A method for monitoring bed mobility

Abstract

The invention involves a method for monitoring bed mobility, - the method comprising detecting a postural change of a target person lying on a target mattress (11TA), which detection is made by use of at least one pressure sensor (231-233) arranged to sense the pressure in at least one air chamber configuration (1101-1103) of the target mattress (11TA), and - determining whether or not the detected postural change of the target person is a tissue offloading postural change which substantially offloads tissue of the target person.

Description

[0001] A METHOD FOR MONITORING BED MOBILITY

[0002] TECHNICAL FIELD

[0003] The invention relates to a method for monitoring bed mobility, and to a computer program, a computer readable medium, and a control unit for carrying out steps of the method. The method may be used to prevent and / or treat pressure ulcers.

[0004] BACKGROUND

[0005] Pressure ulcers, also known as pressure sores, bedsores and pressure injuries, are localized damage to the skin and / or underlying tissue that usually occur over a bony prominence of a person as a result of pressure or pressure in combination with shear. Pressure ulcers commonly occur among individuals with low mobility, i.e. individuals who are unable to change position when for instance lying in a bed.

[0006] Pressure ulcer prevention guidelines promote the use of pressure redistribution through support surfaces such as bed mattresses and wheelchair cushions, in addition to frequent repositioning to offload vulnerable skin sites. Nursing homes and hospitals usually set care processes to assess the risk of their individuals in care and provide interventions to those that need it. This may include using a routine time frame for turning and repositioning a patient to support skin health. However, such programs will increase the workload of the staff working in the care setting and are often not personalized to the individual. This can result in both under and over treatment burden for a given individual.

[0007] In many cases, the use of pressure redistribution mattresses provides a high level of protection from skin damage. Such mattresses, which are becoming more popular, can be connected to a pump that is capable of supplying and evacuating air to or from the mattress to periodically change pressures under the individual. Pressure redistribution mattresses may contain multiple air chambers that are alternately pumped. The pump may follow a particular control strategy for increasing and reducing the air in the chambers. In such a way, the pressure may be varied and thus it provides time for the skin to recover between periods of loading.

[0008] Typically, these mattresses controlled by pumps are implemented through settings, operated by a care worker or healthcare professional. The digital information gathered by the pump is primarily used to control internal pressures within the mattress. However, there is a need for a cost effective way to determine the mobility status of individuals deemed at risk of developing pressure ulcers. Existing sensing technologies, such as wearable sensors, or mattress covers with pressure measuring sensor cells, are relatively costly, and may also complicate the care of the individuals.

[0009] US9295600B2 describes monitoring the activity level of an individual whilst positioned on a support apparatus which includes a plurality of separately or independently controllable inflatable zones. At least one pressure sensor is operably coupled to the interior region of each of the zones and / or the fluid conduit connected thereto. A mobility score is determined based on the person's activity level. It is suggested that “major” movements, “minor” movements, or combinations thereof may be detected and monitored. The documents describes minor changes in the position of a person's body or an extremity that are indicative of person’s mobility, while the person is positioned on the person support apparatus, where the minor changes in position are more significant in degree than physiological functions of the person such as breathing or heart rate. A major position change is exemplified as bed exit or sitting up.

[0010] Further, according to US9295600B2 threshold values for determining when the person's activity constitutes a major or minor movement are calculated or obtained from a lookup table. It is suggested that the activity threshold values are adjusted based on the person's weight, and / or the position of the person support apparatus. The document also suggests deriving an activity score by mapping results of an analysis of sensor inputs to threshold values that are indicative of different activity levels, which threshold values are determined through laboratory testing. For example, in the load cell embodiment, sensor inputs that indicate that each load cell is at about 50% of its dynamic range may indicate that a person is on the person support apparatus and is not moving. There is nevertheless a desire to provide a more accurate method for assessing or monitoring the bed mobility status of a person.

[0011] SUMMARY

[0012] An object of the invention is to provide a more accurate method for assessing or monitoring the bed mobility status of a person.

[0013] The object is reached with a method according to claim 1. Thus, the invention involves a method for monitoring bed mobility,

[0014] - the method comprising detecting a postural change of a target person lying on a target mattress, which detection is made by use of a pressure sensor arranged to sense the pressure in an air chamber configuration of the target mattress, and determining whether or not the detected postural change is a tissue offloading postural change which substantially offloads tissue of the target person.

[0015] The invention thus involves using, for detecting the postural change of the target person, a pressure sensor arranged to sense the pressure in an air chamber configuration of a mattress. Thereby, the sensor could be a pressure sensor provided with a pump for pressure redistribution mattresses. Thereby, the pump employed for the task of identifying tissue offloading postural changes of the target person lying on the mattress. Thereby, no additional sensors are needed for the tissue offloading postural change identification process. This combined use of a sensor allows a relatively low complexity of the system used, and a relatively low cost. It should be noted that in some embodiments, the sensor could be located in the air chamber configuration of the mattress, or in a conduit adapted to communicate with the air chamber configuration. Such a conduit could be adapted to be connected to a pump for pressure redistribution in the mattress.

[0016] The target mattress could be of a variety of types. For example, the target mattress could be provided with multiple air chamber configurations. Thereby, the pressure in the air chamber configurations could be adjusted individually so as to change the pressure distribution between the air chamber configurations. However, in some embodiments, the target mattress may comprise a single air chamber configuration, such as a bladder. In some embodiments, the target mattress may be split into sections, such as a head section and a feet section, wherein one or more of the sections comprise a respective air chamber configuration.

[0017] The method may comprise collecting a work set of sensing data from the pressure sensor arranged to sense the pressure in the air chamber configuration of the target mattress. The detection of the postural change of the target person may be done by analysing the work set of sensing data. It is understood that the collected work set of data may be used to determine whether or not the detected postural change of the target person is a tissue offloading postural change. Such a postural change may be a change from one posture of the target person to another posture of the target person. The postural change may be a change from an identified posture of the target person to another identified posture of the target person.

[0018] Preferably, the method comprises making while the target postural change of the target person takes place a plurality of measurements of the pressure in the air chamber configuration of the target mattress, and determining a time derivative of the pressure measurements, wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is made in dependence on the time derivative. The determination of whether or not the detected postural change of the target person is a tissue offloading postural change may involve comparing the time derivative of the pressure measurements at the postural change to a threshold time derivative which may be predetermined. Thus, in some embodiments, if the determined time derivative of the measurements of the pressure in the air chamber configuration of the target mattress is above the threshold value, it is determined that the detected postural change of the target person is a tissue offloading postural change. In such embodiments, if the determined time derivative of the pressure measurements is below the threshold value, it is determined that the detected postural change of the target person is a tissue non-offloading postural change. Such embodiments are based on an establishment that tissue offloading postural changes provide pressure time derivatives that are higher than pressure time derivatives provided by tissue non-offloading postural changes.

[0019] For example, one or more of the following postural changes may be tissue non-offloading postural changes which do not substantially offload tissue of the target person: head of bed raise, head of bed drop, arm and / or leg elevation. Further, one or more of the following postural changes may be tissue offloading postural changes: pelvic shift and moving up or down in bed, lateral turn, and transfer from lying to sitting. Such postural changes may offload vulnerable bony landmarks in the sacrum, heel or buttocks. Pelvic shift and moving up or down in bed may involve lifting the bottom and moving it up or down in bed. Moving up in bed may be a move away from the foot end of the bed. Moving down in bed may be a move towards the foot end of the bed.

[0020] The invention allows for determinations whether or not detected postural changes of the target person is a tissue offloading postural change to be made with a large degree of accuracy.

[0021] Rather than merely determining whether a person's activity constitutes a major or minor movement, where a minor movement could be a change in the position of the person’s body, or an extremity thereof, the invention allows for determining whether or not the movement is a tissue offloading postural change. It should be noted that what said US9295600B2 describes as minor movements include tissue offloading postural changes as well as tissue non-offloading postural changes, and US9295600B2 does not allow for determining whether or not a movement is a tissue offloading postural change.

[0022] The method may comprise, as exemplified below, taking a clinical action in relation to the target person lying on the target mattress in dependence on the determination whether or not the detected postural change is a tissue offloading postural change.

[0023] Preferably, the method comprises collecting a test set of sensing data, comprising, for each of one or more test mattresses, collecting sensing data from a pressure sensor arranged to sense the pressure in an air chamber configuration of the test mattress, wherein the sensing data is collected while a test person undergoes a plurality of postural changes on the test mattress, which postural changes preferably are predetermined, wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is done in dependence on the test set of sensing data.

[0024] The test mattress could be of a variety of types. For example, the test mattress could be provided with multiple air chamber configurations. However, in some embodiments, the test mattress may comprise a single air chamber configuration, such as a bladder. In some embodiments, the test mattress may be split into sections, such as a head section and a feet section, wherein one or more of the sections comprise a respective air chamber configuration.

[0025] It should be noted that the postures adopted by the test person are preferably performed in an uninterrupted sequence. Thereby, postural changes of the test person may be a change from one posture of the test person to another posture of the test person. Thereby, the postural changes of the test person may be correlated with changes in the signals from the pressure sensor, caused by changes in the pressure in the air chamber configuration of the mattress, in turn caused by the postural changes of the test person. Thereby, characteristics of signals from the pressure sensor correlated with tissue offloading postural changes may be different from the characteristics of signals from the pressure sensor correlated with tissue non-offloading postural changes. This difference in signal characteristics may be used to identify tissue offloading postural changes as different from tissue non-offloading postural changes.

[0026] It should be noted that in some embodiments, the method comprises collecting a work set of sensing data from two or more pressure sensors, each adapted to sense the pressure in a respective of two or more air chamber configurations of a target mattress. Further, in some embodiments, the method comprises collecting a test set of sensing data from two or more pressure sensors arranged to sense the pressure in a respective air chamber configuration of a test mattress. Preferably, the method comprises categorising one or more of the postural changes, which the test person undergoes, as tissue offloading postural changes which substantially offloads tissue of the target person, and categorising one or more of the postural changes, which the test person undergoes, as tissue non-offloading postural changes which do not substantially offload tissue of the target person, wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is done in dependence on the categorisation of the postural changes which the test person underwent. In some embodiments, one or more of the posture events are categorised as tissue offloading posture events, and the remaining of the posture events are categorised as tissue non-offloading posture events.

[0027] The invention also provides a computer program according to claim 16, a computer readable medium according to claim 17, and a control unit according to claim 18.

[0028] Further advantages and advantageous features of the invention are disclosed in the following description and in the dependent claims.

[0029] BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Below, embodiments of the invention will be described with reference to the drawings, in which:

[0031] - fig. 1 depicts parts of a system for preventing and / or treating pressure ulcers, with a portion of a mattress shown in a vertically oriented cross-section,

[0032] - fig. 2 and fig. 3 are flow diagrams depicting steps in a method according to an embodiment of the invention, and

[0033] - fig. 4a - fig. 4g shows examples of postures that may be adopted by test persons,

[0034] - fig. 5 shows signals from pressure sensors in the system in fig. 1, while a test person on the mattress adopts a series of postures,

[0035] - fig. 6 shows peak derivatives for different postural changes, across all test persons in a research project, of pressures measured by one of the sensors in the system in fig. 1, - fig. 7a - fig. 7c show ROC curves for tissue offloading postural changes in the research project, each ROC curve for a respective of the pressure sensors, and

[0036] - fig. 8 shows for comparison a ROC curve for tissue offloading postural changes for signals in an actimetry system used in the research project.

[0037] DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION

[0038] Embodiments of the invention involves a set of parts for preventing and / or treating pressure ulcers. The set of parts may include a plurality of mattresses, herein referred to as a mattress, a second mattress, etc.

[0039] Fig. 1 depicts a part of a mattress assembly 1 comprising a mattress 11TA. The mattress 11, in this example referred to as a target mattress, may be located on a bed frame (not shown), such as a hospital bed frame. The mattress assembly 1 comprises an air system. The air system comprises a plurality of, in this example three, assembly cavity formations

[0040] 1101-1103, 1111-1113. Each assembly cavity formation comprises an air chamber configuration 1101-1103 in the target mattress 11TA, and a conduit 1111-1113 externally of the target mattress 11TA. The conduit 1111-1113 is arranged to communicate with the respective air chamber configuration 1101-1103, and with a pump 2, described below.

[0041] Two of the air chamber configurations comprise elongated chambers 1102a, 1102b, 1103a, 1103b extending in a width direction of the target mattress 11TA. In the length direction of the target mattress, chambers of different such multi -chambered air chamber configurations 1102-1103 are altematingly distributed so that chambers of one air chamber configuration 1102 are separated by chambers of one or more, in this example one, other air chamber configurations 1103. In each multi -chambered air chamber configuration

[0042] 1102-1103, chambers are connected with inter-chamber conduits 1 lOlz, indicted with broken lines in fig. 1.

[0043] One of the air chamber configurations 1101 forms an under-mattress. It is located under the multi-chambered air chamber configurations 1102-1103. The under-mattress air chamber configuration 1101 comprises a single chamber. The pump 2 comprises a compressor 201, a valve assembly 211-213, and one or more, in this example three, mattress interfaces 221-223. An inlet conduit 204 is arranged to guide air from the surrounding atmosphere to the pump 2. The valve assembly comprises three valves 211-213, each arranged to communicate with the compressor and with a respective of the mattress interfaces 221-223.

[0044] The mattress assembly 1 comprises one or more, in this example three, pump interfaces 1121-1123 each adapted to be connected to a respective of the mattress interfaces 221-224 of the pump. The pump interfaces 1121-1123 are arranged to communicate with a respective of the conduits 1111-1113, which in turn communicate with a respective of the air chamber configurations 1101-1103. The pump interfaces are fixed to each other to form an interface assembly.

[0045] The pump comprises three pressure sensors 231-233. Each sensor is located in a respective of the mattress interfaces 221-223.

[0046] The pump 2 comprises a control unit 241 arranged to receive signals from the sensors 231- 233. The control unit 241 is arranged to control the valves 211-213.

[0047] Reference is made also to fig. 2. An embodiment of a method for monitoring bed mobility comprises collecting SI a work set of sensing data from one of the pressure sensors 231- 233, while a target person is lying on the target mattress 11TA. In this example, the work set of sensing data is collected by the sensor 231 arranged to sense the pressure in the single chamber air chamber configuration 1101. The target person may be a patient staying on the target mattress for a long period of time.

[0048] In other embodiments, the work set of sensing data could be collected from a plurality of the pressure sensors 231-233 arranged to sense the pressure in a respective of the air chamber configurations 1101-1103. While the work set of sensing data is collected, the compressor 201 of the pump 2 may be non-operative.

[0049] The work set of sensing data, as well as a test set of sensing data described below, may be collected by sensor signal readings at a predetermined frequency, e.g. 1Hz.

[0050] The method further comprises using the collected work set of data to detect S2a one or more postural changes of the target person lying on the target mattress 11TA. The method further comprises determining S2b based on a test set of sensing data described below, and on a categorisation of postural changes described below, whether or not the detected postural changes of the target person are tissue offloading postural changes.

[0051] The determination S2b whether or not the detected postural changes of the target person are tissue offloading postural changes comprises determining time derivatives of pressure measurements in the work set of data. The time derivatives of the pressure measurements are compared to a threshold time derivative which may be predetermined. Where such a time derivative is above the threshold value, it is determined that a postural change has taken place which is a tissue offloading postural change. Where such a time derivative is below the threshold value, it is determined that no tissue offloading postural change occurred at the time of the measurements for which the time derivative was determined.

[0052] The method may be used for preventing or treating pressure ulcers. Thereby, in dependence on whether or not the detected postural changes are tissue offloading postural changes postural changes, a decision may be made S3 on whether or not to take a clinical action with the target person lying on the target mattress 11TA.

[0053] The decision whether or not to take a clinical action may depend on how many of the detected postural changes are tissue offloading postural changes. The decision whether or not to take a clinical action may also depend on how many tissue offloading postural changes that have occurred within a certain period of time. For example, the method may comprise determining the number of tissue offloading postural changes, compared to the number of tissue non-offloading postural changes. The decision on whether or not to take a clinical action may be made in dependence on said postural change count. If a decision is made to not take any clinical action, the step of collecting SI a work set of sensing data from one of the pressure sensors 231-233 may be repeated.

[0054] The clinical action may comprise controlling the valve assembly 211-213 of the pump 2, and optionally the compressor 201, so as to adjust the pressure in one or more of the air chamber configurations 1101-1103. Thus, if a decision is made S3 to take a clinical action, the valve assembly 211-213 of the pump 2, and optionally the compressor 201, may be controlled so as to adjust S4 the pressure in one or more of the air chamber configurations 1101-1103. For this adjustment control, the pressure sensors 231-233 are used.

[0055] The clinical action may comprise controlling the valve assembly 211-213 and the compressor 201 according to a control strategy. The control strategy may for example comprise controlling the pressures in the air chamber configurations 1101-1103, so that each of them are in a series of mutually different levels held at respective time intervals.

[0056] Thus, the determinations whether or not the detected postural changes are tissue offloading postural changes, may be used to support the making of clinical decisions for patient care. Further, the making of clinical decisions may be automated, as exemplified above. Further, the clinical action performed as a result of a decision, which can be automated or made by a human, may be performed by mattress pressure redistribution, involving use of the one or more sensors used for the collection of the work set of sensing data.

[0057] It should be noted that in some embodiments, the decision whether or not to take a clinical action may be caused by a signal, alerting one or more persons in whose care the target person is. For example, the decision may involve turning the target person. The alert signal may for example be sent by a communication link to a health care provider.

[0058] As understood, the method for monitoring bed mobility may provide data for clinical decisions. Reference is made also to fig. 3. The identification of the one or more postural changes of the target person lying on the target mattress 11TA is done using an algorithm. The algorithm is determined, before said steps of collecting a work set of sensing data and determining whether or not detected postural changes are tissue offloading postural changes, as follows:

[0059] A test set of sensing data is collected S 101. This collection comprises, for each of one or more test mattresses 11TE (fig. 1), collecting sensing data from a pressure sensor 231-233 arranged to sense the pressure in an air chamber configuration 1101-1103 of the test mattresses 11TE. Although the mattress in fig. 1 is herein referred to as a target mattress 11TA as well as a test mattress 11TE, it is understood that they could be different mattresses.

[0060] The sensing data for the test set of sensing data is collected while a test person is adopting a plurality of predetermined lying postures on the test mattress 11TE. Thereby, a sequence of movements may be performed by the test person. Thus, the test person may present a series of postural changes. Each postural change may be a change from one posture of the test person to another posture of the test person. Preferably, the postures are correlated in time with the collected test set of sensing data. Thus, a record of the postures may be correlated in time a record of the collected test set of sensing data. The test person may be a healthy person, engaged to perform the postures for the sensing data test set collection. The test person may be a person different from the target person.

[0061] The postures that are correlated with the collected test set of sensing data may be recorded with the aid of a separate sensing system including e.g. wearable sensors, or a mattress cover with pressure measuring sensor cells. Alternatively, postures that are correlated with the collected test set of sensing data may be observed and noted by another person.

[0062] Where the test set of sensing data is collected on two or more test mattresses 11TE, they are preferably of the same type, e.g. of the same make and model. Thereby, the air chamber configurations 1101-1103 of one of the two or more test mattresses 11TE are identical to the air chamber configurations 1101-1103 of the remaining of the two or more test mattresses 11TE.

[0063] Preferably, the target mattress 11TA on which the work set of sensing data is collected is of the same type as the one or more test mattresses 11TE on which the test set of sensing data is collected. Thereby, the air chamber configurations 1101-1103 of the target mattress 11TA on which the work set of sensing data is collected, are identical to the air chamber configurations 1101-1103 of the one or more test mattresses 11TE on which the test set of sensing data is collected. However, in some embodiments, the target mattress 11TA is of a type that is different from the type of the one or more test mattresses 11TE. In such embodiments, thresholds used in determinations whether postural changes are tissue offloading may be different from corresponding thresholds used in embodiments where the target mattress 11TA and the one or more test mattresses 11TE are of the same type.

[0064] The predetermined postures may include the supine posture. Where the test mattress 11TE is located on an adjustable hospital bed, the supine posture may be followed by raising the head of bed in one or more increments, in turn followed by lowering the head of bed in one or more increments until the supine posture is re-established. Each posture may be held for a predetermined period of time, e.g. of 5 - 10 minutes. Further, the predetermined postures may include tilting of the pelvis, e.g. 30°, in a transverse plane of the test person’s body. For the pelvis tilting, pillows may be placed under the knees and thorax.

[0065] Adopting the predetermined postures may include a series of functional postural changes to simulate normal activities in bed. The functional postural changes may be relatively small in magnitude. The functional postural changes may include an elevation of the upper and lower limbs, small postural adjustments of the pelvis, and shifting such as moving up and down the bed, and transfer from lying to sitting.

[0066] The functional postural changes are preferably performed in a random order to mitigate the risk of order effects. A test set of sensing data was collected in a research project performed by representatives of the University of Southampton in 2022. The test set of sensing data was collected from posture adoptions of fifteen test persons. Fig. 4a - fig. 4g shows examples of postures that may be adopted by such test persons. Fig. 4a shows a supine lying position. Fig. 4b - 4d show positions at a head of bed (HOB) raised to 30°, 60°, and 90°, respectively. Fig. 4e and fig. 4f show lateral rotations of 30° and 90°, respectively. Fig. 4g shows a posture in which the pelvis is in a non-zero angle to a horizontal plane, in this example 30°.

[0067] Fig. 5 shows signals from all pressure sensors 231-233 while one of the test persons in said research project adopted the postures and postural changes PE1-PE14 listed in table 1 below, herein commonly referred to as posture events.

[0068] Postural change Movement description Tissue offloading?

[0069] PE 1 HOB raise to 20° from supine lying No

[0070] PE 2 HOB raise to 40° from HOB 20° No

[0071] PE 3 HOB raise to 60° from HOB 40° No

[0072] PE 4 HOB drop to 40° from HOB 60° No

[0073] PE 5 HOB drop to 20° from HOB 40° No

[0074] PE 6 Supine lying from HOB 20° No

[0075] PE 7 Lateral turn 1 Yes

[0076] PE 8 Supine lying from lateral turn 1 Yes

[0077] PE 9 Lateral turn 2 Yes

[0078] PE 10 Supine lying from lateral turn 2 Yes

[0079] PE 11 Right & left arm and leg elevations No

[0080] PE 12 Pelvic shift & moving up-down in bed Yes

[0081] PE 13 Transfer to sitting Yes

[0082] PE 14 From sitting back to supine lying Yes

[0083] Table 1

[0084] As can be seen, the postural changes in the table are categorised as follows: Lateral turns, to supine lying from lateral turns, (PE7-PE10), pelvic shift and moving up- down bed (PE12), transfer to sitting tissue (PE13), and from sitting back to supine lying (PE 14) are categorised as offloading postural changes. The lateral turn 1 and the lateral turn 2 are lateral turns in opposite directions. PE12 is a pelvic shift with a move up or down in bed.

[0085] Head of bed raise (PE1-PE3), head of bed drop (PE4, PE5), to supine lying from HOB (PE6), arm and / or leg elevation (PEI 1) are categorised as tissue non-offloading postural changes. PEI 1 is an elevation of only one arm or only one leg.

[0086] Signal 1 is from the sensor 231 detecting the pressure in the single chamber air chamber configuration 1101. Signals 2 and 3 are from the sensors 232, 233 detecting the pressures in the multi-chamber air chamber configurations 1102, 1103.

[0087] The test set of data is processed SI 02 to create the algorithm to identify tissue offloading postural changes based on the work set of sensing data. Thus, the tissue offloading postural change identification is done based on the test set of sensing data.

[0088] The processing of the test set of data may include obtaining time derivatives of pressure measurements. Thus, for a specific detected postural change a time derivative of the pressure sensed in the air chamber configurations of the test mattress may be determined. The processing of the test set of data may include an assumption that tissue offloading postural changes provide pressure time derivatives that are higher than pressure time derivatives provided by tissue non-offloading postural changes. The processing may include setting a threshold value for the time derivative of pressure measurements such that the time derivatives for all tissue offloading postural changes that the test person underwent are above the threshold value, and the time derivatives for all tissue nonoffloading postural changes that the test person underwent are below the threshold value.

[0089] More particularly, the processing of the test set of data may include resampling for a common frequency, e.g. 1 Hz, across all collected data to facilitate comparisons, offset deletion, cropping, getting 1st time derivatives, and / or filtering using moving average and median filters. The processing of the test set of data may further include labelling and extracting the peak magnitudes of the predefined postural changes, from the filtered derivatives of all the measurement signals across all test persons adopting the postures. These derivative peaks may be used to estimate signal threshold values for a sensitivity and specificity analysis for postural change classification.

[0090] Fig. 6 shows peak derivatives, across all test persons in said research project, of pressures measured by the sensor 231 detecting the pressure in the single chamber air chamber configuration 1101.

[0091] Sensitivity and specificity to detect tissue offloading postural changes for each sensor signal may be presented as a percentage true positive and true negative tissue offloading postural change detections. Sensitivity may be calculated by assessing the number of cases in which a sensor signal, with respect to a prescribed threshold level, correctly classified a specific tissue offloading postural change, i.e. a true positive. Specificity may be calculated for each postural change as the proportion of cases where the sensor signals did not exceed the selected threshold level during the postural change, i.e. a true negative. This forms part of a standard Receiver Operator Characteristic (ROC) Curve analysis, with an associated area under the ROC curve (AuC).

[0092] In the research project a range of thresholds were tested as part of the ROC analysis. Sensitivity and specificity plots were made for each of the threshold values for postural change discrimination for signals from the sensor 231 detecting the pressure in the single chamber air chamber configuration 1101, from the peak derivatives across all fifteen test persons in the research project.

[0093] As suggested, creating the algorithm may comprise using a Receiver Operating Characteristic (ROC) analysis on the test set of sensing data. The ROC may be used to determine the optimal range of parameters, which discriminate between the presence and absence of tissue offloading postural changes. Each of the sensor signals and the possible thresholds may be examined to identify the combination with the optimal sensitivity and specificity for all evoked postural changes. The area under the ROC curve (AUC), which plots sensitivity versus false positive rate (100 - specificity), may be calculated to assess the overall accuracy in discriminating tissue offloading postural changes for the signal. The AUC for the ROC curves quantifies the probability that a test correctly discriminates the presence or absence of a tissue offloading postural change, and ranges from 0 to 1. Those signals achieving high AUC values (>0.8) from the ROC analysis may be used to develop an efficient algorithm for tissue offloading postural change detection.

[0094] Fig. 7a - fig. 7c show ROC curves for tissue offloading postural changes in the research project, each ROC curve for a respective of the pressure sensors 231-233. It is suggested that AUC values between 0 and 0.5, between 0.5 and 0.8, between 0.8 and 0.9, and above 0.9 indicate poor, moderate, good, and excellent discrimination, respectively, (IBM SPSS statistics V28; IBM, Armonk, New York). The ROC curves for the tissue offloading postural changes have AUC values of 0.89, 0.88, and 0.88, indicating good, and close to excellent discrimination, for tissue offloading postural changes.

[0095] As a comparison, reference is made to fig. 8, showing a ROC curve for tissue offloading postural changes for signals in an actimetry system used in the research project. The AUC for this curve is 0.88. Thus, the method according to this embodiment of the invention provides discrimination for tissue offloading postural changes, at level with that of the actimetry system, which is referred to as the ‘gold standard’.

[0096] Creating the algorithm may further comprise the use of a machine learning model on the test set of sensing data to detect tissue offloading postural changes. Such a model may involve the extraction of data epochs, data reduction (principal component analysis), and learning strategies (KNN, Naive-Bayes, Support Vector Machine).

[0097] Various alternatives to the embodiments described above are possible. In some embodiments, the target mattress 11TA on which the work set of sensing data is collected is a test mattress 11TE on which the test set of sensing data is collected, i.e. the same mattress is used for both collections. In some embodiments, the person adopting the postures for the collection of the test set of sensing data, and the person whose postural changes are identified using the work set of data, are the same person. I.e., the test person may be the target person. For example, while the test set of sensing data is collected, the postures of the person, e.g. a patient, may be observed, e.g. manually or by means of an additional sensing system. These observed postures may be correlated in time with the collected test set of sensing data. Personalized predetermined threshold values which are unique to that patient or target person may then be a determined. The postures adopted by the person during the collection of the test set of sensing data may be not predetermined but may be randomly occurring movements of the person.

[0098] In some embodiments, the pressure sensor 231-233 from which the work set of sensing data is collected is not located in a pump for a pressure redistribution mattress. For example, the pressure sensor may be located in the mattress, or in a conduit connecting the mattress with a pump for pressure redistribution of the mattress. In further embodiments, the mattress may not be a pressure redistribution mattress. For example, the mattress may be a static mattress, e.g. a pressure relieving mattress. Thereby, the pressure sensor may be located in the mattress, or in a space which communicates with an air chamber configuration of the mattress. In some embodiments, the mattress includes one or more local air chambers, which may be relatively small, which may include a respective sensor, to obtain more information.

[0099] It should be noted that the invention is applicable also in cases where the mattress assembly comprises only one assembly cavity formation. Thus, in such examples, the mattress may comprise only one air chamber configuration, e.g. with a single air chamber.

[0100] It is to be understood that the present invention is not limited to the embodiments described above and illustrated in the drawings; rather, the skilled person will recognize that many changes and modifications may be made within the scope of the appended claims.

Claims

CLAIMS1. A method for monitoring bed mobility,- the method comprising detecting a postural change of a target person lying on a target mattress (11TA), which detection is made by use of at least one pressure sensor (231-233) arranged to sense the pressure in at least one air chamber configuration (1101-1103) of the target mattress (11TA), characterised by determining whether or not the detected postural change of the target person is a tissue offloading postural change, which substantially offloads tissue of the target person and comprising, while the detected postural change of the target person takes place, making a plurality of measurements of the pressure in the at least one air chamber configuration of the target mattress (11TA), and determining a time derivative of the pressure measurements, wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is made in dependence on the time derivative.

2. A method according to claim 1 wherein one or more of the following postural changes are categorised as tissue offloading postural changes: lateral turn, pelvic shift and moving up or down in bed, transfer from lying to sitting.

3. A method according to claim 1 or claim 2 wherein one or more of the following postural changes are categorised as tissue non-offloading postural changes: head of bed raise, head of bed drop, arm and / or leg elevation.

4. A method according to any of the previous claims wherein the determination of whether or not the detected postural change of the target person is a tissue offloading postural change comprises comparing the time derivative of the pressure measurements at the postural change to a predetermined threshold time derivative.

5. A method according to claim 4 wherein the predetermined threshold time derivative is based on a test set of sensing data obtained from the target person.

6. A method according to claim 4 or 5 wherein if the determined time derivative of the measurements of the pressure in the at least one air chamber configuration of the target mattress is above the predetermined threshold value, it is determined that the detected postural change of the target person is a tissue offloading postural change.

7. A method according to any of claims 4 to 6 wherein if the determined time derivative of the pressure measurements is below the predetermined threshold value, it is determined that the detected postural change of the target person is a tissue non-offloading postural change.

8. A method according to any one of the preceding claims, comprising taking a clinical action in relation to the target person lying on the target mattress (11TA) in dependence on the determination whether or not the detected postural change is a tissue offloading postural change.

9. A method according to any one of the preceding claims, comprising collecting a test set of sensing data, comprising, for each of one or more test mattresses (11TE), collecting sensing data from at least one pressure sensor (231-233) arranged to sense the pressure in at least one air chamber configuration (1101-1103) of the test mattress (11TE), wherein the sensing data is collected while a test person undergoes a plurality of postural changes on the test mattress (11TE), which postural changes preferably are predetermined, wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is done in dependence on the test set of sensing data.

10. A method according to claim 9, wherein the test set of sensing data is collected on two or more test mattresses (11TE), wherein one or more air chamber configurations (1101-1103) of one of the two or more test mattresses (11TE) areidentical to one or more air chamber configurations (1101-1103) of the remaining of the two or more test mattresses (11TE).

11. A method according to any one of claims 9 or 10, wherein one or more air chamber configurations (1101-1103) of the target mattress (11TA) are identical to one or more air chamber configurations (1101-1103) of the one or more test mattresses (11TE).

12. A method according to any one of claims 9-11, comprising categorising one or more of the postural changes, which the test person undergoes, as tissue offloading postural changes which substantially offloads tissue of the target person, and categorising one or more of the postural changes, which the test person undergoes, as tissue non-offloading postural changes which do not substantially offload tissue of the target person, wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is done in dependence on the categorisation of the postural changes which the test person underwent.

13. A method according to any one of claims 9-12, comprising processing the test set of data to create an algorithm, wherein the determination whether or not the detected postural change is a tissue offloading postural change is done using the algorithm.

14. A method according to claim 13, wherein processing the test set of data to create an algorithm comprises determining for the postural changes that the test person undergoes time derivatives of the pressure sensed in the at least one air chamber configuration (1101-1103) of the test mattress (11TE), wherein the determination whether or not the detected postural change of the target person is a tissue offloading postural change is made in dependence on the determined time derivatives.

15. A method according to any one of claims 13-14, wherein processing the test set of data to create an algorithm comprises using a Receiver Operating Characteristic(ROC) analysis on the test set of sensing data, to characterize the sensitivity and specificity of the detection threshold.

16. A method according to any one of the preceding claims, comprising providing a pump (2), wherein the pump comprises a compressor (201), a valve assembly (211- 213), and one or more mattress interfaces (221-223), wherein the valve assembly is arranged to communicate with the compressor and with the one or more mattress interfaces, wherein a mattress assembly (1), comprising the target mattress (11TA), comprises one or more pump interfaces (1121-1123) each connected to a respective of the mattress interfaces (221-223) of the pump, wherein each pump interface communicates with a respective of the at least one air chamber configurations (1101-1103) of the target mattress (11TA).

17. A method according to claim 16, wherein the pump comprises the pressure sensor (231-233) by use of which the postural change of the target person is detected, wherein the pump forms a space (221-223) arranged to communicate with one of the at least one air chamber configuration (1101-1103) of the target mattress(11TA), wherein the sensor is located in said space (221-223).

18. A method according to any one of claims 16-17, comprising taking a clinical action in relation to the target person lying on the target mattress (11TA) in dependence on the determination whether or not the detected postural change is a tissue offloading postural change, wherein the clinical action comprises controlling the valve assembly (211-213) of the pump (2).

19. A method according to any of claim 9-18 wherein the test person is the target person.

20. A computer program comprising program code means for performing the steps of any one of claims 1-19 when said program is run on a computer.

21. A computer readable medium carrying a computer program comprising program code means for performing the steps of any one of claims 1-19 when said program product is run on a computer.

22. A control unit (241) configured to perform the steps of the method according to any one of claims 1-19.

Images