating the pumping power level or pumping rate of the pump of the pneumatic actuator.

[0049] Another aspect of the present invention is a processing unit for a hemodynamic parameter measurement device, comprising: an input / output unit for operably coupling to a hemodynamic parameter measurement device in use, the hemodynamic parameter measurement device comprising a cuff for partially wrapping around a user's body, and wherein the cuff comprises a pneumatic actuator in the form of an inflatable bladder for changing the pressure applied by the cuff to the body part, and the hemodynamic parameter measurement device further comprises a tissue pressure sensor and operating parameter sensing means, the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body part and the cuff, the operating parameter sensing means for sensing the operating parameters of the pneumatic actuator; and one or more processors configured to execute a method. In some embodiments, the operating parameter sensing means may be an actuator pressure sensor for sensing the pressure within the inflatable bladder. In some embodiments, the actuator operation signal comprises an actuator activity signal indicating the pumping power level or pumping rate of the pump of the pneumatic actuator. The cuff may be configured such that when the user wears the cuff, at least the pressure-sensitive portion of the tissue pressure sensor is arranged between the body part and the actuator. The tissue pressure sensor may comprise a fluid-filled sensing pad fluidly connected to a pressure transducer to read out changes in the pressure applied to the sensor pad. The cuff may be configured such that when the user wears the cuff, the fluid-filled sensor pad is arranged between the actuator of the cuff and the body part.

[0050] The method comprises: receiving, via the input / output unit, a set of sensor signals from the hemodynamic parameter measurement device, wherein the set of sensor signals comprises: a tissue pressure signal indicating the pressure between the surface of the user's body part and the cuff; and an actuator operation signal indicating the operating parameters of the pneumatic actuator. The method further comprises processing the two signals using an artifact detection module to detect the occurrence of any one of a predefined set of different types of artifact events based on the combination of these signals, each artifact event type corresponding to a physical event affecting the sensor readings. The method may further comprise generating an output based on the occurrence of any detected (one or more) artifact events. The method may for example comprise generating a report indicating the type of any one or more detected artifact events and exporting the report as a data item via the input / output unit.

[0051] Another aspect of the present invention is a system. The system includes a processing unit according to any of the embodiments described in the present disclosure, such as the processing unit outlined above. The system further includes a hemodynamic parameter measuring device operably coupled to the processing unit, the hemodynamic parameter measuring device including a cuff for wrapping around a portion of a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure exerted by the cuff on the body portion, and the hemodynamic parameter measuring device further includes a tissue pressure sensor and an actuation operation parameter sensing device, the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body portion and the cuff, the actuation operation parameter sensing device for sensing an operation parameter of the pneumatic actuator, such as an actuator pressure sensor for sensing the pressure within the inflatable bladder or an actuator activity signal indicating the pumping power level or pumping rate of a pump of the pneumatic actuator. The cuff may be configured such that when the user wears the cuff, at least the pressure-sensitive portion of the tissue pressure sensor is arranged between the body portion and the actuator. The tissue pressure sensor may include a fluid-filled sensing pad, the fluid-filled sensing pad being in fluid communication with a pressure transducer to read out changes in the pressure applied to the sensor pad. The cuff may be configured such that when the user wears the cuff, the fluid-filled sensor pad is arranged between the actuator of the cuff and the body portion. These and other aspects of the present invention will be apparent and elucidated with reference to the embodiments described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] For a better understanding of the present invention, and to more clearly show how the present invention may be implemented, reference will now be made, by way of example only, to the accompanying drawings, in which:

[0053] Figure 1 Schematically illustrates an example hemodynamic parameter measuring cuff suitable for a method according to an embodiment of the present invention;

[0054] Figure 2 Outlines the steps of an example method according to one or more embodiments of the present invention;

[0055] Figure 3 Shows a block diagram of an example processing unit and components of a system according to one or more embodiments of the present invention;

[0056] Figure 4 Shows a processing flow according to one or more embodiments of the present invention;

[0057] Figure 5 Illustrates an example recording of a tissue pressure signal and an actuator pressure signal obtained during measurement of hemodynamic parameters using a hemodynamic parameter measuring device;

[0058] Figure 6 Illustrates illustrative characteristic waveform features that allow detection of stick-slip artifacts; and

[0059] Figure 7 Illustrates characteristic waveform features that allow detection of cuff cracking artifacts. Detailed Description

[0060] The present invention will be described with reference to the accompanying drawings.

[0061] It should be understood that the detailed description and specific examples, although indicating exemplary embodiments of the apparatus, systems and methods, are intended for illustrative purposes only and are not intended to limit the scope of the invention. These and other features, aspects and advantages of the apparatus, systems and methods of the present invention will be better understood from the following description, the appended claims and the drawings. It should be understood that these figures are merely schematic and not drawn to scale. It should also be understood that the same reference numerals are used throughout the drawings to indicate the same or similar components.

[0062] The present invention provides an apparatus for detecting and classifying artifact-causing events associated with hemodynamic parameter measurements based on an analysis of tissue pressure signals and actuator operation signals, the tissue pressure signals indicating the pressure between the surface of a body part of a user and a cuff of a measuring device, and the actuator operation signals indicating operation parameters of a pneumatic actuator associated with the cuff.

[0063] As background,[[]] Figure 1 A schematic diagram of an example of a cuff-based hemodynamic parameter measurement device is shown, and embodiments of the present artifact detection method can be advantageously applied using the cuff-based hemodynamic parameter measurement device. The device can be used to detect blood pressure and / or other hemodynamic parameter measurements, such as cardiac output or stroke volume. The figure shows a cuff 2 applied to the upper arm 10 of a patient. The artery 8 of the patient is schematically shown. The device has a different design from the most standard cuff-based blood pressure measurement devices in that it includes a dedicated tissue pressure sensor 4 that is used to hold the dedicated tissue pressure sensor against the skin tissue when the cuff is inflated. The cuff 2 includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure applied by the cuff to the body part 10. The tissue pressure sensor 4 is arranged to sense the pressure between the surface of the user's body part and the cuff. Independent of the tissue pressure sensor, one or more operation parameters of the pneumatic actuator can be sampled. For example, the pressure inside the inflatable bladder can be sensed. An actuator activity signal indicating the pumping power level or pumping rate of the pump of the pneumatic actuator can also be sampled.

[0064] Standard blood pressure measurement cuffs use the airbag air pressure as the only measurement signal for sensing blood pressure. By additionally using the tissue pressure sensor 4, Figure 1The device shown can achieve superior quality results and is also capable of measuring advanced hemodynamic parameters in addition to blood pressure, such as stroke volume, cardiac output, and fluid responsiveness. In particular, in a standard air blood pressure cuff, it is assumed that the damping air destroys >90% of the amplitude and profile of the tissue pressure pulse wave. In contrast, by using a dedicated tissue pressure sensor, Figure 1 the design shown allows for the recording of high-fidelity (HiFi) arterial blood pressure and pulse waveforms.

[0065] The tissue pressure sensor 4 can be implemented as, for example, a fluid-filled bladder that provides a fluid conforming layer between the cuff 2 and the surface of the user's body. In this way, the integrated pneumatic actuator enables the tissue pressure sensor 4 to be hydraulically coupled to the upper arm tissue encapsulated by the rigid housing. When the blood in the artery pulsates, this results in a pressure wave 6 that can be detected by the tissue pressure sensor 4. The tissue pressure sensor 4 can be connected to a pressure transducer via a fluid-filled tube / line. The pressure sensor converts the pressure in the fluid into an electrical signal.

[0066] The operating principle is based on the coupling of the pressure sensor to a body part (such as the arm) in order to transcutaneously record the tissue pressure pulse wave caused by arterial pulsation (such as the brachial artery). In a manner similar to a conventional upper arm air blood pressure cuff, the cuff uses an integrated pneumatic actuator with an elevated clamping pressure to compress the upper arm. However, the actual compression can be achieved via the narrowed diameter of the rigid circumferential housing.

[0067] In some embodiments, the cuff 2 can include a housing structure, wherein the housing structure is arranged to be located between the pneumatic actuator and the body part when the cuff is worn on the body part and is arranged to surround the body part when the cuff is worn. The housing structure can be relatively rigid. Thus, the measurement accuracy can be improved because the pressure in the arm is measured against a relatively rigid support, which prevents the attenuation of the amplitude and shape of the signal. In particular, if the tissue pressure sensor unit is at least partially located between the housing structure and the body part, high accuracy of the signal measured by the tissue pressure sensor unit can be achieved. With this configuration of the tissue pressure sensor unit, the housing structure does not absorb or attenuate the arterial pressure signal.

[0068] The housing structure can include at least partially overlapping portions, and wherein the overlapping portions of the housing structure can move or slide relative to each other, thereby reducing the diameter of the housing structure in response to an increase in the pressure applied by the pneumatic actuator.

[0069] In some embodiments, the housing structure can include a single sheet or layer element rolled into a tubular shape, and wherein the circumferential ends of the rolled sheet or layer overlap to form at least partially overlapping (one or more) portions.

[0070] The above design allows for non-invasive hemodynamic monitoring and is capable of measuring blood pressure as well as cardiac output and other hemodynamic parameters.

[0071] For more extensive details regarding the exemplary hemodynamic parameter measurement device, reference is made to the document EP2953528A1, which describes the cuff design in more detail. The hemodynamic parameter measurement device according to an embodiment of the present invention can be implemented based on the hemodynamic parameter measurement system described in the document EP2953528A1.

[0072] As described above, artifacts may occur in the signals output by the hemodynamic parameter measurement device. This may lead to inaccuracies in any measurement results obtained based on the processing of these signals, for example, different hemodynamic measurement results as well as the blood pressure calculation itself.

[0073] Different artifacts have different causes and require different responses to remedy them. Embodiments of the present invention are based on detecting signal waveform features indicating certain physical interference events of the hemodynamic parameter measurement device or the patient. By detecting these events that cause artifacts, the timing of possible signal artifacts is detected.

[0074] Some artifacts require repeated measurements. Here, knowing the cause of the artifact is helpful because the repeated measurements can be performed in a way that prevents the event causing the artifact from occurring again.

[0075] For some artifacts, when they occur too frequently, the user may consider stopping the measurement and reapplying or replacing the cuff. In such cases, it is helpful if the system can identify such a situation and provide an appropriate recommendation for reapplying or replacing the cuff.

[0076] In some other cases, corrections can be performed to eliminate the artifacts.

[0077] An object of embodiments of the present invention is to at least partially overcome the disadvantages of prior art devices by providing a method for detecting, identifying the type of artifact, and preferably the recommended response action and transmitting it to the user, so that the clinician can determine what remedial measures need to be taken, or for example, whether a correction of the measurement needs to be applied, or whether the artifact needs to be considered in the evaluation of the measurement results.

[0078] In various embodiments of the present invention, we propose a method for detecting and managing artifacts that may occur in prior art hemodynamic parameter measurement devices during the measurement of blood pressure (BP), stroke volume (SV), cardiac output (CO), fluid responsiveness (FRP), and other hemodynamic (HDM) parameters.

[0079] Figure 2The steps of an example method in accordance with one or more embodiments are outlined in block diagram form. Before further explaining in the form of example embodiments, the steps will be narratively summarized.

[0080] The method is a computer-implemented method used during the measurement of hemodynamic parameters by a hemodynamic parameter measurement device, wherein the hemodynamic parameter measurement device includes a cuff that is partially wrapped around a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure exerted by the cuff on the body part, and the hemodynamic parameter measurement device further includes a tissue pressure sensor and an operating parameter sensing device, the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body part and the cuff, and the operating parameter sensing device for sensing the operating parameters of the pneumatic actuator.

[0081] The method includes receiving a set of sensor signals from the hemodynamic parameter measurement device. This includes receiving a tissue pressure signal indicative of the pressure between the surface of the user's body part and the cuff. It also includes receiving an actuator operation signal indicative of the operating parameters of the pneumatic actuator. The method further includes processing these two signals using an artifact detection module to detect the occurrence of any one of a predefined set of different types of artifact events based on the combination of the signals, each artifact event type corresponding to a physical event that affects the sensor readings. The method preferably may also include generating an output indicative of the occurrence of any one of the predefined set of different types of artifact events detected. For example, the method may include generating a report indicative of the type of any one or more detected artifact events and preferably exporting the report as a data item.

[0082] As described above, the method 10 may also be embodied in hardware form (e.g., in the form of a processing unit) configured to execute the method according to any example or embodiment described in this document or according to any claim of this application.

[0083] For further assistance in understanding, Figure 3 A schematic diagram of an example processing unit 32 configured to execute the method according to one or more embodiments of the present invention is presented. The processing unit is shown in the context of a system 30 that includes the processing unit. The processing unit 32 represents one aspect of the present invention alone. The system 30 is another aspect of the present invention. The provided system need not include all of the shown hardware components; it may include only a subset thereof.

[0084] The processing unit 32 includes one or more processors 36 configured to execute a method according to the method outlined above or according to any embodiment described in this document or any claim of this application. In the illustrated example, the processing unit further includes an input / output unit 34 or a communication interface.

[0085] In Figure 3 the illustrated example, the system 30 further includes a hemodynamic parameter measurement device 42, the hemodynamic parameter measurement device including a tissue pressure sensor 44 and an operating parameter sensing device 46, the tissue pressure sensor being arranged to sense the pressure between the surface of a body part of a user and a cuff of the device, the operating parameter sensing device for sensing operating parameters of a pneumatic actuator. For example, the hemodynamic parameter measurement device 42 may include a cuff for wrapping around a portion of a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable balloon for changing the pressure applied by the cuff to the body part. In some embodiments, the operating parameter sensing device may at least include an actuation pressure sensor for sensing the pressure within the inflatable balloon. In some embodiments, the operating parameter sensing device may include an actuator activity signal indicating the pumping power level or pumping rate of a pump of the pneumatic actuator.

[0086] The system 30 may further include a memory 38 for storing computer program code (e.g., computer-readable code) configured to cause one or more processors 36 of the processing unit 32 to execute a method as described above or according to any embodiment described in this disclosure or according to any claim.

[0087] Figure 4 A flowchart outlining a basic processing flow according to one or more embodiments is shown. As described above, the input 52 to the processing operation may include a tissue pressure signal from a tissue pressure sensor, an actuator operating parameter signal (such as actuation pressure or pumping power level or pumping rate of a pump), and optionally a motion or position signal (such as acquired via, for example, an accelerometer or gyroscope coupled to the patient) for indicating the motion or posture of the patient. These inputs are provided to an artifact detection module 54, which analyzes one or more signals to detect artifact features in the signals and classifies the type of artifact-causing event that caused the artifact features present in one or more of the input signals. The output 56 is the type of detected artifact event plus optionally a recommendation to the user of the best subsequent action to remedy or mitigate the artifact. By way of example, the recommended subsequent action may be at least any one or combination of the following actions: repeat the measurement; reapply or replace the cuff and repeat the measurement; retain the measurement but mark the measurement as potentially unreliable due to the artifact; apply model-based correction of the artifact.

[0088] Regarding the artifact detection module 54, this can include one or more algorithms configured to process a received set of sensor signals and generate an output detection of one or more artifact events of different types. Each of the one or more algorithms can be adapted to perform signal processing to detect one or more characteristic features in the waveform of one or more of the received sensor signals. The module is operable to perform multiple detection algorithms, each algorithm dedicated to detecting a specific type of artifact event. By applying multiple algorithms, the presence of each type among the possible types of artifacts in the set of signals is examined. Additionally or alternatively, in some embodiments, the artifact detection module is operable to perform multiple detection modules, each detection module configured to detect a respective one or more characteristic features in the waveform of one or more of the received sensor signals. The outputs of these detection modules can then be used to detect the presence of one or more specific types of artifact events, where each specific type of artifact event is characterized by a specific combination of the presence of different possible characteristic waveform features.

[0089] As a further example, the artifact detection module can employ one or more machine learning algorithms that have been pre-trained to detect one or more types of artifact events based on the input signal. Similarly, in some cases, there can be multiple machine learning algorithms, each machine learning algorithm being trained to detect a specific type of artifact event based on a specific signal combination.

[0090] Now, how the input set of sensor signals can be processed to detect specific types of artifacts will be described in more detail.

[0091] In particular, the artifact detection module 54 can be configured to automatically detect and classify one or more of the following artifacts:

[0092] Motion artifacts, which correspond to movement of the patient or a change in the patient's posture or position;

[0093] Artifact events that correspond to a user touching the fluid line of the hemodynamic parameter measurement device, the fluid line connecting the sensor pad filled with fluid of the pressure sensing device to the pressure transducer;

[0094] Stick-slip events, which are related to the overlapping portions or layers of the hemodynamic parameter measurement cuff (which will be explained in more detail below);

[0095] "Cuff cracking" events, which correspond to loosening or slipping of the cuff fastening device.

[0096] First, the apparatus for detecting patient motion artifact events will be described.

[0097] Movement can be identified by detecting specific pattern features in the waveforms of tissue pressure signals or actuator operation parameter signals. As described above, the tissue pressure signal indicates the pressure between the surface of a user's body part and the cuff, and the tissue pressure signal is measured by a tissue pressure sensor included in the hemodynamic parameter measurement device. As described above, the actuator operation signal indicates the operation parameters of the pneumatic actuator.

[0098] For the purposes of this example, an operation signal in the form of an actuator pressure signal will be used, which indicates the pressure within the inflatable bladder of the pneumatic actuator. However, as a further example, an operation signal in the form of an actuator activity signal can be used, which indicates the pumping power level or pumping rate of the pump of the pneumatic actuator.

[0099] Movement can be detected by detecting irregular patterns in the tissue pressure or actuator pressure signal. As Figure 5 shown, both the actuator pressure signal (AP) 62 and the tissue pressure signal (TP) 64 include a baseline ramp function that increases pressure during the measurement (due to cuff inflation). Superimposed on it, the TP signal has high-frequency oscillations corresponding to the cardiac pulse.

[0100] If, for example, the signal power level in the high-pass filtered AP signal exceeds a predefined threshold, an irregularity in the AP signal indicating movement can be detected. In other words, artifact detection can include detecting movement artifacts based on: high-pass filtering the actuator operation signal; and, detecting the signal power level in the high-pass filtered signal that exceeds a predefined threshold. By signal power level is meant the intensity of the AC component of the high-pass filtered signal. This can be an instantaneous power level. Alternatively, it can be calculated, for example, as the mean of the squares of the high-pass filtered samples over a time window. Alternatively, it can be calculated as the mean of the absolute values of the high-pass filtered samples over a time window. In either case, the resulting metric indicates the intensity of the AC component of the signal over the relevant window.

[0101] Regarding the threshold, this can of course be preset based on the specific specifications of the system. For example, this can be programmed during manufacturing based on empirical testing.

[0102] Additionally or alternatively, an irregularity in the TP signal indicating movement can be detected based on a reduction in the correlation signal, which shows the degree to which the TP signal is correlated with itself over time. In other words, artifact detection can include: detecting movement artifacts based on calculating an autocorrelation metric of the tissue pressure signal; and, detecting that the autocorrelation metric drops below a predefined threshold.

[0103] Expected motion events occur at similar times in both the AP and TP signals. Thus, confirmation of motion artifacts can be achieved by checking for the occurrence of motion detection at similar times in the AP and TP signals. In this way, the presence of motion artifacts in the AP and TP signals can be detected.

[0104] In other words, the detection can include detecting motion artifacts based on: a first detection that includes high-pass filtering the actuator operation signal and detecting a signal power level of the high-pass filtered signal that exceeds a predefined threshold at any time point; and a second detection that includes calculating an autocorrelation measurement of the tissue pressure signal and detecting that the autocorrelation metric drops below a predefined threshold at any time point; and wherein a motion artifact is detected in an event where the time point corresponding to the first detection and the time point corresponding to the second detection are within a predefined proximity of each other. The time points can be the same or within a threshold time range of each other.

[0105] Additionally, in some embodiments, position and / or motion sensors can be included as part of system 30, the position and / or motion sensors being arranged to detect movement of the patient or a change in the patient's attitude or posture. For example, the position and / or motion sensors can include one or more accelerometers attached to the patient's body, such as a triaxial accelerometer. Using such an accelerometer, changes in the projection of the gravitational components on the three axes of the accelerometer can be detected, and this can be used to detect a change in the patient's attitude or posture. Such a change of course indicates movement of the patient, which can result in signal artifacts. A triaxial accelerometer is not required. Any accelerometer can be used to detect motion events associated with the patient. For example, additionally or alternatively, the AC signal power in the motion signal can be detected and used to detect patient movement. For example, if the AC signal power in the motion signal exceeds a predefined threshold, this is an indication of motion.

[0106] As described above, in some embodiments, the tissue pressure sensor can include a compressible fluid pad arranged as part of a cuff, and wherein the fluid-filled pad is fluidly coupled to a pressure transducer to convert the pressure physically applied to the pad (resulting in a change in the internal fluid pressure of the pad) into an electronic pressure readout. In this context, another potential event that can physically interfere with the measurement of hemodynamic parameters and result in signal artifacts is whether the patient or clinician touches the fluid tube or line (which connects the fluid-filled sensor pad to the pressure transducer). If the fluid-filled tube of the tissue pressure measurement system is accidentally touched, this will result in a "ringing pattern" in the TP signal and is not expected to affect the AP signal.

[0107] A "ringing pattern" can be identified in the TP signal based on a reduction in a correlation signal that shows the degree to which the TP signal is correlated with itself over time. Thus, artifact detection can include detecting an artifact event corresponding to a user touching a fluid line of a hemodynamic parameter measurement device, and wherein the detection includes: calculating an autocorrelation metric of the TP signal over a time period; and calculating a noise metric of the actuator operation signal, the noise metric indicating the degree to which the signal varies from a smooth ramp function over the time period; and wherein the touch is detected in response to the autocorrelation of the tissue pressure signal being below a predefined threshold and the noise metric of the actuator operation signal being below a predefined threshold.

[0108] Additionally or alternatively, the detection can include detecting one or more predefined characteristic waveform patterns within the waveform of the tissue pressure signal. In particular, since the "ringing pattern" depends on tube length, tube diameter, and fluid type, it can be characterized according to an expected range for one or more oscillatory properties such as resonant frequency and damping coefficient. Thus, a template or characteristic waveform pattern for the expected "ringing pattern" (e.g., in terms of expected resonant frequency and damping coefficient) can be predefined and used to identify such a "ringing pattern" in the TP signal.

[0109] In the case where such a "ringing pattern" is detected in the TP signal, an indication that the fluid-filled tube has been touched is obtained. Since it is expected that the AP remains unaffected, in some embodiments, the detection can also depend on detecting the absence of the waveform pattern in the actuator operation signal and / or detecting the absence of any oscillatory components in the actuator operation signal.

[0110] Another class of artifact events are those that represent physical disturbances in the state of the cuff itself.

[0111] As a first example, a hemodynamic parameter measurement cuff can include a device with at least partially overlapping layers or layer portions. In some embodiments, the cuff includes a plurality of overlapping layers or layer portions that are, for example, slidable relative to each other to change the diameter of the cuff. For example, in some embodiments, the cuff can include a housing structure that is arranged to be located between the pneumatic actuator and the body part when the cuff is worn on the body part and is arranged to surround the body part when the cuff is worn. The housing structure can include an overlapping portion, and wherein the overlapping portion of the housing structure can move or slide relative to each other, thereby reducing the diameter of the housing structure in response to an increase in the pressure applied by the pneumatic actuator. In some embodiments, the housing can include a single sheet or layer element that is rolled into a tubular shape, and wherein the circumferential ends of the rolled sheet or layer overlap to form the overlapping portion.

[0112] Stick-slip events may occur between these one or more overlapping layers or portions. During a stick-slip artifact, there is friction within the housing that prevents the circumference of the housing from gradually decreasing over time. At regular intervals, this friction is "released / removed" and the following chain reaction occurs: the cuff circumference suddenly decreases; the actuator pressure signal (AP) suddenly decreases; the tissue pressure signal (TP) suddenly increases. Artifact detection may include detecting an artifact event corresponding to a stick-slip event associated with the overlapping portion.

[0113] Figure 6 Examples of the waveform characteristics of each of the AP 62 and TP 64 signals corresponding to a severe stick-slip event are shown. For illustration, in fact, for purposes of illustration, Figure 6 A series of stick-slip events is shown. Each event is characterized by an upward step in the TP signal 64, or in other words, the upward inflection point of the tissue pressure signal exceeds a first predefined gradient threshold. The event is characterized by a downward step in the AP signal 62, or in other words, the downward inflection point in the actuator operation signal exceeds a second predefined gradient threshold. In particular, the inflection point or step may be an inflection point or step in the baseline of the corresponding signal. Note that Figure 6 The specific example shown represents a particularly severe artifact event and thus shows a somewhat larger artifact size than might typically be expected in most cases, and also shows a repeated series of artifact events. In the case of less severe artifacts, the stick-slip events may appear as having a slightly shorter plateau in the tissue pressure signal 64 and / or a slightly smaller step size in the tissue pressure signal 64.

[0114] Thus, in the case where the cuff includes a plurality of at least partially overlapping or stacked material layers, detection of an artifact event corresponding to a stick-slip event associated with the layers may be performed, and wherein the stick-slip event is detected based on: detecting an upward inflection point in the tissue pressure signal (TP) that exceeds a first predefined gradient threshold, and detecting a downward inflection point in the actuator pressure signal (AP) that exceeds a second predefined gradient threshold.

[0115] Furthermore, it is expected that these inflection points will occur at similar times. Thus, detection can be further confirmed by: first detecting the occurrence of both of the above signal characteristics in the TP and AP signals, and then determining whether they occur within a predefined time window of each other.

[0116] For example, the signal characteristics themselves (e.g., inflection points) can be detected by using the time derivative of each of the TP and AP signals.

[0117] Another type of artifact related to interference from the measurement cuff is the occurrence of so-called cuff cracking. This is a term in the art and is familiar to those skilled in the art. Cuff cracking corresponds to the sudden loosening or slipping of the fastening device used to hold the cuff in the wrapped position around the body part. Most commonly, such a fastening device takes the form of a hook-and-loop type fastener, e.g., Velcro. At some point, the Velcro (or other fastener) of the cuff will come loose, so that the cuff acquires a slightly larger circumference. Due to the larger circumference, it is expected to observe: a sudden decrease in tissue pressure and a sudden decrease in actuator pressure. This can be detected by identifying the inflection points in these signals. In particular, the inflection point can be an inflection point in the baseline of the corresponding signal.

[0118] In other words, artifact detection can include detecting an artifact event corresponding to the loosening or slipping of the cuff fastening device, and wherein said detection is achieved based on: detecting a downward inflection point in the actuator operation signal exceeding a first predefined gradient threshold; and detecting a downward inflection point in the tissue pressure signal exceeding a second predefined gradient threshold.

[0119] For illustration, Figure 7 illustrates the appearance of the signal waveform characteristics corresponding to cuff cracking in the tissue pressure signal 64. Under normal circumstances, during the measurement of hemodynamic parameters, as the pneumatic actuator pressure increases (cuff tightening), the circumference of the cuff gradually decreases. However, when cuff cracking occurs, both the tissue pressure and the actuator pressure suddenly turn downward.

[0120] In addition, after the artifact, both the tissue pressure and the actuator pressure signals show an additional persistent (negative) offset. This is also visible in Figure 7 wherein a negative "jump" can be observed in the TP signal 64, which results in a persistent negative offset. Therefore, another way to detect cuff cracking is to analyze the persistent negative jumps in the TP and AP signals. In other words, the detection can also include detecting the downward inflection points in each of the tissue pressure and actuator operation signals (as described above), followed by the signal portion of each corresponding signal, the signal portion representing a negative offset relative to the corresponding signal before the downward inflection point; and wherein said detection further includes determining whether the duration of the signal portion exceeds a predefined threshold.

[0121] According to one or more embodiments, the method may include a single processing operation applied to tissue pressure signals and actuator operation signals, which is configured to detect any of a plurality of different artifacts. This may involve the application of a series of detection modules, each configured to detect a particular one of a set of characteristic waveform features, and wherein different waveform features in the set of waveform features are characteristic of different possible artifacts. Some waveform features may be characteristic of more than one type of artifact. Thus, the process effectively potentially detects multiple different types of artifacts in a single processing operation.

[0122] As an illustrative example, the processing operation may be implemented generally as follows.

[0123] One component of the processing operation may include analyzing the AC signal power level of the actuator pressure signal to detect one or more waveform features in the actuator pressure signal.

[0124] Another component of the processing operation may include detecting a negative drop in the actuator operation parameter signal.

[0125] Another component of the processing operation may include detecting a positive spike in the actuator operation parameter signal.

[0126] Another component of the processing operation may include detecting a positive pressure spike in the tissue pressure signal.

[0127] Another component of the processing operation may include detecting a negative pressure drop in the tissue pressure signal.

[0128] Another component of the processing operation may include detecting a ringing transient in the tissue pressure signal.

[0129] Another component of the processing operation may include calculating the autocorrelation (or auto-correlation) of the AC component of the tissue pressure signal and detecting a decrease in the autocorrelation. In cases where a motion or position sensor is included as part of the hemodynamic parameter measurement device (e.g., an accelerometer) to detect user motion or posture, another component of the processing operation may include analyzing the motion / posture signal.

[0130] By combining and analyzing the outputs of all the individual processing operation components, the type(s) of the artifact(s) present can be classified based on knowledge of a particular combination of waveform features associated with the occurrence of different artifact events.

[0131] As briefly mentioned above, optionally, the method may further include generating a recommendation for a response action for each or at least a subset of any detected artifact events.

[0132] For example, recommendations for further actions on how to overcome or avoid the repeatedly detected artifacts and thereby improve the measurement of hemodynamic parameters can be generated. The output of this step will be recommendations for further actions, which can be communicated to the user via a user interface device (such as by means of a display and / or an acoustic output device). For example, this can be the display of a patient monitor to which the hemodynamic parameter measurement device is communicatively coupled, or a dedicated display of the hemodynamic parameter measurement device. For example, the display output can provide a report indicating the type of each detected artifact and the response actions recommended for each detected artifact.

[0133] In the case of an artifact event in the form of a patient movement event being detected, the recommended response action can be to repeat the hemodynamic parameter measurement. This can be communicated to the clinician user. In some embodiments, feedback is additionally provided, which advises the clinician user to pay attention to checking that the patient does not move during the repeat measurement.

[0134] In the case of an artifact event in the form of the user touching the fluid tube / line of the hemodynamic parameter measurement cuff being detected, the recommended response action can likewise be to repeat the measurement. In some embodiments, feedback is additionally provided, which advises the clinician user to pay attention to checking that the patient or the clinician user does not touch the fluid-filled tube during the measurement.

[0135] In the case of an artifact event in the form of a stick-slip event associated with the material layer of the hemodynamic parameter measurement cuff being detected, the recommended response action can likewise be to repeat the measurement. In some embodiments, the method can further include tracking the recurrence of stick-slip events on a predefined number of measurements during a single measurement session, where the measurement session is a session in which the hemodynamic parameter measurement cuff is continuously applied to a single patient. If the number of stick-slip events during the measurement session exceeds a predefined threshold, this can be reported to the user via the user interface. A recommendation to rewind and / or replace the cuff can also be communicated to the user.

[0136] In the case of an artifact event in the form of cuff cracking (sliding or loosening of the cuff fastening device) being detected, the recommended response action can be to repeat the measurement. Additionally, in some embodiments, the method can further include tracking the recurrence of loosening or sliding (cuff cracking) events on a predefined number of measurements during a single measurement session (the measurement session is defined as a session in which the hemodynamic parameter measurement cuff is continuously applied to a single patient). If the number of cuff cracking occurrences during the measurement session exceeds a predefined threshold, this can be reported to the user via the user interface, and a recommendation to rewind the cuff of the hemodynamic parameter measurement device onto the patient and repeat the hemodynamic parameter measurement, or to replace the cuff of the hemodynamic parameter measurement device with a new cuff and repeat the hemodynamic parameter measurement, is further transmitted to the user interface.

[0137] In some embodiments, the recommendation for repeated measurements may depend on the amplitude or severity of the artifact exceeding a predefined threshold. For example, in the case of a cuff cracking event, the amplitude of the waveform features corresponding to the cuff cracking event can be determined, and a recommended response action for repeated measurements is generated only when this amplitude exceeds the threshold. The same method can be applied to any other type of artifact event mentioned in the present disclosure. In some cases, different recommendations can be issued based on the amplitude or severity of the artifact exceeding different thresholds. For example, in the case of cuff cracking, there may be a higher threshold for the amplitude of the cuff cracking event, and wherein when this higher threshold is breached, a recommended response action is communicated to the user, recommending re - fitting the hemodynamic parameter measurement cuff or replacing the hemodynamic parameter measurement cuff and repeating the measurement.

[0138] Alternatively or additionally, in some embodiments, a correction can be applied to the sensor signal and / or blood pressure or other hemodynamic measurements to correct the detected artifact.

[0139] In the case of cuff cracking, the correction can be performed based on the processing of the acquired tissue pressure signal 64. As described above, cuff cracking causes a drop in the acquired actuator pressure signal (AP), followed by a drop in the tissue pressure signal (TP). In the absence of any artifacts, the ideal signal waveform takes the form of a uniformly increasing (ramp) actuator pressure baseline, where the tissue pressure baseline also increases due to the increase in actuator pressure. When cuff cracking occurs, this regular ramp of actuator pressure is disrupted, and both the AP and TP signals exhibit a downward inflection in their respective baselines. The effective result is that for a given segment of the applied (actuator) pressure ramp, the tissue pressure signal includes more than one measured signal portion. If the tissue pressure signal is fed into the measurement algorithm without any modification, this disrupts the resulting hemodynamic measurement. The solution is that the specific tissue pressure signal intervals obtained for replicated actuator pressure intervals can be averaged in order to obtain a sequence of single corrected tissue pressure pulses in a monotonically increasing order across the entire actuator pressure range. The corrected tissue pressure signal no longer contains a drop in the average or baseline tissue pressure level. Existing algorithms for a given hemodynamic measurement can be directly applied to the corrected tissue pressure signal in order to determine blood pressure and / or other hemodynamic parameters.

[0140] The averaging of the replicated tissue pressure intervals can be achieved using a weighted average, where an equal weight of 0.5 is applied to each replicated tissue pressure interval (assuming each tissue pressure signal interval is replicated only once).

[0141] Alternatively, the weighting of the replicated tissue pressure intervals can be a function of each signal-to-noise ratio (SNR) in the individual tissue pressure intervals, as follows.

[0142] The SNR for each in the individual tissue pressure intervals can be determined based on the amplitude spectrum of the tissue pressure signal. The absolute value of the sum of the complex amplitudes of the frequency components at the pulse rate and its harmonics indicates the total amplitude of the cardiac signal, referred to as the cardiac amplitude_i, where the integer i can be 1 or 2, corresponding to the first tissue pressure signal interval and the second tissue pressure signal interval acquired during the first and second applications of the repeated actuator pressure intervals.

[0143] The square root of the sum of the squares of the absolute values of the amplitudes of the remaining frequency components indicates the total noise in the signal, referred to as the noise amplitude_i, where i can be 1 or 2, corresponding to the first tissue pressure signal interval and the second tissue pressure signal interval acquired during the first and second applications of the repeated actuator pressure intervals. These metrics can be used to obtain an SNR measure, which is defined as SNR_i = cardiac amplitude_i / noise amplitude_i.

[0144] Then the factors SNR_1 / (SNR_1 + SNR_2) and SNR_2 / (SNR_1 + SNR_2) can be used respectively to determine the weighting of the first tissue pressure signal interval and the second tissue pressure signal interval acquired during the first and second applications of the repeated actuator pressure signal intervals, in order to obtain a single tissue pressure signal interval over the replicated actuator pressure intervals.

[0145] Alternatively, the SNR for each in the individual tissue pressure signal intervals can be determined based on the power spectrum. The sum of the power of the frequency components at the pulse rate and its harmonics indicates the total power of the cardiac signal, referred to as the cardiac power_i, where the integer i can be 1 or 2, corresponding to the first tissue pressure signal interval and the second tissue pressure signal interval acquired during the first and second applications of the repeated actuator pressure intervals. The sum of the power of the remaining frequency components indicates the total noise in the signal, referred to as the noise power_i, where i can be 1 or 2, corresponding to the first tissue pressure signal interval and the second tissue pressure signal interval acquired during the first and second applications of the repeated actuator pressure intervals. These metrics can be used to obtain an SNR measure, which is defined as SNR_i = cardiac power_i / noise power_i.

[0146] Then, the factors SNR_1 / (SNR_1+SNR_2) and SNR_2 / (SNR_1+SNR_2) can be used separately to determine the weighting of the first tissue pressure signal interval and the second tissue pressure signal interval collected during the first and second applications of the repeated actuator pressure signal, so as to obtain a single tissue pressure signal interval over the replicated actuator pressure intervals. The algorithm for hemodynamic measurement can be directly applied to the corrected tissue pressure signal to determine blood pressure and / or other hemodynamic parameters, since the corrected tissue pressure signal no longer contains the drop in the average tissue pressure level. In the foregoing paragraph, the averaging process has been explained for tissue pressure intervals collected twice for a portion of the actuator pressure ramp. In the case where there is more than one cuff crack and tissue pressure intervals have been collected three or more times for a portion of the actuator pressure ramp, the averaging process can be generalized to work with more than two intervals.

[0147] In addition, a predefined limit can be established for the maximum number of pressure intervals that can be averaged to obtain correction. In the case where the number of replicated pressure intervals due to cuff crack exceeds the predefined limit, no averaging attempt will be made. Instead, this can be reported to the user via the user interface, and a recommendation can be further transmitted to the user interface to re-wrap the cuff of the hemodynamic parameter measurement device around the patient and repeat the hemodynamic parameter measurement, or to replace the cuff of the hemodynamic parameter measurement device for the patient and repeat the hemodynamic parameter measurement.

[0148] The corrected measurement results can be displayed to the user via the user interface, optionally except that the displayed result is a suggestion of the corrected measurement result, so that the user can consider this when interpreting the result and can decide whether repeated measurement is needed. Optionally, in this case, the method can further include tracking the frequency of performing cuff-crack correction during a predefined number of measurement processes within a single measurement session (applying the cuff to the patient continuously once). If the number of corrections exceeds a predefined threshold, this can also be reported to the user, and a suggestion to re-wrap and / or replace the cuff can be generated.

[0149] In addition, regardless of the type of artifact, generally, the process of detecting an artifact and determining a recommended response action can be implemented in at least two ways. In the first method, artifact detection, classification, and providing a recommended follow-up action can be completed after a full measurement cycle. A full measurement cycle includes the gradual inflation / deflation of the cuff and a full cycle of acquiring measurement signals during that time. In the second method, artifact detection, classification, and providing a recommended response action can be performed in real time during the course of the measurement cycle. In this case, in response to detecting an artifact event, the system can immediately and directly abort the measurement. In other words, in response to detecting an artifact, the method can include generating a control signal for output to a hemodynamic parameter measurement device to control the device to abort the current hemodynamic parameter measurement.

[0150] Since measurements that continue to be affected by artifacts will lack clinical utility, this method saves clinical time. In addition, this method is more comfortable for the patient because the entire inflation cycle is only completed once when a good measurement result is likely.

[0151] Regarding the recommended response actions, these can be communicated to the user via a user interface device. The user interface can include a display. The user interface can include one or more other sensory output devices, such as an acoustic output device. As a non-limiting example, the recommendations and / or other advice can be provided via an audiovisual message, via a visual message, or via an audio message.

[0152] As briefly mentioned above, one or more classical algorithms or one or more machine learning algorithms can be used to perform artifact detection. Now regarding the use of machine learning algorithms, each of the one or more machine learning algorithms can be a classification algorithm configured to classify one or a set of signals as containing a specific artifact or not containing an artifact. A battery of such algorithms can be included, with one trained to detect each type of artifact. Each of the one or more machine learning algorithms can be trained by leveraging a database of the acquired measurement data (with various types of artifacts annotated in the data). That is, the database can contain the following annotations: the location of the artifact, the duration of the artifact, the type of the artifact, and the possible severity of the artifact.

[0153] Although in the examples discussed above, reference has been made to the use of an actuator pressure signal, more generally, any operating signal indicating an operating parameter of a pneumatic actuator may be used. For example, examples include an actuator activity signal indicating the pumping power level or pumping rate of a pump of a pneumatic actuator. This is in effect an alternative to the actuator pressure signal and is related to the physical interference of the measuring device in a similar manner to the actuator pressure signal. In fact, in all of the above methods for detecting a particular type of artifact, the reference to the actuator pressure signal may be replaced by a different operating parameter such as pumping power level or pumping rate, and the method will still work in the same way.

[0154] In summary, the methods and systems of embodiments of the present invention present an algorithm for automatic artifact detection and classification that provides information about the possible presence of artifacts and thus provides information about the signal quality during hemodynamic parameter measurement. The proposed solution analyzes the tissue pressure signal (TP) together with a signal indicating an operating parameter of a pneumatic actuator indicating a blood pressure signal and an optional motion sensor signal to determine and identify the likelihood of artifacts in the acquired signal data. Through the detection process, the type of artifact is also determined. This can be reported to the clinician / user via a user output, preferably also with a recommendation for further action.

[0155] The general workflow according to a set of advantageous embodiments can be summarized as follows. First, the sensor signals received from the hemodynamic parameter measurement device are analyzed for the presence of artifacts. This can be done based on, for example, amplitude information, correlation information, or template matching. The signal preferably includes at least a tissue pressure signal indicating the pressure between the surface of the user's body part and the cuff, and an actuator operating signal indicating an operating parameter of the pneumatic actuator. If artifacts are detected, they may optionally be further analyzed to determine the characteristics of the artifacts. For example, the start and end times of the artifact and / or the duration of the artifact may be determined. The amplitude corresponding to the waveform characteristics of the artifact may be determined in order to derive a severity index for a given artifact. For example, some of the artifacts discussed above are associated with inflection points in the signal. The amplitude of this inflection point can be measured and used as an indicator of severity.

[0156] Once the artifacts are located in time, they can be classified to determine whether they involve at least one of the following types of artifacts: motion artifacts, touching of the fluid tube connecting the pressure fluid sensor pad to the pressure transducer; stick-slip events related to the layers of the cuff; and, cuff cracking (loosening or slipping of the cuff fastening device).

[0157] Finally, a report can be generated for output to the user, the report including the type of artifact detected (e.g., one of the types described above), and preferably including a recommended follow-up action (e.g., at least one of the recommended follow-up actions described above).

[0158] In some embodiments, if the occurrence of an artifact is detected during the measurement process, the method may include automatically and directly aborting the measurement, and preferably reporting to the user the type of artifact detected, and preferably providing a recommended further action. In this way, the total measurement duration can be shortened, which saves time for the clinical team and provides greater comfort to the patient.

[0159] By detecting the occurrence of events that cause artifacts, embodiments of the present invention provide indicators for signal quality assessment and thus for assessment of the reliability of hemodynamic parameter measurements. Thus, the embodiments support clinicians in decision-making and improve the clinical workflow by avoiding errors in the clinical assessment of patients and avoiding false alarms, particularly in cases where repeated recommendations on how to remedy the events that cause artifacts are provided.

[0160] The embodiments of the present invention described above employ a processing unit. The processing unit can generally include a single processor or multiple processors. It can be located in a single containing device, structure, or unit, or it can be distributed among multiple different devices, structures, or units. Thus, a reference to a processing unit adapted or configured to perform a particular step or task can correspond to a step or task performed individually or in combination by any one or more of a plurality of processing components. Those skilled in the art will understand how such a distributed processing unit can be implemented. The processing unit includes a communication module or an input / output for receiving data and outputting data to other components.

[0161] One or more processors of the processing unit can be implemented in a variety of ways using software and / or hardware to perform the various functions required. Processors generally employ one or more microprocessors that can be programmed using software (e.g., microcode) to perform the required functions. A processor can be implemented as a combination of dedicated hardware for performing some functions and one or more programmed microprocessors and associated circuitry for performing other functions.

[0162] Examples of circuits that can be employed in various embodiments of the present disclosure include, but are not limited to, conventional microprocessors, application specific integrated circuits (ASICs), and field programmable gate arrays (FPGAs).

[0163] In various embodiments, a processor may be associated with one or more storage media, such as volatile and non-volatile computer memories such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that, when executed on one or more processors and / or controllers, perform the required functions. The various storage media may be fixed within the processor or controller or may be transportable such that one or more programs stored thereon may be loaded into the processor.

[0164] By studying the drawings, the disclosure, and the appended claims, those skilled in the art can understand and realize variations of the disclosed embodiments when practicing the claimed invention. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

[0165] A single processor or other unit may implement the functions of several items recited in the claims.

[0166] The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0167] A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium provided together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

[0168] If the term "adapted to" is used in the claims or the specification, it should be noted that the term "adapted to" is intended to be equivalent to the term "configured to".

[0169] Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A method (10) for use during measurement of hemodynamic parameters by a hemodynamic parameter measurement device (42), wherein, the hemodynamic parameter measurement device includes a cuff for partially wrapping around a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable balloon for changing the pressure exerted by the cuff on the body part, and the hemodynamic parameter measurement device further includes a tissue pressure sensor (44) and an operating parameter sensing device (46), the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body part and the cuff, the operating parameter sensing device for sensing the operating parameters of the pneumatic actuator; wherein, the method (10) includes: receiving (12) a set of sensor signals from the hemodynamic parameter measurement device, wherein the set of sensor signals includes: a tissue pressure signal (18) indicative of the pressure between the surface of the user's body part and the cuff, and an actuator operation signal (20) indicative of the operating parameters of the pneumatic actuator; processing (22) both signals using an artifact detection module to detect the occurrence of any one of a predefined set of different types of artifact events based on a combination of the signals, each artifact event type corresponding to a physical event affecting the sensor readings; and generating a report indicating the type of any one or more detected artifact events.

2. The method according to claim 1, wherein, the method is performed in real time during measurement of hemodynamic parameters.

3. The method according to claim 1 or 2, wherein, the artifact detection is performed based on detecting one or more characteristic features in the waveform of one or more of the received sensor signals.

4. The method according to any one of claims 1 - 3, wherein, the artifact detection includes detecting motion artifacts based on: high-pass filtering the actuator operation signal; and detecting a signal power level in the high-pass filtered signal that exceeds a predefined threshold.

5. The method according to any one of claims 1 - 4, wherein, the artifact detection includes: detecting motion artifacts based on calculating an autocorrelation metric of the tissue pressure signal; and detecting that the autocorrelation metric drops below a threshold.

6. The method according to any one of claims 1 - 5, wherein, the detection includes detecting motion artifacts based on: a first detection, which includes: high-pass filtering the actuator operation signal, and detecting a signal power level in the high-pass filtered signal that exceeds a predefined threshold at any time point; and a second detection, which includes: calculating an autocorrelation metric of the tissue pressure signal, and detecting that the autocorrelation metric drops below a threshold at any time point; wherein, in events where the time points corresponding to the first detection and the second detection are within a predefined proximity of each other, motion artifacts are detected.

7. The method according to any one of claims 1 - 6, wherein, The set of sensor signals further includes position and / or motion sensor signals indicative of the patient's posture and / or motion, and wherein the method includes detecting the occurrence of an artifact event including movement of the patient or a change in the patient's posture.

8. The method according to any one of claims 1-7, wherein, the artifact detection includes detecting an artifact event corresponding to a user touching a fluid line of the hemodynamic parameter measurement device, and wherein the detection includes: calculating an autocorrelation metric of the tissue pressure signal over a time period; and calculating a noise metric of the actuator operation signal, the noise metric indicating the degree of variation of the signal from a smooth ramp function over the time period; wherein the touch is detected in response to the autocorrelation of the tissue pressure signal being below a predefined threshold and the noise metric of the actuator operation signal being below a predefined threshold.

9. The method according to any one of claims 1-8, wherein, the artifact detection includes detecting an artifact event corresponding to a user touching a fluid line of a hemodynamic parameter measurement cuff, and wherein the detection includes detecting one or more predefined characteristic waveform patterns within the waveform of the tissue pressure signal.

10. The method according to any one of claims 1-9, wherein, the cuff includes a plurality of at least partially overlapping material layer portions; and wherein the artifact detection includes detecting an artifact event corresponding to a stick-slip event associated with the layer, and wherein the stick-slip event is detected based on: detecting an upward inflection point in the tissue pressure signal that exceeds a first predefined gradient threshold, and detecting a downward inflection point in the actuator operation signal that exceeds a second predefined gradient threshold.

11. The method according to any one of claims 1-10, wherein, the artifact detection includes detecting an artifact event corresponding to loosening or slipping of a cuff fastening device, and wherein the detection is based on: detecting a downward inflection point in the actuator operation signal that exceeds a first predefined gradient threshold, and detecting a downward inflection point in the tissue pressure signal that exceeds a second predefined gradient threshold; and optionally, wherein, the detection further includes detecting the downward inflection points in each of the tissue pressure signal and the actuator operation signal, followed by a signal portion of each respective signal, the signal portion representing a negative offset relative to the respective signal before the downward inflection point; and the detection further includes determining whether the duration of the signal portion exceeds a predefined threshold.

12. The method according to any one of claims 1-11, wherein, the operating parameter sensing device includes an actuation pressure sensor for sensing the pressure within the inflatable bladder, and the actuator operation signal includes an actuator pressure signal indicative of the pressure within the inflatable bladder; and / or wherein the actuator operation signal includes an actuator activity signal indicative of the pumping power level or pumping rate of the pump of the pneumatic actuator.

13. The method according to any one of claims 1-12, wherein, In response to detecting an artifact, the method includes generating a control signal for output to the hemodynamic parameter measurement device to control the device to abort the current hemodynamic parameter measurement.

14. The method according to any one of claims 1-13, wherein, the method further includes determining a recommended response action for the user to take based on applying a recommendation module to the output of the artifact detection module; and wherein the generated report also indicates the recommended response action for each detected artifact; and optionally, wherein, the recommended response action is to repeat the hemodynamic parameter measurement; the recommended response action is to re-wrap the cuff of the hemodynamic parameter measurement device around the user and repeat the hemodynamic parameter measurement; and / or the recommended response action is to replace the cuff of the hemodynamic parameter measurement device with a new cuff and repeat the hemodynamic parameter measurement.

15. A computer program product comprising code units configured to run on a processor operatively coupled to a hemodynamic parameter measurement device, the hemodynamic parameter measurement device including a cuff for partially wrapping around a portion of a user's body, and wherein, the cuff includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure applied by the cuff to the body part, and the hemodynamic parameter measurement device further includes a tissue pressure sensor and operating parameter sensing means, the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body part and the cuff, the operating parameter sensing means for sensing the operating parameters of the pneumatic actuator; and wherein the code units are configured to cause the processor to execute the method according to any one of claims 1-14 when run.

16. A processing unit (32) for a hemodynamic parameter measurement device, comprising: an input / output unit (34) for operatively coupling with a hemodynamic parameter measurement device (42) in use, the hemodynamic parameter measurement device including a cuff for partially wrapping around a portion of a user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable bladder for changing the pressure applied by the cuff to the body part, and the hemodynamic parameter measurement device further includes a tissue pressure sensor (44) and operating parameter sensing means (46), the tissue pressure sensor being arranged to sense the pressure between the surface of the user's body part and the cuff, the operating parameter sensing means for sensing the operating parameters of the pneumatic actuator; one or more processors (36) configured to execute a method that includes: receiving, via the input / output unit, a set of sensor signals from the hemodynamic parameter measurement device, wherein the set of sensor signals includes: a tissue pressure signal indicating the pressure between the surface of the user's body part and the cuff, and an actuator operation signal indicating the operating parameters of the pneumatic actuator; Process the signal using the artifact detection module to detect the occurrence of any one of a predefined set of different types of artifact events based on a combination of the signals, each artifact event type corresponding to a physical event that affects the sensor readings; and Generate a report indicating the type of any one or more detected artifact events and preferably export the report as a data item via the input / output unit.

17. A system (30), comprising: The processing unit (32) according to claim 16; and A hemodynamic parameter measurement device (42), which is operatively coupled to the processing unit, the hemodynamic parameter measurement device includes a cuff for partially winding around a part of the user's body, and wherein the cuff includes a pneumatic actuator in the form of an inflatable airbag for changing the pressure applied by the cuff to the body part, and the hemodynamic parameter measurement device further includes a tissue pressure sensor (44) and an actuation operation parameter sensing device (46), the tissue pressure sensor is arranged to sense the pressure between the surface of the user's body part and the cuff, and the actuation operation parameter sensing device is used to sense the operation parameters of the pneumatic actuator, for example, an actuator pressure sensor for sensing the pressure in the inflatable airbag.

Images