Wound treatment system

Abstract

An exemplary system includes a memory and one or more processors in communication with the memory, the one or more processors being configured to receive patient information and select at least one NPWT parameter setting based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at a wound site. The one or more processors are further configured to control fluid at the wound site via an NPWT dressing based on the at least one selected NPWT parameter setting.

Description

Technical Field

[0001] The present disclosure relates to wound monitoring.

Background Art

[0002] Smart wound dressings, wound treatment, and wearable sensors are used to monitor healing and administer treatment to a wound. The treatment can be adjusted based on the monitoring of healing.

Summary of the Invention

[0003] Generally, the present disclosure describes a system and technique for selecting wound treatment parameter settings based on a causal model that determines the current causal relationship between wound treatment parameter settings and, for example, the effect of wound treatment administered according to the wound treatment settings.

[0004] For example, a negative pressure wound therapy (NPWT) system can be configured to receive patient information including the effect of controlling fluid at the wound site, and select NPWT parameter settings for controlling fluid at the wound site based on a causal model that determines the current causal relationship between the NPWT parameter settings and, for example, the effect of controlling fluid at the wound site according to the selected NPWT parameter settings.

[0005] The disclosed system and technique can lead to improved patient outcomes, for example, by improving treatment control decisions. For example, the disclosed system and technique can implement an advanced deep causal learning (DCL) algorithm to make control decisions regarding a large and complex input dataset related to the adjustment of control parameters for NPWT and wound perfusion / instillation treatment. A DCL-based feedback control system can enable multi-objective improvement of patient outcomes, such as improving the healing rate while reducing patient discomfort, wound infection, maceration around the wound, and other undesirable complications.

[0006] In one example, the present disclosure describes a method comprising receiving patient information, selecting at least one NPWT parameter setting for controlling fluid at a wound site via an NPWT dressing based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, and controlling fluid at the wound site via the NPWT dressing based on the selected at least one NPWT parameter setting.

[0007] In another example, the present disclosure describes a system comprising a memory and one or more processors in communication with the memory, the one or more processors configured to receive patient information, select at least one NPWT parameter setting based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, and control fluid at the wound site via an NPWT dressing based on the selected at least one NPWT parameter setting.

[0008] In another example, the present disclosure describes a computer-readable medium comprising instructions that, when executed, cause one or more processors to receive patient information, select at least one NPWT parameter setting based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, and control fluid at the wound site via an NPWT dressing based on the selected at least one first NPWT parameter setting.

[0009] In another example, the present disclosure is based on a causal model that determines the current causal relationship between means for receiving patient information and a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, and includes means for selecting at least one NPWT parameter setting for controlling fluid at the wound site via an NPWT dressing, and means for controlling fluid at the wound site via the NPWT dressing based on the selected at least one NPWT parameter setting.

[0010] Details of one or more examples are set forth in the accompanying drawings and the description below. Other features, objects, and advantages will be apparent from the description and drawings, and from the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0011]

Figure 1

[0012]

Figure 2

[0013]

Figure 3

[0014]

Figure 4

[0015]

Figure 5

[0016]

Figure 6

[0017]

Figure 7

[0018]

Figure 8

[0019]

Figure 9

[0020]

Figure 10

[0021]

Figure 11

DETAILED DESCRIPTION OF THE INVENTION

[0022] In an example, the present disclosure describes a system and technique for selecting wound treatment parameter settings based on a causal model that determines the current causal relationship between wound treatment parameter settings and, for example, the effect of wound treatment performed according to the wound treatment settings.

[0023] Sensor technology can improve wound care, enable and enhance a telemedicine-based healthcare paradigm. For example, system control parameters of negative pressure wound therapy (NPWT) and irrigation / drip wound therapy (such as pressure, flow rate and / or dwell time (in the case of drip), temperature and content of irrigation / drip solution, etc.) can be adjusted to improve patient outcomes (e.g., promoting healing and / or adjusting control parameters to reduce patient discomfort based on feedback). In some examples, such adjustment of control parameters can enable telemedicine by automating system adjustment decisions (e.g., a clinician need not observe the patient and / or wound to determine modification of treatment parameters).

[0024] Some techniques may be unable to handle data complexity, dynamics, and interactions and thus may be limited in improving patient outcomes. For example, control decisions to adjust system control parameters of NPWT and irrigation / drip therapy can be based on diverse and complex real-time dynamic data sets (such as patient demographic information, patient medical history, on-demand user input, spatio-temporal dynamic sensor measurements of wound tissue, dynamic patient biometrics, etc.).

[0025] For example, a wound care specialist can regularly measure the wound shape and determine whether and / or if healing has occurred based on the reduction in wound area / volume over time. These regular measurements require the removal of the dressing and a detailed examination of the wound, such as inserting a ruler into the wound to assess wound depth. Removal of the dressing can disrupt the wound bed and, for example, add to the exposure time of the wound to a non-sterile environment. If the wound care specialist determines that the wound area / volume has not decreased, the treatment strategy can be changed. For example, the wound care specialist can change the NPWT and temporarily pause the NPWT (e.g., "vac vacation"). In some examples, the wound care specialist can perform an infusion and / or irrigation regimen as part of the treatment such that the wound is periodically filled with fluid for a set dwell time and drained by a new application of negative pressure (in the case of infusion) or continuously washed with fluid simultaneously with the application of negative pressure (in the case of irrigation) to, for example, cleanse the slough and remove trace amounts of microorganisms. In some examples, the wound care specialist can locally deliver concentrated oxygen to the wound bed to improve the oxygen gradient at the wound-tissue interface. As used herein, "fluid" refers to any substance that deforms or "flows" when subjected to one or more external forces (e.g., pressure, gravity, etc.). The fluid can be used according to various techniques such as the NPWT described herein regardless of whether it is in a liquid state (e.g., saline, distilled water, etc.), a gaseous state (e.g., a mixture of gases such as air, a gaseous element such as pure oxygen, etc.), a plasma state, or other state.

[0026] In some instances, a wound care specialist can make various control decisions to adjust the system control parameters of NPWT and irrigation / drip therapy based on experience, intuition, and discontinuous subjective measurement criteria. For example, a wound care specialist can, based on experience, intuition, and discontinuous subjective measurement criteria, determine the length of time to pause negative pressure, whether to initiate negative pressure after a pause and when to initiate it, the length of time to apply negative pressure, whether to cycle the pause and the "on" and "off" times of the cycle, the time to end negative pressure therapy, the purge rate, whether to irrigate with fluid, whether to drip fluid at a given "on" and "off" time of the cycle, the fill and / or concentration of the drip fluid for each cycle, the drip delivery rate, the dwell time, and the purge rate, as well as any other suitable NPWT control decisions. In some instances, it can be difficult for a wound care specialist and / or the current NPWT control method to determine the causal relationship between control decisions and treatment outcomes and / or metrics in order to base control decisions on objective criteria rather than subjective criteria, for example, based on experience, intuition, and discontinuous subjective measurement criteria, but rather based on, for example, cause and effect.

[0027] The disclosed systems and techniques can result in improved patient outcomes, for example, by improving treatment control decisions. For example, the disclosed systems and techniques can implement an advanced deep causal learning (DCL) algorithm to make control decisions regarding a vast and complex input dataset related to the adjustment of control parameters for NPWT and wound irrigation / drip therapy. A DCL-based feedback control system can enable multi-objective improvement of patient outcomes, such as improving the healing rate while reducing one or more of patient discomfort, wound infection, maceration around the wound, or other undesirable complications.

[0028] In some examples, a control system can monitor and adjust stimuli to improve and promote tissue health. For example, a skin and / or wound health sensor can be combined with one or more wound treatments with a stimulus and DCL-based control algorithm to adjust the treatment and improve patient outcomes. The control system can receive algorithm inputs, and the DCL model and / or algorithm can be configured to select control parameter settings, such as NPWT parameter settings, and a wound treatment and / or treatment system, such as an NPWT system, can be configured to provide wound treatment based on the selected control parameter settings.

[0029] In some examples, the algorithm input can include patient information. The patient information can include patient input, real-time and / or recorded patient biometric sensor data, real-time and / or recorded wound sensor data, and any other suitable patient and / or wound information. Patient input, such as user input received from a clinician and / or patient, can include patient demographics, patient health record data, input data regarding patient discomfort, clinician input regarding prognosis, a list of treatment options, and any other suitable patient input information. In some examples, the patient input can enable a clinician to control the treatment regimen. Real-time and / or recorded patient biometric sensor data can include blood pressure, heart rate, body temperature, blood glucose level, albumin, prealbumin, tissue oxygen concentration, oxygenated hemoglobin level, and the like. Real-time and / or recorded wound sensor data can include impedance-based wound monitoring, imaging, temperature and / or pressure measurements, and / or any other suitable wound sensor data.

[0030] In some examples, the DCL model and / or algorithm can be configured to process large, spatio-temporally dynamic, and often interrelated amounts of input data to adjust control parameters related to treatment options. The DCL model and / or algorithm can be configured to determine the current causal relationship between a control parameter setting and the effect of providing wound treatment based on the control parameter setting.

[0031] In some examples, a wound treatment and / or treatment system can include NPWT and a combined NPWT infusion system (e.g., V.A.C. VERAFLO™ treatment) or NPWT and a combined NPWT perfusion system. The term "negative pressure" refers to an absolute pressure lower than the absolute atmospheric pressure at the location where the device is used. Thus, a defined level of negative pressure in a region is a relative measure between the absolute atmospheric pressure and the absolute pressure in that region. A description that the negative pressure is decreasing means that the pressure within the region is shifting towards atmospheric pressure (e.g., the absolute pressure is increasing). When using numerical values, a negative sign is placed in front of the number of the pressure value to indicate that the value is negative pressure relative to atmospheric pressure.

[0032] FIG. 1 is a diagram showing an exemplary system 2 according to the techniques described in the present disclosure. As shown in FIG. 1, system 2 includes a patient 4, an NPWT dressing 20, a treatment system 12, and a server 24 that can communicate via a network 10.

[0033] The treatment system 12 can be configured to receive algorithm inputs, select control parameter settings for wound treatment based on a DCL model and / or algorithm, and perform wound treatment by a wound treatment and / or treatment system, such as an NPWT system, according to the selected control parameter settings. In the illustrated example, the treatment system 12 includes an NPWT device 14, a sensor 16, and a computing device 18.

[0034] The NPWT device 14 can comprise a system for effecting fluid delivery to a wound dressing, such as an NPWT dressing 20. The NPWT device 14 can include a reservoir and a negative pressure source coupled to the reservoir and the NPWT dressing 20. In some examples, the NPWT device 14 can further include one or more of a fluid flow device in fluid communication with the fluid supply reservoir, such as a pump, valve, or generator. As used herein, "fluid" refers to any substance that deforms or "flows" when subjected to one or more external forces (e.g., pressure, gravity, etc.). The fluid can be used in accordance with various techniques such as NPWT described herein, regardless of whether it is in a liquid state (e.g., saline, distilled water, etc.), a gaseous state (e.g., a gaseous mixture such as air, a gaseous element such as pure oxygen), a plasma state, or other state.

[0035] The NPWT device 14 can be, for example, a Veraflo™ therapy system, a V.A.C. ULTA™ therapy system that can include an INFOV.A.C.™ canister, a V.A.C.® therapy system, and an ActiV.A.C.™ therapy system manufactured by 3M™ Company (St. Paul, Minnesota). The NPWT dressing 20 can be, for example, a V.A.C. VERAFLO™ dressing, a V.A.C. VERAFLO™ large dressing, a V.A.C. VERAFLO™ CLEANSE dressing, and a V.A.C. VERAFLO CLEANSE CHOICE™ dressing manufactured by 3M™ Company (St. Paul, Minnesota).

[0036] In some examples, the NPWT device 14 can utilize gravity fluid flow from the fluid supply reservoir to the NPWT dressing 20 without using a pumping device. For example, the fluid flow device of the NPWT device 14 can be a valve (e.g., a solenoid actuated pinch valve) configured to control the fluid flow between the fluid supply reservoir and the NPWT dressing 20. In some examples, the negative pressure source can draw fluid from the fluid supply reservoir to the NPWT dressing 20, for example, without the aid of a gravity supply or pumping action from the fluid flow device.

[0037] In some examples, the negative pressure source can comprise a diaphragm vacuum pump. In some examples, the NPWT device 14 can include a filter or muffler coupled to the negative pressure source, for example, to reduce the operating noise of the negative pressure source and / or to filter the air exiting the negative pressure source.

[0038] The fluid flow device of the NPWT device 14 can comprise a pump, such as a peristaltic pump, a centrifugal pump, or other suitable pump. In other examples, the fluid flow device of the NPWT device 14 can comprise a gravity supply system, instead of (or in combination with) a pump, to deliver fluid to the NPWT dressing 20. For example, a valve between the gravity supply system and the NPWT dressing 20 can be used to restrict the fluid flow to the NPWT dressing 20 when a predetermined pressure is reached.

[0039] In some examples, the NPWT device 14 can also include a vent on the reservoir and a check valve configured to allow flow in the direction from the NPWT dressing 20 towards the negative pressure source and to restrict fluid flow in the reverse direction. The NPWT device 14 can also include a pressure sensor coupled to the NPWT dressing 20, as well as pressure sensors coupled to the negative pressure source and the wound dressing 20.

[0040] In some examples, the NPWT device 14 can operate in three modes. In a first mode, the negative pressure source can operate to generate a negative pressure applied to the NPWT dressing 20 while the fluid flow device of the NPWT device 14 is not operating. In a second mode, the negative pressure source may not operate, but the fluid flow device of the NPWT device 14 can operate to provide fluid flow to the NPWT dressing 20. In a third mode, neither the negative pressure source nor the fluid flow device of the NPWT device 14 may operate.

[0041] In an example of the operation of the NPWT device 14, the negative pressure source can be operative to generate a negative pressure applied to the reservoir and the NPWT dressing 20. The pressures in the negative pressure source as well as the reservoir and the NPWT dressing 20 can be monitored by one or more pressure sensors. When a desired level of negative pressure (e.g., -125 mmHg) is reached, the negative pressure source can be stopped and the vent can open the reservoir to vent, for example, to the atmosphere. In some examples, a check valve can maintain the negative pressure applied to the NPWT dressing 20, which can be monitored by one or more pressure sensors. In some examples, the check valve can be a duckbill type or a ball check type or a flap type valve.

[0042] The fluid flow device of the NPWT device 14 can then operate to initiate fluid delivery to the NPWT dressing 20. In some examples, the fluid flow device of the NPWT device 14 can be configured to flow at various rates, such as about 70-90 mL / min when the fluid is a liquid having a viscosity in the general range of saline, or lower in the case of certain gases such as oxygen. As fluid is pumped from the fluid flow device of the NPWT device 14 to the NPWT dressing 20, the pressure in the NPWT dressing 20 (which can be monitored by one or more pressure sensors) can be increased. When the NPWT dressing 20 reaches a predetermined pressure, a pressure sensor (which can be used to detect both positive and negative pressures) can send a control signal to a control device (e.g., a control switch or actuator) of the NPWT device 14 to restrict the fluid flow from the fluid flow device of the NPWT device 14 to the NPWT dressing 20. The increase in pressure in the NPWT dressing 20 can be used as an indicator that the fluid from the fluid flow device of the NPWT device 14 has sufficiently filled the NPWT dressing 20. By monitoring the pressure of the NPWT dressing 20 using one or more pressure sensors of the NPWT dressing 20 and / or the NPWT device 14, the NPWT device 14 can reduce the likelihood that the NPWT dressing 20 is overfilled. This can reduce fluid waste and leakage of the NPWT dressing 20 associated with overfilling. It will be understood that an interface circuit (not shown for simplicity of illustration) may be utilized to generate a sufficiently strong control signal and implement control logic.

[0043] The fluid flow device of the NPWT device 14 can be a valve (e.g., a solenoid-actuated pinch valve) that restricts fluid flow from a fluid supply reservoir or from a pump that can operate to provide fluid flow. The operation of the fluid flow device of the NPWT device 14 (e.g., the position of the valve or the operation / stop of the pump) can be automatically changed when a predetermined pressure is reached in the NPWT dressing 20. In some examples, the predetermined pressure in the NPWT dressing 20 at which the operation of the fluid flow device of the NPWT device 14 is changed can be about 1.0 mmHg (gauge pressure measured by a pressure sensor). In some examples, the predetermined pressure can be -10 to 10 mmHg, including values of -10, -9, -8, -7, -6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 mmHg, or any value in between.

[0044] In other examples, the user can monitor one or more pressure sensors of the NPWT dressing 20 and / or the NPWT device 14 and manually control the operation of the fluid flow device of the NPWT device 14 when the NPWT dressing 20 reaches a predetermined pressure. For example, the user may stop the fluid flow device of the NPWT device 14 by operating a control switch of the NPWT device 14, or may restrict the fluid flow from the fluid flow device of the NPWT device 14 by closing a valve.

[0045] When the NPWT dressing 20 is sufficiently filled with fluid, the user can continue with the desired fluid instillation and vacuum therapy treatment. In some examples, the NPWT device 14 and the NPWT dressing 20 can be used for the instillation cycle, which can provide an advantage for wound dressings on joints (e.g., knees) where the wound or dressing volume can be affected by the patient's body position. In some examples, by utilizing the foam NPWT dressing 20, the volume of the NPWT dressing 20 can change over time, partially due to the compressive deformation of the foam. For example, the volume of the foam may decrease over time as the foam is subjected to pressure. This change in the volume occupied by the foam can affect the volume of fluid required to fill the NPWT dressing 20. In some examples, the NPWT device 14 can utilize pressure measurements to indicate when the NPWT dressing 20 has received a sufficient volume of liquid. In some examples, the NPWT device 14 can utilize an absorption layer within the wound dressing instead of or in addition to a reservoir.

[0046] The sensor 16 can include, but is not limited to, a pressure sensor, a flow sensor, a resistor, a current, a voltage, and / or an impedance sensor (as described below with reference to FIG. 5), a temperature sensor, a humidity sensor, a blood pressure sensor, a heart rate sensor, a blood glucose sensor, an albumin and / or prealbumin sensor, an oxygen sensor, a tissue oxygen concentration sensor, an oxygenated hemoglobin value sensor, a biometric sensor including, but not limited to, these, a wound sensor, a body sensor, an accelerometer, a carbon dioxide sensor, a pH sensor, a patient activity sensor including, but not limited to, these, an analyte sensor configured to measure biomarkers, proteins, growth factors, cytokines, foreign DNA, microorganisms, etc., an optical sensor configured to measure the reflectivity, transmittance, absorbance, and / or opacity of light of one or more wavelengths of a wound and / or fluid, a viscosity sensor, a turbidity sensor, a specific gravity sensor, or any other suitable sensor, and can include any one or more of these.

[0047] Computing device 18 can be configured to process data and / or information from either the sensor 16 and / or the user device 6 and the server 24 and the database 8, and automatically control the NPWT device 14, for example, directly connected to the network 10 or via a connection to the network. For example, the computing device 18 can collect, gather, and / or collate patient information from a plurality of NPWT patients including patient 4. The computing device 18 can be configured to transmit patient information, for example of patient 4, to the user device 6, the server 24, the database 8, and / or any other suitable device or database.

[0048] In some examples, the computing device 18 can be configured to execute a treatment parameter unit 22. For example, the computing device 18 can be configured to execute the treatment parameter unit 22 after receiving patient information to determine one or more NPWT parameters and / or settings based on the patient information, for example one or more NPWT parameters. In other examples, the computing device 18 can be configured to receive one or more NPWT parameters and / or settings from another device that can output patient information of patient 4, for example, and execute the treatment parameter unit 22, and control the NPWT device 14 based on the received one or more NPWT parameters and / or settings. In other words, the treatment parameter unit 22 can be executed by a different device, for example, the user device 6, the server 24, or any other suitable device, and the computing device 18 can send and receive patient information, such as sending the information of patient 4 to another device that executes the treatment parameter unit 22, receive one or more NPWT parameters and / or settings determined by, for example, the treatment parameter unit 22, and be configured to control the NPWT device 14 to provide NPWT according to the received one or more NPWT parameters and / or settings.

[0049] The treatment parameter unit 22 can be configured to receive patient information and select at least one NPWT parameter and / or setting to control fluid at the wound site, for example, via the NPWT device 14 and the NPWT dressing 20. In some examples, the treatment parameter unit 22 is based on a causal model, such as a deep causal learning (DCL) model, that determines the current causal relationship between a set of NPWT parameters and / or settings and a set of effects by controlling fluid at the wound site, and can be configured to select at least one NPWT parameter and / or setting. In some examples, the treatment parameter unit 22 controls fluid at the wound site via the NPWT dressing 20 based on the selected at least one NPWT parameter setting, receives a measure of the effect by controlling fluid at the wound site, for example, via one or more of the sensors 16, and can be configured to adjust the DCL model based on the received measure of the effect by controlling fluid at the wound site. In some examples, the selected at least one NPWT parameter setting can include negative pressure level, negative pressure cycle, continuous application of negative pressure, fluid flow rate, fluid volume, fluid pressure, fluid temperature, fluid composition, fluid dwell time, and fluid purge time. In some examples, controlling fluid at the wound site can include providing fluid to the wound site via the NPWT dressing 20, for example, via the NPWT device 14. In some examples, the user can control fluid at the wound site based on information from the treatment parameter unit 22 that can be displayed by one or more devices, such as the user device 6, for example, via the NPWT device 14.

[0050] In some examples, the patient information includes at least one of user-entered patient information, patient biometric information, and wound measurement information, such as wound measurement information including a measure of the effect by controlling fluid at the wound site. In some examples, the wound measurement information and the measure of the effect by controlling fluid at the wound site can include at least one of a measured impedance value of the wound tissue, an oxygen measurement value of the wound bed, an oxygen measurement value of the fluid, a carbon dioxide measurement value of the wound bed, a temperature measurement value, an analyte sensor measurement value, and / or an optical measurement value of the wound bed. In some examples, the patient information can include an aggregation of patient information of multiple patients, for example, stored in the database 8. In some examples, the patient information includes any information and / or data from the sensor 16.

[0051] The user device 6 can be any suitable device suitable for communicating with a server or computing device of a company, organization, or institution, a personal computing device, a clinician's computer terminal, or the treatment system 12 and / or the network 10, and / or implementing any of the computing functions attributable to any computing device herein, such as the computing device 18.

[0052] In other examples, the network 10 can include a public network such as the Internet. Although shown as a single entity, the network 10 may include a combination of public networks and / or private networks. In some examples, the network 10 can include one or more of a wide area network (WAN) (e.g., the Internet), a local area network (LAN), a virtual private network (VPN), or another wired or wireless communication network.

[0053] Server 24 can be configured to collect, gather, and / or aggregate patient information and to execute treatment parameter unit 22 and, in some instances, to control NPWT device 14 and / or sensor 16, for example remotely. Server 24 can collect, gather, and / or aggregate patient information from one or more user devices 6, one or more computing devices 18, or any other suitable device and / or from direct input from a clinician and / or patient 4.

[0054] FIG. 2 is a block diagram showing an exemplary computing device 28 configured to execute a causal model in accordance with the techniques of the present disclosure. Computing device 28 can be an example of server 24 of FIG. 1 or computing device 18 of FIG. 1, can be included in server 24, can communicate with the server, or can be a device separate from server 24. The architecture of computing device 28 shown in FIG. 2 is shown for illustrative purposes only and computing device 28 should not be limited to this architecture. In other examples, computing device 28 can be configured in various ways.

[0055] As shown in the example of FIG. 2, computing device 28 includes one or more processors 30, one or more user interface (UI) devices 32, one or more communication units 34, and one or more memory units 36. The memory 36 of computing device 28 includes an operating system 38, a UI module 40, a telemetry module 42, and a treatment parameter unit 48 that are executable by processor 30. The various components, units, or modules of computing device 28 are coupled (physically, communicatively, and / or operably) using a communication channel for component - to - component communication. In some examples, communication channel 26 can include a system bus, a network connection, an inter - process communication data structure, or any other means for communicating data. Although shown separately in the example of FIG. 2, it will be understood that in other examples, memory 36 and processor 30 may be integrated into a single hardware unit such as a system - on - chip (SoC). Whether implemented as separate components or integrated, memory 36 and processor 30 provide a computer platform for executing operating system 38. Next, operating system 38 provides a multitasking operating environment for executing one or more software applications that computing device 28 can run.

[0056] In one example, the processor 30 can include one or more processors configured to implement functions and / or process instructions for execution within the computing device 28. For example, the processor 30 may be capable of processing instructions stored by the memory 36. The processor 30 can include, for example, a microprocessor, a single-core processor, a multi-core processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a processing circuit (e.g., a fixed-function circuit, a programmable circuit, or any combination of a fixed-function circuit and a programmable circuit), or an equivalent discrete or integrated logic circuit, or any combination of the foregoing devices or circuits.

[0057] The memory 36 can be configured to store information (e.g., data and / or executable instructions) within the computing device 28 during operation. The memory 36 can include a computer-readable storage medium or a computer-readable storage device. In some examples, the memory 36 can include one or more short-term memories or long-term memories. The memory 36 can include, for example, one or more of random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), on-chip memory (e.g., in the case of an SoC implementation), off-chip memory, magnetic disks, optical disks, flash memory, or forms of electrically programmable memory (EPROM) or electrically erasable programmable memory (EEPROM). In some examples, the memory 36 is used to store program instructions executed by the processor 30. The memory 36 can be used by software or an application operating on the computing device 28 (e.g., the therapy parameter unit 48) to temporarily store information during program execution.

[0058] The computing device 28 can communicate with external devices via one or more networks, such as network 10 in FIG. 1, or via wireless signals, using the communication unit 34. The communication unit 34 can be a network interface such as an Ethernet interface, an optical transceiver, a radio frequency (RF) transceiver, or any other type of device capable of transmitting and receiving information. Other examples of interfaces can include Wi-Fi (trademark), near field communication (NFC), or Bluetooth (registered trademark) wireless. In some examples, the computing device 28 wirelessly communicates with external devices such as user device 6 or account device 8 in FIG. 1 using the communication unit 34.

[0059] The UI device 32 can be configured to operate as both an input device and an output device. For example, the UI device 32 can be configured to receive tactile input, voice input, or visual input from a user of the computing device 28. In addition to receiving input from the user, the UI device 32 can be configured to provide output to the user using tactile stimuli, voice stimuli, or video stimuli. In one example, the UI device 32 can be configured to output content such as a graphical user interface (GUI) for display on a display device. The UI device 32 can include a presence detection display that displays a GUI and receives input from the user using capacitive, inductive, and / or optical detection at or near the presence detection display.

[0060] Other examples of the UI device 32 include a mouse, keyboard, voice response system, video camera, microphone, or any other type of device for detecting commands from a user, a sound card, a video graphics adapter card, or any other type of device for converting a signal into a suitable form understandable by a human or a machine. Additional examples of the UI device 32 include a speaker, a cathode ray tube (CRT) monitor, a liquid crystal display (LCD), an organic light emitting diode (OLED), or any other type of device capable of generating an output understandable by a user or a machine.

[0061] The operating system 38 controls the operation of the components of the computing device 28. For example the operating system 38 facilitates communication between, in one example, the UI module 40, the telemetry module 42, and the treatment parameter unit 48, and the processor 30, the UI device 32, the communication unit 34, and the memory 36. The UI module 40, the telemetry module 42, and the treatment parameter unit 48 can each include program instructions and / or data stored in the memory 36 and executable by the processor 30. For example, the treatment parameter unit 48 can include instructions for causing the computing device 28 to implement one or more of the techniques described in this disclosure.

[0062] The computing device 28 can include additional components not shown in FIG. 2 for clarity. For example, the computing device 28 may include a battery that powers the components of the computing device 28. Similarly, the components of the computing device 28 shown in FIG. 2 may not be required in all examples of the computing device 28 that are consistent with this disclosure.

[0063] In the example shown in FIG. 2, the treatment parameter unit 48 includes a deep causal learning (DCL) unit 50, patient information and DCL input 52, NPWT parameter settings 54, and wound measurements 56. The treatment parameter unit 48 can be an example of the treatment parameter unit 22 in FIG. 1.

[0064] The DCL unit 50 can be configured to execute a control system, such as the control system 100 of FIG. 3 described below, that selects NPWT parameter settings 54 applied to an NPWT system, such as the NPWT device 14 and the NPWT dressing 20. In some examples, the DCL unit 50 can be configured to execute a control system based on a causal model, such as the causal models 110 and / or 210 described below with reference to FIGS. 3 and 4. The DCL unit 50 can be configured to receive patient information and determine one or more NPWT parameters and / or settings based on the patient information, such as one or more NPWT parameters. In some examples, the DCL unit 50 can be configured to receive patient information and select at least one NPWT parameter and / or setting to control fluid at the wound site, for example, via the NPWT device 14 and the NPWT dressing 20 of FIG. 1. The DCL unit 50 can be configured to select at least one NPWT parameter and / or setting based on a causal model that determines the current causal relationship between a set of NPWT parameters and / or settings and a set of effects by controlling fluid at the wound site. In some examples, the DCL unit 50 can be configured to receive a measure of the effect of controlling fluid at the wound site, for example, via one or more of the sensors 16, and adjust the DCL model based on the received measure of the effect of controlling fluid at the wound site. In some examples, the at least one selected NPWT parameter setting can include a negative pressure level, a negative pressure cycle parameter (in the case of drip therapy), continuous application of negative pressure, fluid flow rate, fluid volume, fluid pressure, fluid temperature, fluid composition, fluid dwell time (in the case of drip therapy), and fluid purge time (in the case of drip therapy).

[0065] In some examples, the DCL unit 50 can be software, hardware, or a combination thereof configured to execute a causal model by execution of the DCL unit 50 by a computing device such as, for example, computing device 28, computing device 18, user device 6, server 24, or any suitable computing device. The DCL unit 50 can be configured to select at least one NPWT parameter and / or setting based on received patient information and / or other DCL inputs. For example, the patient information and / or additional DCL inputs may be utilized by the causal model during the execution phase and / or may update the causal model. Updating the causal model is part of the execution phase (i.e., when the DCL is operating / executing), and the experiments contribute to improving either the model (exploration) or patient outcomes (exploitation).

[0066] In some examples, the DCL unit 50 can be configured to determine and / or measure the current causal relationship between the NPWT parameter settings 54, the wound measurements 56, and the treatment targets 58 by executing a DCL model such as, for example, DCL models 110 and / or 210 described below with respect to FIGS. 3 and 4. In contrast, conventional ML and / or AI models and / or algorithms, or other models and / or algorithms, for example, computing systems and / or software configured according to classification, regression, dimensionality reduction, and / or clustering ML models, rely on determining the correlation between parameters and responses, but in some examples do not measure the causal effects and uncertainties regarding causal effects as configured to be done by the DCL unit 50. For example, the DCL unit 50 can be configured to determine and / or measure an effect measure between a parameter setting and a response and / or estimate a confidence interval of the true average effect and the effect measure of the parameter setting that represents the current level of uncertainty regarding the causal effect, as described below with reference to the causal model 110 of FIG. 3.

[0067] In some examples, the patient information and DCL input 52 can include patient health record data such as demographic information, e.g., age, gender, ethnicity, weight, body mass index, etc., personal and / or family medical history and co-morbidities, past and current diagnoses, e.g., diabetes, obesity, cardiovascular disease, cholesterol, blood pressure, etc., prescription medications including dosage and frequency of use, blood test results and values, genetic test results, allergies and allergy test results, and any other suitable patient health record data. The patient information and DCL input 52 can include situation inputs from a clinician or patient, e.g., current symptoms, results from past and recent wound examinations, e.g., wound images over time, measurements of wound shape over time (length, width, depth, volume), infections and suspected infections, infection history and test results regarding culture counts and speciation over time, frequency and history of wound dressing changes, NPWT treatment constraints (e.g., minimum and / or maximum negative pressure determined by a clinician), discomfort metrics (e.g., pain scores and / or evaluations over time, history of pain associated with dressing changes), patient activity schedules (e.g., rest / sleep schedule, work schedule, activity / exercise schedule), etc. The patient information and DCL input 52 can include any applicable established prognostic models, such as the current level of confidence of the clinician and / or patient regarding the current treatment.

[0068] Patient information and DCL input 52 can further include measured patient biometrics such as real-time measurement and recording of patient biometric information, e.g., body temperature, heart rate, blood pressure, systemic blood oxygenation, and / or any data / information from sensor 16, and patient activity by measurement, e.g., rest / sleep time, movement / accelerometer data that can measure / detect patient itching and / or scratching at or near a wound and / or dressing. Patient information and DCL input 52 can include wound-related measurements such as real-time and historical measurements of a wound and / or wound bed, e.g., impedance measurements of the wound and spatial mapping (which can indicate the current healing stage such as granulation tissue thickness, degree of epithelial coverage, inflammation, proliferation, repair, etc., transition of the healing stage, progress of healing, stalling of healing, etc.), measurements of oxygen, carbon dioxide, temperature, pH, etc. of the wound, analyte sensor measurements of biomarkers, proteins, growth factors, cytokines, foreign DNA, etc., and optical measurements such as reflection and / or absorption spectra and / or one or more wavelengths of the wound bed. Patient information and DCL input 52 can include real-time and historical measurements around and / or near the wound bed, e.g., impedance measurements and spatial mapping across the wound bed that can indicate, for example, wound bed depth, impedance measurements and spatial mapping around the circumferential segments of the wound bed, such as tunneling and undermining and their severity (e.g., relative size of the void space from tunneling and / or undermining) and / or maceration, and can detect subcutaneous features such as tunneling and undermining and their severity (e.g., relative size of the void space from tunneling and / or undermining) and / or maceration, and impedance measurements and spatial mapping around the circumferential segments of the wound bed, and optical measurements around the wound such as reflection and / or absorption spectra and / or one or more wavelengths.

[0069] In some examples, patient information and DCL input 52 can further include component measurements of ionic strength and / or the presence of analytes (e.g., biomarkers, proteins, growth factors, cytokines, foreign DNA, etc.), oxygen, carbon dioxide, temperature, pH measurements, viscosity, turbidity, and / or specific gravity of the collected fluid, optical measurements indicating transparency and / or opacity of the fluid, and optical measurements related to reagent-based assays, e.g., absorbance, fluorescence, luminescence measurements, etc., real-time and historical measurements of exudate and / or infusion fluid collection containers.

[0070] In some examples, the patient information and DCL input 52 can further include any patient information, measurements, and data aggregated from multiple patients. The patient information and DCL input 52 can further include criteria for classifying "raw measurement data", such as the current, impedance, voltage, etc. of sensors and / or detectors. The patient information and DCL input 52 can further include criteria for ranking the importance of the data included in the patient information and DCL input 52.

[0071] The NPWT parameter settings 54 can include any parameter settings that can be adjusted in an NPWT and irrigation / drip wound treatment system, such as in the NPWT device 14 and the NPWT dressing 20. For example, the NPWT parameter settings 54 can include the negative pressure levels of one or more pressure schedules and / or regimens, negative pressure cycles (in the case of drip irrigation), and / or "on" and "off" times (in the case of drip irrigation), duration and negative pressure level (continuous basis during drip intervals or in the case of irrigation). In some drip-based scenarios, the NPWT parameter settings 54 can include a stepped or discrete pressure schedule that includes different negative pressures for the same and / or different durations. The NPWT parameter settings 54 can include, for example, a negative pressure versus time waveform for a continuous pressure schedule in the case of irrigation treatment, such as a sine wave, triangular wave, square wave, or any other waveform. The NPWT parameter settings 54 can include the negative pressure pump-down time and pump-up time, such as the ramp rate, the overall length of the treatment, and the time between dressing changes. The NPWT parameter settings 54 can include drip parameters such as fluid delivery fill time and volume rate, fluid volume, fluid pressure, fluid dwell time, fluid purge time and volume rate, fluid composition (e.g., ionic strength and electrolyte composition, antibacterial agent, chemical curation agent, preservative, astringent, hormone, cytokine, anti-inflammatory agent, immune response activator, immune response inhibitor, presence and dosage of gases such as oxygen, carbon dioxide, etc.), fluid fill, dwell, and draw time profiles (e.g., slough, biofilm, high frequency dynamics that can mitigate and / or inhibit ingrowth of tissue into the dressing, fill / purge duty cycle, and / or waveforms such as sine wave, triangular wave, square wave, or any other waveform).

[0072] In some examples, the treatment parameter unit 48 can include one or more NPWT treatment targets or goals, such as treatment target 58. For example, the DCL unit 50 can control and automate a multivariate dynamic system to improve and / or optimize multiple treatment targets 58. In some examples, the DCL unit 50 can improve and / or optimize multiple treatment targets 58 in parallel and / or simultaneously, for example, for one or more patients.

[0073] Treatment targets 58 can include rapid granulation tissue formation, rapid epithelialization, reduction of wound healing stall, reduction of the occurrence of tunneling and undermining regions, rapid contraction of tunneling and undermining regions while reducing abscess formation, reduction of biofilm or infection, improvement of the quality of regenerated tissue, improvement of graft integration, reduction of ingrowth of tissue into the dressing, reduction of patient pain and discomfort, improvement of patient sleep and recovery, reduction of dressing changes (e.g., only when necessary), and reduction of maceration around the wound.

[0074] In some examples, a causal model and / or DCL algorithm, such as DCL unit 50, can transfer learning from a given patient to subsequent patients, as opposed to other machine learning (ML) and / or artificial intelligence (AI) algorithms, models, or schemes. A causal model and / or DCL algorithm, such as DCL unit 50, can determine causal and effect relationships between NPWT parameter settings 54, wound measurements 56, and treatment targets 58. In contrast, other ML and / or AI algorithms can determine correlations between NPWT parameter settings 54, wound measurements 56, and treatment targets 58 and may not be able to quickly (e.g., with fewer iterations) or accurately determine the exact NPWT parameter settings 54 to yield a desired treatment target 58. For example, other ML and / or AI algorithms may not be able to determine and / or measure the current causal relationships between NPWT parameter settings 54, wound measurements 56, and treatment targets 58, as described below with reference to FIG. 3 and NPWT settings 104 and responses 130. Thus, other ML and / or AI algorithms mainly rely on exploring determined correlations rather than determining, understanding, and / or measuring causal relationships and may spend computational time and / or resources exploring "weaker" correlations, e.g., correlations without a causal relationship between NPWT parameter settings 54, wound measurements 56, and treatment targets 58 and / or with low causal impact. For example, other ML and / or AI algorithms may not be configured to separate causal relationships from historical data and may not be configured to quantify the relationship with the accuracy required to facilitate decision-making, e.g., the fact that a particular NPWT parameter setting correlates with a wound measurement may not be sufficient to recommend changing that particular NPWT parameter or changing it by a particular amount. The larger the number of dimensions of a problem, such as the determination of NPWT parameter settings 54, and the number of features, the higher the likelihood of spurious correlations occurring and the much more difficult it becomes for any algorithm to identify the main variables, e.g., NPWT parameter settings 54, that actually and reliably promote a desired outcome.

[0075] In some examples, causal models and / or DCL algorithms such as DCL unit 50 can aggregate learning from multiple patients, thereby increasing statistical utility, and DCL unit 50 can create smaller aggregations / clusters that represent sub-populations over time and adjust NPWT parameter settings 54 for the sub-populations. In some examples, causal models and / or DCL algorithms such as DCL unit 50 can determine which context data results in different courses of action for a particular patient, e.g., for personalized medicine, e.g., different NPWT parameter settings 54.

[0076] FIG. 3 is a conceptual diagram showing a control system 100 for selecting NPWT parameter settings 104 applied to an NPWT system 102 according to one or more techniques of the present disclosure. Each of the NPWT parameter settings 104 defines a respective setting for a plurality of controllable elements of the NPWT system 102 and can be substantially the same as the NPWT parameter settings 54. Generally, the controllable elements of the NPWT system 102 are elements that can be controlled by the system 100 and can take on a plurality of different possible settings. The control system 100 can be implemented and / or executed, for example, by the DCL unit 50 described above.

[0077] During operation, the control system 100 can repeatedly select NPWT parameter settings 104 and monitor a response 130 to the control settings 104. The response 130 can be measured by the sensor 16. For example, the sensor 16 may measure the impedance of the wound, the sensor 16 may measure blood pressure, the sensor 16 may measure biomarkers, and the measurement values can be the response 130.

[0078] System 100 can calculate performance metrics based on response 130, for example, control NPWT system 102 to calculate a range of values including one or more single values representing the performance of the system when improving and / or maximizing the quality and / or effectiveness of the applied NPWT. The measure of quality used by control system 100 can be one or more predetermined values for response 130. An exemplary performance metric that combines all quality measurements used in the system is the weighted sum of the selected quality measurements.

[0079] As another example, the performance metric can be the weighted sum of the differences between each measure of quality and the baseline or desired value of the measure of quality, for example, to cause the system to attempt to reduce and / or minimize deviations outside the acceptable range of each measure of quality. For example, the performance metric may be the weighted sum of the difference between response 130 and treatment target 58. Another example of such a performance metric is the weighted sum of a function that is zero when the quality measurement is within the acceptable range and equal to the difference from the quality measurement to the nearest endpoint of the acceptable range when the quality measurement is outside the acceptable range.

[0080] Control system 100 can also monitor NPWT characteristics 140 of NPWT system 102. Generally, NPWT characteristics 140 can include any data that characterizes NPWT system 102 that can change the effect of NPWT parameter settings 104 on response 130 but is not considered in NPWT parameter settings 104, for example, data that cannot be controlled by control system 100. For example, NPWT characteristics 140 of NPWT system 102 can include environmental conditions during treatment, such as ambient humidity and / or ambient temperature of the environment where the patient is located during treatment. In some cases, these measures may be controllable by NPWT system 102, for example, NPWT system 102 may control ambient humidity and ambient temperature. In these cases, the adjustable measures will be included as NPWT parameter settings 104 rather than NPWT characteristics 140.

[0081] The control system 100 can use the response 130 to update a causal model 110 that models the causal relationship between the NPWT parameter setting 104 and the response 130. For example, rather than or in addition to there being a correlation between the NPWT parameter setting 104 and the response 130, the causal model 110 models how different settings of different elements affect the value of the response 130. In particular, for each NPWT parameter setting 104 of the NPWT system 102 and for each different type of response 130, the causal model 110 measures the causal effect of different possible NPWT parameter settings 104 on the response 130 and the current level of uncertainty of the control system 100 regarding the causal effect of the possible NPWT parameter settings 104. As a specific example, for each different possible NPWT parameter setting 104 of a given controllable element and for each different type of response 130, the causal model 110 includes an effect measure value representing the effect of the possible NPWT parameter setting 104 on the response 130 compared to other possible NPWT parameter settings 104 of the controllable element, e.g., an estimated value of the true average effect of the possible NPWT parameter setting 104, and a confidence interval of the effect measure value representing the current level of uncertainty of the control system 100 regarding the causal effect, e.g., a 95% confidence interval.

[0082] In some examples, the control system 100 calculates a confidence interval that specifies, for example, 95% of the upper and lower bounds of the effect of the NPWT parameter setting 104 on system performance, as indicated by the response 130. Specifically, this enables the control system 100 to identify cases where the selection of different NPWT parameter settings 104 results in (clinically) significant or non-significant differences.

[0083] If there is no causal relationship between one or more NPWT parameter settings 104 and one or more responses 130, the NPWT parameter settings 104 that do not form a causal relationship with respect to the change in the response 130 may not be updated or changed during the inspection. For example, the control system 100 may refrain from inspecting controllable elements that do not result in a significant difference. For example, as long as the upper and lower limits of the confidence interval of one or more specific NPWT parameter settings 104 do not indicate a clinically meaningful difference even for the maximum effect, and as long as there is a cost to continue inspecting / exploring one or more specific NPWT parameter settings 104, the control system 100 can omit those one or more specific NPWT parameter settings 104 from further exploration / experimentation / iteration and / or can indicate that it can be omitted. For example, as long as there is a cost, such as a health-related, time, money, power, or any other cost, to continue the experiment / iteration regarding one or more specific NPWT parameter settings 104, the control system 100 may stop and / or indicate to stop the experiment / iteration regarding those specific NPWT parameter settings 104 because the cost may exceed any benefit of controlling the NPWT system 102 to improve the performance metric and / or to determine a further causal relationship between the NPWT parameter settings 104 and the response 130.

[0084] In contrast, conventional ML and / or AI models and / or algorithms, or other models and / or algorithms can be correlation-based models rather than causal models. For example, classification, regression, dimensionality reduction, and / or clustering ML models rely on determining the correlation between parameters and responses, but do not measure causal effects and the uncertainty regarding causal effects, such as the effect measure between a parameter setting and a response, and / or the estimated value of the true average effect of a parameter setting, and the confidence interval of the effect measure representing the current level of uncertainty regarding the causal effect.

[0085] In some examples, before initiating control of the NPWT system 102, the control system 100 can receive an external input 106. The external input 106 can include data received by the control system 100 from any of a variety of sources. For example, the external input 106 can include data received from a user of the control system 100, data generated by another control system that previously controlled the NPWT system 102, data generated by a machine learning model, data generated by one or more sensors 16, or some combination thereof.

[0086] Generally, the external input 106 at least specifies possible initial values (e.g., NPWT parameter settings 104) for settings of controllable elements of the NPWT system 102 and which responses 130 the control system 100 is to track during operation.

[0087] For example, the external input 106 can cause the control system 100 to track measurements of a particular sensor 16 of the NPWT system 102, performance metrics such as merit indices, any other objective function derived from particular sensor measurements and improved and / or optimized by the control system 100 while controlling the NPWT system 102, or both.

[0088] The control system 100 can use the external input 106 to generate an initial possible set of values for the controllable elements, for example, an initial probability distribution (the "baseline probability distribution") over the NPWT parameter settings 104. By using the external input 106 to initialize these baseline probability distributions, the control system 100 can constrain the selection of the NPWT parameter settings 104 to not violate any constraints imposed by the external data 106 and, optionally, to not deviate from the historical range of NPWT parameter settings 104 previously used to control the NPWT system 102. For example, if there are certain ranges of NPWT parameter settings 104 known to be harmful, the external input 106 can define those ranges so that the control system 100 does not select NPWT parameter settings 104 within those ranges.

[0089] The control system 100 can also use the external input 106 to initialize a set of internal parameters 120, for example, to assign baseline values to the internal parameters 120. Generally, the internal parameters 120 define how the control system 100 selects the NPWT parameter settings 104, assuming the current causal model 110, for example, assuming the current causal relationships determined by the control system 100 and the system uncertainty regarding the current causal relationships. The internal parameters 120 can also define how the control system 100 updates the causal model 110 using the received response 130.

[0090] The control system 100 can update at least some of the internal parameters 120 while updating the causal model 110. That is, during the operation of the control system 100, one or more of the internal parameters 120 may be fixed to an initialized baseline value, but the control system 100 can repeatedly adjust other internal parameters 120 during operation to enable the control system 100 to measure more effectively and, in some cases, utilize causal relationships. For example, to control the NPWT system 102 during operation, the control system 100 may repeatedly identify procedural instances within the NPWT system 102 based on the internal parameters 120.

[0091] Each procedural instance can be a set of one or more entities within the NPWT system 102 that can be associated with a time frame. Entities within the NPWT system 102 can be subsets of the NPWT system 102, such as appropriate or inappropriate subsets. In particular, an entity can be a subset of the NPWT system 102 that can obtain a response 130 and be affected by an applied control setting, such as an NPWT parameter setting 104. For example, if the NPWT system 102 includes multiple physical entities that can obtain measurements from the sensor 16, a given procedural instance can include an appropriate subset of the physical entities to which a set of NPWT parameter settings 104 can be applied. The number of subsets into which the entities within the NPWT system 102 can be divided can be defined by the internal parameters 120.

[0092] The way the control system 100 divides entities into subsets at any given point during the operation of the control system 100 can be defined by internal parameters 120 that define the spatial extent of the NPWT parameter settings 104 applied by the control system 100 to the instance. The spatial extent of the instance identifies a subset of the responses 130 that can be assigned to the instance and, for example, enables the responses 130 obtained from that subset to be associated with the instance.

[0093] For example, a procedural instance can include one or more machines that operate using the NPWT parameter settings 104. The spatial extent can define the number and type of machines of the procedural instance. The control system 100 can obtain a response 130 to the NPWT parameter settings 104 selected for a given group of machines. For example, as described above, the control system 100 can select NPWT parameter settings 104 related to one or more measurements of one or more sensors 16, and then the control system 100 can use the selected NPWT parameter settings to track the selected performance metrics of the NPWT treatment.

[0094] The length of the time frame associated with the entity of any given procedural instance can be further defined by the internal parameter 120. In particular, the time frame that the control system 100 can assign to any given procedural instance can be defined by the internal parameter 120 that defines the time range of the NPWT parameter settings 104 applied by the control system 100. This time frame, for example, the time range of the instance, can define whether a future response 130 that the control system 100 can determine is caused by the NPWT parameter settings 104 selected for the procedural instance.

[0095] Since the internal parameter 120 changes during the operation of the control system 100, the instances generated by the system 100 can also change. That is, the control system 100 can change the way in which a procedural instance is identified when the control system 100 changes the internal parameter 120. The control system 100 can then select the NPWT parameter settings 104 for each instance based on the internal parameter 120 and optionally based on the NPWT characteristics 140.

[0096] In some examples, such as when the control system 100 is exploring the space of possible NPWT parameter settings 104, the control system 100 can select the NPWT parameter settings 104 for all of the instances based on the baseline probability distribution. In other examples, for instance, when the control system 100 is leveraging causal relationships that have already been determined to improve and / or optimize the objective function, the control system 100 can select the NPWT parameter settings 104 for some of the instances (e.g., "hybrid instances") using the current causal model 110 while continuing to select the NPWT parameter settings 104 for other instances (e.g., "baseline instances") based on the baseline probability distribution. More specifically, at any given point in time during the operation of the control system 100, the internal parameter 120 can define the ratio of hybrid instances to the total number of instances. The control system 100 can also determine, for each instance, based on the internal parameter 120, which response 130 can be associated with the instance, e.g., which to use when updating the causal model 110.

[0097] Next, for each instance, the control system 100 can select and / or set the NPWT parameter settings 104 and monitor the response 130. The control system 100 maps the response 130 to an instance-specific impact measurement value and determines a causal model update 150 that can be used to update the current causal model 110 using the impact measurement value. The control system 100 can determine, based on the internal parameters 120, which historical procedural instances (and associated responses 130) can be considered by the causal model 110, and can determine the causal model update 150 based only on these determined historical procedural instances. A set of internal parameters 120 that define a data inclusion window may determine the historical procedural instances considered by the causal model 110. The data inclusion window can specify, at any given point in time, one or more historical time frames in which a procedural instance, e.g., the response 130 associated with that procedural instance, should occur in order for that procedural instance to be considered by the causal model 110. Updating the causal model 110 can be part of the execution phase (i.e., when the causal model 110 is operating and / or being executed), and the experiment can contribute to improving the model (exploration) or patient outcomes (utilization).

[0098] For those internal parameters 120 that are being changed by the control system 100 during operation, the control system 100 can periodically generate internal parameter updates 160 based on the causal model 110, for example, to update the internal parameters 120 maintained by the control system 100. In other words, when the causal model 110 changes during the operation of the control system 100, the control system 100 can also update the internal parameters 120 to reflect the change in the causal model 110. In some examples, such as when the control system 100 assigns some NPWT parameter settings 104 to utilize the current causal model 110, the control system 100 can also use the difference in system performance between the "hybrid" instance and the "baseline" instance to determine the internal parameter updates 160.

[0099] FIG. 4 is a conceptual diagram showing an exemplary causal model 210 according to one or more techniques of the present disclosure. The causal model 210 can be an example of the causal model 110 and can be executed by hardware and / or software such as the DCL unit 50. In some examples, the causal model 210 includes some processes, and in other examples, the causal model 210 can include other processes, data, inputs, outputs, feedback, interactions, etc. In the illustrated example, the causal model 210 includes an experimental unit generator 212, a treatment assignment unit 214, a search unit 216, a baseline monitor 218, a data inclusion window unit 220, and a clustering unit 222.

[0100] The experimental unit generator 212 can be configured to generate spatio-temporal experimental units. The experimental units can be the minimum spatio-temporal extent that prevents the current causal relationship between the generated causal knowledge, e.g., parameters such as NPWT parameter settings 104 and / or 54, and the set of effects by controlling systems such as the NPWT system 102 and / or the treatment system 12, from deteriorating due to carry-over effects. In some examples, to buffer the carry-over effects from previous experimental units, the experimental unit generator 212 can limit the data recording of the causal model to the second half of the time range of each experimental unit. Additionally, for each independent variable (which can be, for example, one or more NPWT parameter settings 104 and / or 54), the experimental unit generator 212 can systematically vary and / or explore the spatio-temporal extent of the experimental unit, first within and then outside the minimum and maximum of the spatio-temporal extent and / or constraint data, to determine an improved and / or optimal experimental unit size corresponding to an average effect size within the 95% confidence interval (p = 0.05) from the asymptotic mean effect for a large spatio-temporal extent.

[0101] The treatment assignment unit 214 can assign treatments to the experimental units. For example, the treatment assignment unit 214 may assign treatments to the experimental units according to a predetermined procedure such as randomization without replacement and cancellation. The treatment assignment unit 214 can also assign independent variable levels to probabilistically equivalent experimental units using the constraint that the relative frequency of the assignment matches the relative frequency specified by one or more exploration and / or utilization trade-offs. In some examples, the treatment assignment unit 214 can calculate the D-score and the confidence interval around the D-score for each independent variable level by taking the difference between the average effect when "on" and the average effect when "off" over a data inclusion window (described later) so as to provide an unbiased estimate of the causal effect of the independent variable level / treatment assignment on the utility function.

[0102] The exploration unit 216 can determine whether to allocate experimental units towards making probabilistically optimal decisions and / or towards improving the accuracy of probability estimates. For example, the exploration unit 216 can determine whether to allocate experimental units by probability matching. In some examples, the exploration unit 216 can change the aggressiveness of the exploration / exploitation ratio, control the exploration / exploitation ratio, and determine an aggressiveness that improves and / or maximizes a utility (e.g., including reducing and / or minimizing disappointment) as measured and / or monitored by a baseline monitor 218 (described below). In some examples, when the cost of performing a treatment (including opportunity cost) is non-uniform between independent variable levels, a Bonferroni-corrected confidence interval can be calculated such that more evidence is required to utilize a more costly treatment.

[0103] The baseline monitor 218 can determine, for example, by statistical power analysis, the number of baseline experimental units required to monitor the performance difference between the baseline trials and treatment assignments. The baseline monitor 218 can assign independent variable levels randomly sampled according to the normative operating range data to the baseline experimental units. The difference between the baseline trials and the exploration / exploitation trials can provide an unbiased measure of the utility of the internal parameters (e.g., clustering, data inclusion window, exploration / exploitation aggressiveness) of the causal model 210, enabling such parameters to be objectively adjusted. The baseline trials can also enable exploration of the entire search space defined by the hard constraints.

[0104] The data inclusion window unit 220 can use factorial analysis of variance (ANOVA) for blocked time ranges to analyze the influence of blocked time ranges on the strength and direction stability of the interaction between independent variables and utility functions. For example, for each independent variable, the data inclusion window unit 220 can identify a Pareto optimal data inclusion window that improves and / or maximizes both the experimental test output (across all experimental unit clusters and the entire decision search space) and the statistical significance of the causal effect. In some examples, such an inclusion window can prevent the causal model 210 from overfitting the data and keep the causal model 210 highly responsive to dynamic changes in the underlying system structure.

[0105] The clustering unit 222 can manage dimensions. For example, thereby, the causal model 210 can learn a method of conditionally assigning independent variable levels based on the factor interaction between the effect of the independent variable level and the attributes of the experimental units that cannot be manipulated by the control system (e.g., gender, age, comorbidities, weather, demand, etc.). The clustering unit 222 can pool experimental units into clusters where the similarity within the cluster of the influence of the independent variable on the utility is maximized and the difference between clusters is maximized. In some examples, the clustering unit 222 can use factorial ANOVA to find factors that explain the maximum amount of variance between clusters and use stepwise statistical test output analysis to select some factors that result in clusters having sufficient statistical test output to find available effects. In some examples, the clustering unit 222 can control the clustering decision by continuously examining the clustering decision and using baseline monitoring to objectively explore and utilize the influence of the clustering decision on the utility.

[0106] FIG. 5 is a flowchart of an exemplary method of selecting at least one NPWT parameter of a wound treatment system by one or more techniques of the present disclosure. FIG. 5 is described using the computer-based system 2 of FIG. 1, the computing device 28 of FIG. 2, and the control system 100 of FIG. 3, but it should be understood that the methods described herein may include and / or utilize other systems and methods in other examples.

[0107] The computing device 28 can receive patient information (302). For example, the computing device 28 may receive patient information and DCL input 52. In some examples, the patient information may include user-entered patient information, patient biometric information, impedance measurements of wound tissue, oxygen measurements of the wound bed, oxygen measurements of the fluid, carbon dioxide measurements of the wound bed, temperature measurements, analyte sensor measurements, or optical measurements of the wound bed, each of which may be for an individual patient or an aggregation of multiple patients.

[0108] The DCL unit 50 can select at least one NPWT parameter setting 54 and / or 104 for controlling fluid at the wound site based on a causal model that determines the current causal relationship between the set of NPWT parameters settings 54 and / or 104 and the set of effects by controlling fluid at the wound site, such as causal model 110 (304). In some examples, the causal model, such as causal model 110, can measure the current causal relationship between the set of NPWT parameter settings 54 and / or 104 and the set of effects by controlling fluid at the wound site, for example, by measuring the causal effect of different possible NPWT parameter settings 104 on the response 130 and the current uncertainty level of the control system 100 regarding the causal effect of the possible NPWT parameter settings 104, as described above with reference to FIG. 3. As used herein, "fluid" refers to any substance that deforms or "flows" when subjected to one or more external forces (e.g., pressure, gravity, etc.). The fluid can be used in accordance with various techniques such as NPWT described herein, regardless of whether it is in a liquid state (e.g., saline, distilled water, etc.), a gaseous state (e.g., a mixture of gases such as air, a gaseous element such as pure oxygen), a plasma state, or other states.

[0109] The computing device 28 can control fluid at the wound site via the NPWT dressing 20 based on the selected at least one NPWT parameter setting 54 and / or 104 (306). For example, the computing device 28 may operate the NPWT device 14 in accordance with the at least one selected NPWT parameter setting 54 and / or 104.

[0110] The computing device 28 and the DCL unit 50 can receive a measure of the effect by controlling fluid at the wound site (308). For example, the DCL unit 50 may receive one or more measurements from one or more sensors 16, such as an impedance measurement system that may be included in the NPWT wound dressing 20 or alternatively in the wound site, to measure the impedance of the wound.

[0111] The DCL 50 can adjust the causal model (310). For example, the DCL 50 may generate a causal model update 150 and an internal parameter update 160 and adjust the causal model for update based on the received measure of the effect by controlling fluid at the wound site. The DCL 50 can update and / or adjust the width of the confidence intervals of the estimated values of the cause and effect, such as the upper and lower limits of the influence of the NPWT parameter settings 104 on the system performance.

[0112] The DCL unit 50 can select at least one second NPWT parameter setting 54 and / or 104 to control fluid at the wound site via the NPWT dressing 20 (312). For example, the previously selected NPWT parameter settings 54 and / or 104 can include a selection of a plurality of settings, and the second NPWT parameter settings 54 and / or 104 can be different parameters that can similarly include a selection of a plurality of settings. In some examples, the DCL unit 50 can select at least one second NPWT parameter setting 54 and / or 104 based on the causal model, such as the adjusted and / or updated causal model 110, thereby determining the current causal relationship (which can be different from before adjustment) between the set of NPWT parameter settings 54 and / or 104 and the set of effects by controlling fluid at the wound site.

[0113] Computing device 28 can control fluid at the wound site via NPWT dressing 20 based on at least one selected second NPWT parameter setting 54 and / or 104 (314). The computing device 28 can then repeat steps (308)-(314) to continue both, for example, adjusting and / or updating the causal model 110 and controlling the NPWT system 102 by selecting the NPWT parameter settings 54 and / or 104 based on the adjusted and / or updated causal model 110 (316).

[0114] FIG. 6 is a plot 600 of exemplary complex impedance of a tissue site at a predetermined frequency F1 measured at multiple time points after a wound occurs at the tissue site. The example shown in FIG. 6 shows the progression of the impedance of the tissue site at a predetermined frequency F1 as the wound progresses through multiple healing stages. In the illustrated example, impedance measurements 602-612 correspond to the average value of impedance measurements at eight wound tissue sites at frequency F1 and correspond to six times after wound occurrence, for example, day 0, day 3, day 7, day 10, day 14, and day 16. Plot 600 is a plot of resistance versus reactance in which each average impedance measurement 602-612 averaged over eight wound tissue sites is represented by a pair of resistance (i.e., the real component of its complex impedance) and reactance (i.e., the imaginary component of its complex impedance) values. In other words, plot 600 is a two-dimensional (2D) plot of average impedance measurements at different times, and the dimensions of plot 600 are resistance and reactance. In the illustrated example, the average impedance measurements 602-612 are values at frequency F1, for example 85 kHz.

[0115] In the illustrated example, the progression of impedance over time on plot 600 follows changes in resistance and reactance values that are similar to vectors in "paths" 622 and 624, e.g., in a 2D resistance-reactance plane. Paths 622 and 624 may correspond to changes in the impedance of a wound tissue site as the wound tissue progresses through different healing stages. In the illustrated embodiment, the wound tissue site may begin, for example, around the time of the occurrence of a wound that does not exhibit tissue growth at stage 1 (inflammation), and the electrical properties of the wound tissue may be mainly resistive with little reactance. The resistance / reactance values of impedance measurements 602 and 604 at day 0 and day 3 respectively from the occurrence of the wound may include information indicating that the wound tissue site is in the first stage, e.g., the inflammatory stage of healing.

[0116] As the wound site heals and progresses to stage 2 (proliferation, e.g., granulation), the impedance can "follow" path 622 where the reactance remains near 0 and the resistance decreases. In other words, the wound tissue site may change from being substantially resistive during stage 1 to increased conductivity. In some embodiments, the transition from stage 1 to stage 2 may be determined based on a change in the resistance of the wound tissue site below a threshold R1. In the illustrated example, the resistance value of impedance measurement 606 corresponding to day 7 is less than the threshold resistance R1 and has a reactance near 0, indicating that the wound tissue site has progressed from stage 1 to stage 2. In some embodiments, the impedance of the wound tissue site may remain relatively conductive during stage 2. In the example shown in FIG. 6, the values of impedance measurements 606, 608, and 610 corresponding to day 7, day 10, and day 14 respectively have relatively low resistance values and reactance values near 0, e.g., relatively close to the origin of plot 600. The resistance / reactance values of impedance measurements 606, 608, and 610 may include information indicating that the wound tissue site is in the second stage, e.g., the proliferative stage of healing.

[0117] As the healing of the wound tissue site progresses to stage 3, for example, remodeling / re-epithelialization, the impedance of the wound tissue site can "change" to a negative reactance value gradually away from the origin. For example, the measured impedance value of the wound tissue site can follow a path 624 whose direction changes such that it becomes substantially negative along the reactance axis. In the illustrated example, the impedance measurement value 612 includes a relatively low resistance value and a gradually negative reactance value. The resistance / reactance value of the impedance measurement value 612 corresponding to the 16th day may include information indicating that the wound tissue site is in the 3rd stage, for example, the repair stage of healing. During the 3rd stage, the impedance may remain on the path 624. For example, as more epithelium grows on the wound tissue site, the wound tissue site can transition from substantial conductivity to substantial capacitance (indicating negative reactance). In some examples, when the wound tissue site is completely re-epithelialized, the electrical properties of the tissue site, such as impedance, are the same as those of the non-wound tissue before and / or near the wound, as shown on the plot 600 as the impedance 614 at the 4th completely healed stage.

[0118] Figures 7-11 show exemplary scenarios of a DCL-based feedback control type wound treatment system. In each of Figures 7-11, a causal model, such as the causal model 110 executed by the computing system 28 via the DCL unit 50, can determine the NPWT parameter settings 54 and / or 104 based on the current causal relationship between the NPWT parameter settings 54 and / or 104 and the effects of controlling the fluid at the wound site, such as those measured by various sensors 16. In particular, the effects of controlling the fluid, such as healing and / or the healing stage, can be measured by a wound bed impedance measurement system that can be integrated into or added to the NPWT dressing 20. Figures 7-11 include plots showing impedance measurement values that can indicate the healing and / or healing stage of the wound site, in conjunction with the applied NPWT parameter settings.

[0119] FIG. 7 is a diagram of an exemplary scenario 700 of wound measurement values and NPWT parameter settings over a period of time according to the techniques described in the present disclosure. Scenario 700 includes complex impedance measurement plots 702, 704, and 706 at different times, a wound bed resistance plot 708, and NPWT parameter settings 710 for vacuum pressure.

[0120] In some examples, a DCL-based feedback control type wound treatment system, such as system 2 and / or control system 100, can use real-time measurements of wound bed impedance (along with other factors such as patient information and DCL input 52, wound measurement values 56, and treatment targets 58) to adjust the on / off state of negative pressure (as well as other NPWT parameter settings 54 and / or 104). In some examples, control system 100 can operate within defined constraints, for example, it is only possible to adjust the negative pressure at two levels of 0 mmHg or -125 mmHg, increase or decrease the pressure at a constant and non-adjustable rate, and cycle the negative pressure on and off in cycles with a finite duration. In the illustrated example, the performance metric can be a decrease and / or minimization of wound bed resistance below a threshold resistance (Rgran) value indicating the start of granulation, and can be related to treatment target 58 progressing from the inflammatory phase to the proliferative phase. Control system 100 can make control decisions regarding the negative pressure level 710, for example, it can switch between -125 mmHg and 0 mmHg in the illustrated example when the value of wound bed resistance has not decreased. When the wound bed begins to granulate, as indicated by the wound bed resistance decreasing below the Rgran threshold, control system 100 can determine to apply a constant negative pressure of -125 mmHg.

[0121] FIG. 8 is a diagram of another exemplary scenario 800 of wound measurement values and NPWT parameter settings over a period of time according to the techniques described in this disclosure. Scenario 800 includes complex impedance measurement plots 702-706 and 802 at different times, a wound bed resistance plot 708, a wound bed reactance time derivative plot 804, NPWT parameter settings 710 for vacuum pressure, and an electrical stimulation control parameter plot 806. In the illustrated example, scenario 800 is a continuation of scenario 700 over time.

[0122] In the illustrated example, when the wound bed is filled with an appropriate amount of granulation tissue (e.g., in two weeks as shown), the control system 100 can interrupt the NPWT treatment (e.g., plot 710 returns to 0 mmHg in two weeks) and initiate another treatment, such as a treatment that may be appropriate for re-epithelialization, such as electrical stimulation or oxygen delivery (e.g., plot 806 transitions from the off state to the on state in two weeks). The control system 100 can automatically determine such treatment changes based on the current sensor data and history data set and previous learning from other patients.

[0123] In the illustrated example, the control system 100 can operate within the same defined constraints as described above, e.g., it is only possible to adjust the negative pressure at two levels of 0 mmHg or -125 mmHg, increase or decrease the pressure at a constant and non-adjustable rate, and cycle the negative pressure on and off in finite durations. In the illustrated example, after the granulation tissue fills the wound cavity space along the time axis of the plot in about two weeks, epithelialization begins and the complex impedance measurement values can include an increasing negative reactance value 808. The increasing negative reactance value 808 can indicate, for example, a reduction in the wound bed reactance time derivative 804 as shown in two weeks. The control system 100 can then determine to modify the control parameters by shutting off the negative pressure.

[0124] In some examples, the control system 100 can include other treatment systems for increasing and / or improving re-epithelialization, such as electrical stimulation or oxygen delivery. The control system 100 can, for example, continue to administer treatment by the other treatment system when used in conjunction with NPWT, or determine to begin administering these other treatments when the NPWT treatment is stopped, based on a decrease in the time derivative of the reactance 804 indicating the progression of the wound from the granulation / growth phase to the re-epithelialization / repair phase.

[0125] FIG. 9 is a diagram of another exemplary scenario 900 of wound measurements and NPWT parameter settings over a period of time according to the techniques described in this disclosure. Scenario 900 includes a wound bed resistance plot 708, a wound bed reactance time derivative plot 804, an NPWT parameter setting 710 of vacuum pressure, a peripheral and / or periwound moisture vs. time plot 902, and an NPWT parameter setting 904 of wound drip / irrigation fill level.

[0126] In some examples, excessive moisture in the wound area can result in maceration of the tissue surrounding the wound, e.g., deterioration of the tissue surrounding the wound that can increase the wound size over time. An NPWT system that provides wound irrigation or periodic wound drip to promote granulation can cause maceration of the tissue surrounding the wound if not controlled and / or not taken into account. In the illustrated example, the control system 100 further controls the NPWT parameter settings to reduce and / or eliminate maceration of the tissue surrounding the wound.

[0127] For example, based on impedance measurements and other patient information, the control system 100 can initiate a combination cycle of vacuum / drip / dwell at time T1 to improve granulation. The control system 100 can flow fluid to the wound during the pressure-off time, as indicated, for example, by a 100% drip / irrigation fill level 904 when the vacuum pressure 710 is 0 mmHg and a 0% drip / irrigation fill level 904 when the vacuum pressure 710 is -125 mmHg.

[0128] In the illustrated example, the control system 100 can improve granulation, for example, as indicated by a decrease in the wound resistance 708 between times T1 and T2. The moisture sensor 16 can measure the moisture of the tissue around the wound, and the moisture measurement value 902 around the wound can indicate that the tissue around the wound is approaching moisture saturation from T1 to T2. Then, the control system 100 can modify the treatment, for example, by shutting off fluid perfusion / drip. In addition, the control system 100 can change the duty cycle of the vacuum pressure 710, for example, by increasing the duration of the pressure of -125 mmHg, which can help draw up moisture from the wound area.

[0129] In the illustrated example, the moisture 902 around the wound decreases after T2, and the wound resistance 708 decreases below a threshold (e.g., Rgran) between times T2 and T3, indicating, for example, that the wound has progressed to the granulation stage, and the time derivative 804 of the wound resistance drops below the threshold, for example, at time T3. The control system 100 can determine to end the negative pressure treatment cycle and provide a continuous / constant vacuum negative pressure 710 at time T3.

[0130] The control system 100 can determine to terminate the NPWT treatment at time T4. For example, the control system 100 may determine that the wound has entered the re-epithelialization / repair stage based on a decrease in the time derivative 804 of the wound reactance at or near time T4. Although described by way of example with respect to drip, it should be understood that the various aspects of the technology described herein with respect to FIG. 9 are also applicable to cases of wound perfusion where the wound is simultaneously washed by a flowing fluid and a negative pressure is applied.

[0131] FIG. 10 is a diagram of another exemplary scenario 1000 of wound measurement values and NPWT parameter settings over a period of time according to the techniques described in the present disclosure. Scenario 1000 includes a wound bed resistance plot 708, a wound bed reactance time derivative plot 804, an NPWT parameter setting 710 for vacuum pressure, a microbial sensor plot 1002, and an NPWT parameter setting 904 for wound drip / irrigation fill level.

[0132] In some examples, during wound treatment, the wound may become infected. The control system 100 can provide rapid infection detection and treatment. For example, based on impedance measurements and other patient information, the control system 100 may initiate a combination cycle of vacuum / drip / retention at time T1 to improve granulation. The control system 100 can flow standard drip fluid into the wound during pressure-off times, as indicated, for example, by a 100% drip fill level 904 when the vacuum pressure 710 is 0 mmHg and a 0% drip fill level 904 when the vacuum pressure 710 is -125 mmHg.

[0133] Between times T1 and T2, microbial sensor data 1002 from measurements of the drip fluid collected in the collection container of the NPWT system can indicate an increasing probability of infection. The control system 100 can determine to inject an antibacterial agent into the standard drip fluid at time T2. Microbial sensor data 1002 from measurements of the perfusion / drip fluid collected in the collection container of the NPWT system can indicate a decreasing probability of infection after time T2. The control system 100 can determine to stop the antibacterial agent drip cycle at time T3.

[0134] In some examples, the control system 100 can inject any suitable agent, such as a chemical curing agent, a preservative, an astringent, a hormone, a cytokine, an anti-inflammatory agent, an immune response activator, an immune response inhibitor, and an ionized gas, into the drip fluid, either separately from or in addition to the antibacterial agent. Although described by way of example with respect to drip infusion, it should be understood that the various aspects of the techniques described herein with respect to FIG. 10 are also applicable to cases of wound perfusion where the wound is simultaneously washed with a flowing fluid and a negative pressure is applied.

[0135] FIG. 11 is a diagram of another exemplary scenario 1100 of wound measurements and NPWT parameter settings over a period of time according to the techniques described in the present disclosure. Scenario 1100 includes a wound bed resistance plot 708, a wound bed reactance time derivative plot 804, an NPWT parameter setting 710 for vacuum pressure, a patient input pain level 1102, and an NPWT parameter setting 904 for wound drip infusion / perfusion fill level.

[0136] In some examples, the control system 100 can provide a customized treatment. For example, based on impedance measurements and other patient information, the control system 100 may initiate a combined cycle of vacuum / perfusion at time T1 to improve granulation. The control system 100 can flow a standard drip fluid into the wound during the pressure-off time, as indicated, for example, by a 100% drip fill level 904 when the vacuum pressure 710 is 0 mmHg and a 0% drip fill level 904 when the vacuum pressure 710 is -125 mmHg.

[0137] In some examples, a patient can input a pain level over time to the control system 100. In the illustrated example, the patient input pain level 1102 increases between times T1 and T2. In response, the control system 100 can reduce, for example, the negative pressure used during the vacuum on cycle at time T2 from -125 mmHg to -100 mmHg such that the cycle level is -100 mmHg when the vacuum is on and 0 mmHg when the vacuum is off, as illustrated. In some examples, the return to ambient pressure, for example 0 mmHg, during the vacuum off time may cause and / or be associated with pain, and the control system 100 can adjust the negative pressure used during the vacuum off cycle at time T2, for example, from 0 mmHg to -25 mmHg such that the cycle level is -125 mmHg (or -100 mmHg) when the vacuum is on and -25 mmHg when the vacuum is off (not shown). Additionally, the control system 100 can exchange the infusion fluid for a fluid containing an analgesic at time T2. In some examples, the control system 100 can adjust from infusion to perfusion, for example, by providing a fluid and / or a fluid containing an analgesic at a specific flow rate over a period of time.

[0138] In the illustrated example, the patient input pain level 1102 decreases after T2, and the wound bed resistance 708 decreases below a threshold (e.g., Rgran) between times T2 and T3, indicating, for example, that the wound has progressed to the granulation phase. The control system 100 can then end the negative pressure treatment cycle and provide a sustained / constant vacuum negative pressure 710 at time T3 and can determine to increase the vacuum negative pressure 710 if, for example, the patient's pain level does not increase at T4. Although described by way of example with respect to infusion, it should be understood that the various aspects of the techniques described herein with respect to FIG. 11 are also applicable to cases of wound perfusion where the wound is simultaneously irrigated with a flowing fluid and a negative pressure is applied.

[0139] The techniques described in this disclosure may be implemented at least partially in hardware, software, firmware, or any combination thereof. For example, various aspects of the described techniques may be implemented within one or more processors including one or more microprocessors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), processing circuits (e.g., fixed function circuits, programmable circuits, or any combination of functional and programmable circuits), or any other equivalent integrated circuit or discrete logic circuit, and any combination of such components. The term "processor" or "processing circuit" may generally refer to any of the foregoing logic circuits alone or in combination with other logic circuits, or other equivalent circuits. A control unit including hardware may also execute one or more techniques of this disclosure.

[0140] Such hardware, software, and firmware may be implemented within the same device or in separate devices to support the various techniques described in this disclosure. Further, any of the described units, modules, or components may be implemented together or separately as discrete but interoperable logic devices. The depiction of different features as modules or units is intended to highlight different functional aspects and does not necessarily imply that such modules or units must be implemented by separate hardware, firmware, or software components. Rather, the functions associated with one or more modules or units may be performed by separate hardware, firmware, or software components, or may be integrated within common or separate hardware, firmware, or software components.

[0141] The techniques described in this disclosure may also be embodied or encoded in a product that includes a computer-readable storage medium having instructions encoded thereon. When the instructions included in or encoded on the computer-readable storage medium are executed by one or more processors, one or more programmable processors or other processors may be caused to perform one or more of the techniques described herein, either on a product that includes the computer-readable storage medium or by the instructions that are encoded. Examples of computer-readable storage mediums include random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electronically erasable programmable read-only memory (EEPROM), flash memory, hard disks, compact disk ROM (CD-ROM), floppy disks, cassettes, magnetic media, optical media, or other computer-readable media. In some embodiments, the manufactured article may include one or more computer-readable storage mediums.

[0142] In some embodiments, the computer-readable storage medium may include a non-transitory medium. The term "non-transitory" may mean that the storage medium is not embodied in a carrier wave or a propagated signal. In certain embodiments, the non-transitory storage medium may store data that may change over time (e.g., in RAM or a cache).

[0143] The following examples may illustrate one or more aspects of this disclosure.

[0144] Example 1: A method includes receiving patient information and selecting at least one NPWT parameter setting for controlling fluid at a wound site via an NPWT dressing based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, and controlling fluid at the wound site via the NPWT dressing based on the selected at least one NPWT parameter setting.

[0145] Example 2: The method of Example 1 further includes receiving a measure of the effect of controlling fluid at the wound site and adjusting a causal model based on the received measure of the effect of controlling fluid at the wound site.

[0146] Example 3: The method of Example 2, wherein at least one NPWT parameter setting is at least one first NPWT parameter setting, and the method further includes selecting at least one second NPWT parameter setting for controlling fluid at the wound site based on the adjusted causal model and controlling fluid at the wound site via an NPWT dressing based on the selected at least one second NPWT parameter setting.

[0147] Example 4: The method according to any one of Examples 1 to 3, wherein the patient information includes at least one of user input patient information, patient biometric information, and wound measurement information, and the wound measurement information includes a measure of the effect of controlling fluid at the wound site.

[0148] Example 5: The method according to any one of Examples 1 to 4, wherein the patient information includes an aggregation of patient information of a plurality of patients.

[0149] Example 6: The method according to any one of Examples 1 to 5, wherein the wound measurement information and the measure of the effect of controlling fluid at the wound site include at least one of an impedance measurement value of the wound tissue, an oxygen measurement value of the wound bed, an oxygen measurement value of the fluid, a carbon dioxide measurement value of the wound bed, a temperature measurement value, an analyte sensor measurement value, and an optical measurement value of the wound bed.

[0150] Example 7: The method according to any one of Examples 1 to 6, wherein at least one first NPWT parameter setting and at least one second NPWT parameter setting each include a setting of one NPWT parameter of a set of NPWT parameters, and the set of NPWT parameters includes at least one of a negative pressure level, a negative pressure cycle, continuous application of negative pressure, fluid flow rate, fluid volume, fluid pressure, fluid temperature, fluid composition, fluid residence time, or fluid purge time.

[0151] Example 8: The method of Example 1 or 2, wherein at least one NPWT parameter setting includes values within a predetermined range.

[0152] Example 9: The method of any one of Examples 1-8, wherein controlling fluid at the wound site includes providing fluid to the wound site via an NPWT dressing.

[0153] Example 10: A system including a memory and one or more processors in communication with the memory, the one or more processors configured to receive patient information, determine a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at a wound site, select at least one NPWT parameter setting based on a causal model, and control fluid at the wound site via an NPWT dressing based on the selected at least one NPWT parameter setting.

[0154] Example 11: The system of Example 10, wherein the one or more processors are further configured to receive a measure of the effect of controlling fluid at the wound site and adjust the causal model based on the received measure of the effect of controlling fluid at the wound site.

[0155] Example 12: The system of Example 11, wherein at least one NPWT parameter setting is at least one first NPWT parameter setting, and the one or more processors are further configured to select at least one second NPWT parameter setting for controlling fluid at the wound site based on the adjusted causal model and control fluid at the wound site via an NPWT dressing based on the selected at least one second NPWT parameter setting.

[0156] Example 13: The system of any one of Examples 10-12, wherein the patient information includes at least one of user-input patient information, patient biometric information, and wound measurement information, and the wound measurement information includes a measure of the effect of controlling fluid at the wound site.

[0157] Example 14: A system according to any one of Examples 10 to 13, wherein the patient information includes an aggregation of patient information of a plurality of patients.

[0158] Example 15: A system according to any one of Examples 10 to 14, wherein the measure of the effect by controlling fluid at the wound measurement information and the wound site includes at least one of a measured impedance value of the wound tissue, an oxygen measurement value of the wound bed, an oxygen measurement value of the fluid, a carbon dioxide measurement value of the wound bed, a temperature measurement value, an analyte sensor measurement value, and an optical measurement value of the wound bed.

[0159] Example 16: A system according to any one of Examples 10 to 15, wherein each of at least one first NPWT parameter setting and at least one second NPWT parameter setting includes a setting of one NPWT parameter of a set of NPWT parameters, and the set of NPWT parameters includes at least one of a negative pressure level, a negative pressure cycle, continuous application of negative pressure, fluid flow rate, fluid volume, fluid pressure, fluid temperature, fluid composition, fluid residence time, or fluid purge time.

[0160] Example 17: A system according to Example 10 or 11, wherein at least one NPWT parameter setting includes a value within a predetermined range.

[0161] Example 18: A system according to any one of Examples 1 to 8, wherein controlling fluid at the wound site includes providing fluid to the wound site via an NPWT dressing.

[0162] Example 19: A computer-readable medium includes instructions that, when executed, cause one or more processors to receive patient information and select at least one NPWT parameter setting based on a causal model that determines the current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, and control fluid at the wound site via an NPWT dressing based on the selected at least one first NPWT parameter setting.

[0163] Example 20: At least one NPWT parameter setting is at least one first NPWT parameter setting, and the computer-readable medium further includes instructions that, when executed, cause one or more processors to receive a measure of the effect of controlling fluid at the wound site, adjust a causal model based on the received measure of the effect of controlling fluid at the wound site, select at least one second NPWT parameter setting for controlling fluid at the wound site based on the adjusted causal model, and control fluid at the wound site via an NPWT dressing based on the selected at least one second NPWT parameter setting, the computer-readable medium of Example 19.

[0164] Example 21: A system includes means for receiving patient information, means for selecting at least one NPWT parameter setting for controlling fluid at a wound site based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects of controlling fluid at the wound site, and means for controlling fluid at the wound site via an NPWT dressing based on the selected at least one NPWT parameter setting.

[0165] Example 22: The system of Example 21 further includes means for implementing any of the methods of Examples 2-9. Various examples have been described. These and other embodiments are within the scope of the following claims.

Claims

Claim 1 Receiving patient information, Based on a causal model that determines the current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling fluid at the wound site, selecting at least one NPWT parameter setting for controlling the fluid at the wound site via an NPWT dressing, Based on the selected at least one NPWT parameter setting, controlling the fluid at the wound site via the NPWT dressing, A method comprising. Claim 2 Receiving a measure of the effect of controlling the fluid at the wound site, Based on the received measure of the effect of controlling the fluid at the wound site, adjusting the causal model, The method according to claim 1, further comprising. Claim 3 The at least one NPWT parameter setting is at least one first NPWT parameter setting, and the method comprises Based on the adjusted causal model, selecting at least one second NPWT parameter setting for controlling the fluid at the wound site, Based on the selected at least one second NPWT parameter setting, controlling the fluid at the wound site via the NPWT dressing, The method according to claim 2, further comprising. Claim 4 The patient information includes at least one of user input patient information, patient biometric information, and wound measurement information, and the wound measurement information includes the measure of the effect of controlling the fluid at the wound site. The method according to claim 1. Claim 5 The patient information includes an aggregation of patient information of a plurality of patients. The method according to claim 1. Claim 6 The wound measurement information and the measure of the effect of controlling the fluid at the wound site include at least one of a measured impedance value of wound tissue, an oxygen measurement value of the wound bed, an oxygen measurement value of the fluid, a carbon dioxide measurement value of the wound bed, a temperature measurement value, an analyte sensor measurement value, and an optical measurement value of the wound bed. The method according to claim 1. Claim 7 Each of the at least one first NPWT parameter setting and the at least one second NPWT parameter setting includes a setting of one NPWT parameter out of a set of NPWT parameters, and the set of NPWT parameters includes at least one of negative pressure level, negative pressure cycle, continuous application of negative pressure, fluid flow rate, fluid volume, fluid pressure, fluid temperature, fluid composition, fluid residence time, or fluid purge time. The method according to claim 1.

8. The method according to claim 1, wherein the at least one NPWT parameter setting includes a value within a predetermined range.

9. The method according to claim 1, wherein controlling the fluid at the wound site includes providing the fluid to the wound site via the NPWT treatment dressing.

10. A memory, One or more processors in communication with the memory, Receiving patient information, Selecting at least one NPWT parameter setting based on a causal model that determines a current causal relationship between a set of negative pressure wound therapy (NPWT) parameter settings and a set of effects by controlling a fluid at a wound site, Controlling the fluid at the wound site via an NPWT dressing based on the selected at least one NPWT parameter setting. One or more processors configured as such, A system comprising.

11. The one or more processors, Receiving a measure of the effect by controlling the fluid at the wound site, Adjusting the causal model based on the received measure of the effect by controlling the fluid at the wound site. The system according to claim 10, further configured as such.

12. The at least one NPWT parameter setting is at least one first NPWT parameter setting, and the one or more processors, Selecting at least one second NPWT parameter setting for controlling the fluid at the wound site based on the adjusted causal model, Controlling the fluid at the wound site via the NPWT dressing based on the selected at least one second NPWT parameter setting. The system according to claim 11, further configured as such.

13. The system according to claim 10, wherein the patient information includes at least one of user input patient information, patient biometric information, and wound measurement information, and the wound measurement information includes the measure of the effect by controlling the fluid at the wound site.

14. The system according to claim 10, wherein the patient information includes an aggregation of patient information of a plurality of patients.

15. The system according to claim 10, wherein the wound measurement information and the measure of the effect by controlling the fluid at the wound site include at least one of a measured impedance value of wound tissue, an oxygen measurement value of the wound bed, an oxygen measurement value of the fluid, a carbon dioxide measurement value of the wound bed, a temperature measurement value, an analyte sensor measurement value, and an optical measurement value of the wound bed.

16. Each of the at least one first NPWT parameter setting and the at least one second NPWT parameter setting includes a setting of one NPWT parameter of a set of NPWT parameters, and the set of NPWT parameters includes at least one of a negative pressure level, a negative pressure cycle, a continuous application of negative pressure, a fluid flow rate, a fluid volume, a fluid pressure, a fluid temperature, a fluid composition, a fluid residence time, or a fluid purge time. The system according to claim 10.

17. The system according to claim 10, wherein the at least one NPWT parameter setting includes a value within a predetermined range.

18. The system according to claim 10, wherein controlling the fluid at the wound site includes providing the fluid to the wound site via the NPWT treatment dressing.

Images