Negative and / or positive pressure therapy devices and methods with natural language interfaces

Abstract

A negative and / or positive pressure therapy system can include a therapy device having a pressure source configured to provide a pressure therapy to a wound site covered by a wound dressing and a controller configured to control operation of the pressure source. The system can include a storage medium storing executable instructions that, when executed by one or more processors, cause the one or more processors to: receive a natural language query associated with providing the pressure therapy, the natural language query issued by a user of the therapy device; process the natural language query using one or more machine learning models trained using only data provided by a manufacturer of the therapy device and determine a response; and provide the response to one or more of the user or the controller.

Description

[0001] Cross-references to related applications

[0002] This application claims priority to UK Patent Application No. GB ​​2403957.0, filed on 20 March 2024, UK Patent Application No. GB ​​2409875.8, filed on 8 July 2024, and UK Patent Application No. GB ​​2414397.6, filed on 1 October 2024, each of which is incorporated herein by reference in its entirety. Technical Field

[0003] The implementation schemes described herein relate to devices, systems, and methods for treating wounds, for example, using dressings in combination with negative and / or positive pressure therapy. Background Technology

[0004] Modern clinical wound care for mammals is performed by individuals with experience and training ranging from highly experienced clinicians specializing in the subject to caregivers and patients without formal training who can be categorized as lay users. Wound care is performed in a wide range of clinical settings, from specialist medical units to personal homes and battlefield centers. Over the past two decades, wound care medical devices have become more complex than simple dressings, often including electronic controllers for operating the device. Increased focus has also been placed on wound and tissue site preparation and wound prevention. The range of modern wound care medical devices includes active treatment devices (e.g., local negative pressure, electrical stimulation, ultrasound therapy), wound measurement, assessment, and monitoring devices (e.g., infection monitoring devices using optical, temperature, and pH technologies including UV and infrared), wound bed preparation devices (e.g., waterjet scalpels, cold plasma rods), and wound prevention devices (patient pressure monitors, patient position sensors). All of these devices present challenges for human factors experts in designing user-friendly yet safe user interfaces that are adaptable to multiple languages ​​and levels of expertise. Furthermore, due to the challenges of resource allocation in healthcare, caregivers often find it difficult to contact specialists or peers for assistance.

[0005] Many different types of wound dressings are known for aiding the healing process in humans and animals. These different types of wound dressings include many different types of materials and layers, such as gauze, padding, foam pads, or multilayer wound dressings. Local negative pressure (TNP) therapy, sometimes also called vacuum-assisted closure, negative pressure wound therapy, or decompression wound therapy, is widely recognized as a beneficial mechanism for improving wound healing rates. This type of treatment is suitable for a wide range of wounds, such as incisions, open wounds, and abdominal wounds. TNP therapy helps close and heal wounds by reducing tissue edema, promoting blood flow, stimulating granulation tissue formation, removing excess exudate, and reducing bacterial load. Therefore, it reduces wound infection. Moreover, TNP therapy allows for less external disturbance to the wound and promotes faster healing. Summary of the Invention

[0006] A negative pressure and / or positive pressure therapy system may include a treatment device having a pressure source configured to provide pressure therapy to a wound site covered by a wound dressing, a controller configured to control the operation of the pressure source, and communication circuitry configured to facilitate wireless communication with the treatment device. The system may include a storage medium storing executable instructions that, when executed by one or more processors, cause one or more processors to receive information associated with providing pressure therapy from the communication circuitry. The instructions may cause one or more processors to receive natural language queries associated with providing pressure therapy. The natural language queries may be issued by a user of the treatment device. The instructions may cause one or more processors to process the natural language queries and determine a response using one or more machine learning models trained solely on data provided by the manufacturer of the treatment device. The instructions may cause one or more processors to provide a response to one or more of the user or the controller. The treatment device may provide negative pressure wound therapy.

[0007] The negative pressure and / or positive pressure therapy systems described in any of the preceding paragraphs and / or any systems, devices, or apparatuses disclosed herein may include one or more of the following features: A response may include a set of instructions for adjusting the operation of the pressure source. Providing a response to the controller may cause the controller to adjust the operation of the pressure source. The controller may deactivate or activate the pressure source. Natural language queries may relate to one or more of the status of the therapy device or the status of a wound dressing. The controller may be configured to provide alarms associated with the provision of pressure therapy. Natural language queries may be associated with alarms. A response may include instructions for remedying the alarm.

[0008] The negative pressure and / or positive pressure therapy systems described in any of the preceding paragraphs and / or any systems, devices, or apparatuses disclosed herein may include one or more of the following features: The pressure source may include a negative pressure source configured to provide negative pressure therapy to a wound site. A natural language query may be associated with selecting a wound dressing from a plurality of wound dressings suitable for the treatment device. Information may include one or more images of the wound site. A response may include one or more of the type, shape, and size of a wound dressing suitable for providing pressure therapy to the wound site. One or more images may include the outline of an incision to be made at the wound site. The response may be determined based on the outline of the incision made at the wound site. Determining the response may include generating a shape of a wound dressing that matches the shape of the wound site. Providing a response may include displaying the shape of a wound dressing superimposed on the wound site. Providing a response may include displaying a wound dressing superimposed on the wound site.

[0009] The negative pressure and / or positive pressure therapy systems described in any of the preceding paragraphs and / or any systems, devices, or apparatuses disclosed herein may include one or more of the following features: Providing a response may include displaying a response. One or more processors may reside in one or more servers. Natural language queries may include one or more of speech or text. Information may include one or more of image data or data collected by one or more sensors.

[0010] A method for controlling a negative pressure and / or positive pressure therapy system may include providing pressure therapy from a pressure source to a wound site covered by a wound dressing by a treatment device. The pressure source may be controlled by a controller. The method may include receiving a natural language query associated with the provision of pressure therapy by one or more processors located remotely to the treatment device. The natural language query may be issued by a user of the treatment device. The method may include using one or more machine learning models trained solely using data provided by the manufacturer of the treatment device by one or more processors located remotely to the negative pressure wound therapy device. The method may include processing the natural language query and determining a response by one or more processors located remotely to the negative pressure wound therapy device. The method may include providing a response by one or more processors located remotely to the user or a controller. The treatment device may provide negative pressure wound therapy.

[0011] The methods in any of the preceding paragraphs and / or any of the methods disclosed herein may include one or more of the following features: A response may include a set of instructions for adjusting the operation of the pressure source. Providing a response to the controller may cause the operation of adjusting the pressure source. The operation of adjusting the pressure source may include deactivating or activating the pressure source. Natural language queries may relate to one or more of the status of the treatment device or the status of the wound dressing. The method may include an alarm provided by the treatment device in association with the provision of pressure therapy. Natural language queries may be associated with the alarm. A response may include instructions for remedying the alarm.

[0012] The methods in any of the preceding paragraphs and / or any of the methods disclosed herein may include one or more of the following features: The method may include processing a natural language query and determining a response based on one or more images of the wound. The natural language query may be associated with selecting a wound dressing from a plurality of wound dressings suitable for a treatment device. The response may include one or more of the type, shape, and size of a wound dressing suitable for providing negative pressure therapy to the wound site. One or more images may include the outline of an incision to be made at the wound site. The response may be determined based on the outline of the incision made at the wound site. Providing a response may include displaying a wound dressing superimposed on the wound site. Determining a response may include generating a shape of a wound dressing that matches the shape of the wound site. Providing a response may include displaying the shape of a wound dressing superimposed on the wound site.

[0013] A negative pressure and / or positive pressure therapy system may include a treatment device having a pressure source configured to provide negative pressure therapy to a wound site covered by a wound dressing, a controller configured to control the operation of the pressure source, and communication circuitry configured to facilitate wireless communication with the treatment device. The system may include a storage medium storing executable instructions that, when executed by one or more processors, cause the processors to initiate an interactive session in a baseline state using one or more machine learning models. This interactive session in the baseline state utilizes one or more machine learning models trained solely on data provided by the manufacturer of the treatment device to provide a response to a natural language query issued by a user of the treatment device. The instructions may cause the one or more processors to receive information associated with providing pressure therapy from the communication circuitry. The instructions may cause the one or more processors to update the state of the interactive session, the updated state including the baseline state and also including information associated with providing pressure therapy. The instructions may cause the one or more processors to receive a natural language query associated with providing pressure therapy, issued by a user of the treatment device. The instructions may cause the one or more processors to process the natural language query and determine a response using one or more machine learning models trained solely on data provided by the manufacturer of the treatment device and using the information associated with providing pressure therapy. Instructions can cause one or more processors to respond to one or more users or controllers. Treatment devices can provide negative pressure wound therapy.

[0014] The negative pressure and / or positive pressure therapy systems described in any of the preceding paragraphs and / or any systems, devices, or apparatuses disclosed herein may include one or more of the following features. Information associated with providing pressure therapy may include at least one of the following: the user's level of knowledge, data collected by one or more sensors, or changes in pressure therapy parameters over a period of time. Determining a response may include assigning higher priority to information associated with a baseline state than to information associated with providing pressure therapy.

[0015] A medical system may include a medical device comprising a processing module configured to provide processing to a patient, a controller configured to control the operation of the processing module, and communication circuitry configured to facilitate wireless communication with the medical device. The system may include a storage medium storing executable instructions that, when executed by one or more processors, cause one or more processors to receive information associated with providing processing from the communication circuitry. The instructions may cause one or more processors to receive a natural language query associated with providing processing. The natural language query may be issued by a user of the medical device. The instructions may cause one or more processors to process the natural language query and determine a response using one or more machine learning models trained solely on data provided by the manufacturer of the medical device. The instructions may cause one or more processors to provide a response to one or more of the user or the controller.

[0016] This document discloses kits that include any of the treatment devices described in the preceding paragraphs and / or any of the devices, equipment, or systems disclosed herein, as well as one or more wound dressings.

[0017] Any feature, component, or detail of any arrangement or implementation scheme disclosed in this application (including, but not limited to, any device implementation scheme in the device implementation scheme and any negative pressure wound therapy implementation scheme in the negative pressure wound therapy implementation scheme disclosed herein) may be interchangeably combined with any other feature, component, or detail of any arrangement or implementation scheme disclosed herein to form new arrangements and implementation schemes. Attached Figure Description

[0018] Figure 1A and Figure 1B An example of a negative pressure wound therapy system is shown.

[0019] Figure 2A This is an isometric view of the negative pressure wound therapy device and the canister, showing the canister removed from the device's pump assembly.

[0020] Figure 2B yes Figure 2A The rear view of the negative pressure wound therapy device shown.

[0021] Figure 2C Examples Figure 2A The top surface of the negative pressure wound therapy device shown depicts the user interface.

[0022] Figure 3 A schematic diagram of the control system of a negative pressure wound therapy device is shown.

[0023] Figure 4 Another negative pressure wound therapy system is illustrated.

[0024] Figure 5A , Figure 5B and Figure 5C An example of a canisterless negative pressure wound therapy device is shown.

[0025] Figures 6A to 6E and Figures 7A to 7B An example of a negative pressure wound therapy system with a natural language interface is shown.

[0026] Figures 8A to 8D Examples of suitable wound dressings for applying negative pressure wound therapy are provided.

[0027] Figures 9A to 9C and Figure 10 The selection and placement of appropriate wound dressings for applying negative pressure wound therapy are illustrated.

[0028] Figures 11 to 14 An example of a support is shown for selecting and placing a suitable wound dressing for applying negative pressure wound therapy.

[0029] Figure 15 The options for handheld devices that form part of regulatory information and instructions are illustrated.

[0030] Figure 16 Enhanced images of wounds are shown, identifying the location and percentage of certain bacteria and living tissue. Detailed Implementation

[0031] The embodiments disclosed herein relate to systems and methods for treating and / or monitoring wounds. Some embodiments of the negative pressure wound therapy devices disclosed herein may include a negative pressure source configured to be connected and / or fluidly coupled to a wound covered by a wound dressing via a fluid flow path and to provide negative pressure to the wound.

[0032] The term "wound" is used throughout this instruction manual. The term "wound" should be interpreted broadly and encompasses any open or closed wound, or other wound that results from a tear, cut, or puncture of the skin, or from trauma causing bruising or any other superficial or other condition or defect on the patient's skin, or which benefits from pressure treatment. Therefore, a wound is broadly defined as any area of ​​damaged tissue that may or may not produce fluid. Examples of such wounds include, but are not limited to, abdominal wounds or other large or open wounds resulting from surgery, trauma, sternotomy, fasciotomy, or other conditions, lacerations, acute wounds, chronic wounds, subacute and lacerated wounds, traumatic wounds, flaps and skin grafts, lacerations, abrasions, contusions, burns, diabetic ulcers, pressure ulcers, stomas, surgical wounds, traumatic ulcers, and venous ulcers.

[0033] The implementation schemes of the systems and methods disclosed herein can be used in conjunction with local negative pressure (“TNP”) or decompression therapy systems. In short, negative pressure wound therapy helps close and heal many forms of “difficult-to-heal” wounds by reducing tissue edema, promoting blood flow and granulation tissue formation, or removing excess exudate, and can reduce bacterial load (and thus reduce the risk of infection). Additionally, this treatment allows for less disturbance to the wound, resulting in faster healing. TNP therapy systems can also aid in the healing of surgically closed wounds by removing fluid. TNP therapy can help stabilize tissue at juxtaposed closure sites. Another beneficial use of TNP therapy can be found in grafts and flaps, where removing excess fluid is important and close proximity of the graft to the tissue is necessary to ensure tissue viability.

[0034] As used herein, a decompression level or negative pressure level (e.g., -X mmHg) represents a pressure level relative to normal ambient atmospheric pressure, which may correspond to 760 mmHg (or 1 atm, 29.93 inHg, 101.325 kPa, 14.696 psi, etc.). Therefore, a negative pressure value of -X mmHg reflects a pressure X mmHg lower than 760 mmHg, or in other words, a pressure of (760-X) mmHg. Furthermore, a negative pressure “less” or “smaller” than X mmHg corresponds to a pressure closer to atmospheric pressure (e.g., –40 mmHg is less than –60 mmHg). A negative pressure “more” or “larger” than –X mmHg corresponds to a pressure further away from atmospheric pressure (e.g., –80 mmHg is more than –60 mmHg). In some cases, a local ambient atmospheric pressure is used as a reference point, which may not necessarily be, for example, 760 mmHg.

[0035] In addition to or as a substitute for decompression therapy, the systems and methods disclosed herein can be used in conjunction with other types of treatments, such as perfusion, ultrasound, heating or cooling, nerve stimulation, etc. In some cases, the disclosed systems and methods can be used for wound monitoring without the application of additional treatment. The systems and methods disclosed herein can be used in conjunction with dressings, including compression dressings, decompression dressings, etc.

[0036] Healthcare providers, such as clinicians and nurses, can prescribe TNPs that specify, for example, stress levels or application times. However, the healing process is different for each patient, and the prescription can affect the healing process in ways that the clinician or healthcare provider did not intend when designing the prescription. Healthcare providers may attempt to adjust the prescription as the wound heals (or does not heal), but this process may require various appointments that can be time-consuming and repetitive. The embodiments disclosed herein provide systems, devices, or methods for efficiently adjusting TNP prescriptions and delivering effective TNP treatment.

[0037] Wound Treatment System

[0038] Figure 1A A negative pressure wound treatment system 100' (sometimes referred to as a decompression or negative pressure wound therapy system, TNP system, or wound treatment system) is schematically illustrated. In any embodiment disclosed herein, although not essential, the negative pressure wound treatment system 100' may include a wound filler 102 placed on or inside a wound 104 (which may be a cavity). The wound 104 may be sealed by a wound covering 106 (which may be a drape) such that the wound covering 106 is in fluid communication with the wound 104. The combination of the wound filler 102 and the wound covering 106 may be referred to as a wound dressing. A tube or conduit 108' (also referred to herein as a flexible suction adapter or fluid connector) may be used to connect the wound covering 106 to a wound treatment device 110' (sometimes referred to integrally or partially as a "pump assembly") configured to supply reduced pressure or negative pressure. The conduit 108' may be a single-lumen tube or a multi-lumen tube. A connector may be used to removably and selectively connect a tube or conduit of device 110' to conduit 108'.

[0039] In any system disclosed herein, the wound treatment device may be canisterless, wherein, for example, but not limited to, wound exudate is collected in a wound dressing or transferred via a catheter for collection at another location. However, any wound treatment device disclosed herein may include or support a canister.

[0040] Additionally, for any wound treatment system disclosed herein, any wound treatment device may be attached to, supported by, or adjacent to a wound dressing. Wound filler 102 may be of any suitable type, such as hydrophilic or hydrophobic foam, gauze, inflatable bags, etc. Wound filler 102 may conform to wound 104 such that wound filler 102 substantially fills the cavity of wound 104. Wound covering 106 may provide a substantially fluid-impermeable seal over wound 104. Wound covering 106 may have a top side and a bottom side. The bottom side may be adhesively (or sealed in any other suitable manner) to wound 104, for example, by sealing with the skin surrounding wound 104. Conduit 108 or any other conduit disclosed herein may be formed of polyurethane, PVC, nylon, polyethylene, silicone, or any other suitable material.

[0041] Wound covering 106 may have a port (not shown) configured to receive the end of conduit 108. In some cases, conduit 108 may otherwise pass through or beneath wound covering 106 to supply reduced pressure to wound 104 in order to maintain a desired reduced pressure level in wound 104. Conduit 108 may be any suitable article configured to provide at least substantially sealed fluid flow pathway or path between wound treatment device 110' and wound covering 106 to supply reduced pressure provided by wound treatment device 110' to wound 104.

[0042] Wound covering 106 and wound filler 102 may be provided as individual products or as an integrated single unit. In some cases, wound filler is not provided, and the wound covering itself may be considered as a wound dressing. The wound dressing may then be connected via conduit 108 to a negative pressure source of the wound treatment device 110'. In some cases, although not required, the wound treatment device 110' may be miniaturized and portable, but a larger conventional negative pressure source (or pump) may also be used.

[0043] Wound covering 106 may be positioned above the wound site to be treated. Wound covering 106 may form a substantially sealed cavity or enclosure above the wound. Wound covering 106 may have a membrane with high water vapor permeability to allow excess fluid to evaporate, and may have a superabsorbent material contained therein to safely absorb wound exudate. In some cases, components of the TNP system described herein may be particularly suitable for open wounds with small amounts of wound exudate.

[0044] Wound treatment device 110' can operate with or without an exudate tank. In some cases, as shown, wound treatment device 110' may include an exudate tank. In some cases, wound treatment device 110' and catheter 108' are configured such that catheter 108' can be quickly and easily removed from wound treatment device 110' to facilitate or improve wound dressing or pump changes (if necessary). Any pump assembly disclosed herein may have any suitable connection between catheter 108' and pump.

[0045] The wound treatment device 110' can deliver a negative pressure of approximately -80 mmHg, or between approximately -20 mmHg and -200 mmHg. It should be noted that these pressures are relative to normal ambient atmospheric pressure; therefore, -200 mmHg would actually be approximately 560 mmHg. In some cases, the pressure range can be between approximately -40 mmHg and -150 mmHg. Alternatively, pressure ranges up to -75 mmHg, up to -80 mmHg, or above -80 mmHg can be used. Additionally, in some cases, pressure ranges below -75 mmHg can be used. Alternatively, the wound treatment device 110' can supply a pressure range exceeding approximately -100 mmHg or even -150 mmHg.

[0046] As will be described in more detail below, the negative pressure wound therapy system 100' can be configured to provide a connection 332 to a standalone or remote computing device 334. The connection 332 can be wired or wireless (e.g., Bluetooth, Bluetooth Low Energy (BLE), Near Field Communication (NFC), WiFi, or cellular). The remote computing device 334 can be a smartphone, tablet, laptop, or another standalone computer, server (such as a cloud server), another pump device, etc.

[0047] The wound treatment device 110' can transmit data to a cloud 690 that can host one or more cloud-based servers 692. The cloud 690 can host and process the data and provide one or more responses. The cloud 690 can communicate with a remote computing device 334 that can receive data from the wound treatment device 110'. The cloud 690 can host an artificial intelligence (AI) system that can be configured for natural language processing to assist the operation of the wound treatment device 110' or provide treatment, as described herein. The AI ​​system may include one or more NLP models as described herein.

[0048] The wound treatment device 110' can wirelessly transmit data to the remote computing device 334, and the remote computing device 334 can upload the data received from the wound treatment device 110' to the cloud 690 for processing. The cloud 690 can transmit the response back to the remote computing device 334.

[0049] Figure 1B Another negative pressure wound therapy system 100 is illustrated. The negative pressure wound therapy system 100 may have any other negative pressure wound therapy systems disclosed herein (including, but not limited to, those shown below). Figure 1A The negative pressure wound treatment system 100' shown is or Figure 4 Any component, feature, or other detail of the negative pressure wound therapy system 400 shown herein may be combined with or replaced. Figure 1BThis refers to any component, feature, or other detail of the negative pressure wound treatment system 100 shown and / or described herein. The negative pressure wound treatment system 100 may have a wound covering 106 over the wound 104, which can seal the wound 104. A conduit 108, such as a single-lumen or multi-lumen tube, may be used to connect the wound covering 106 to a wound treatment device 110 (sometimes referred to integrally or partially as a “pump assembly”), which is configured to supply reduced pressure or negative pressure. The wound covering 106 may be in fluid communication with the wound 104.

[0050] refer to Figure 1B The catheter 108 may have a bridging portion 130 and an applicator 132 at the distal end of the bridging portion 130, thereby forming a flexible suction adapter (or catheter) 108. The bridging portion may have a proximal portion and a distal portion (the distal portion being closer to the wound 104 than the proximal portion). A connector 134 may be disposed at the proximal end of the bridging portion 130 for connection to at least one channel in a passage that can extend along... Figure 1B The bridging portion 130 of the catheter 108 shown extends in length. A cap 140 may be coupled to a portion of the catheter 108 and, in some cases, may be attached to a connector 134, as shown. The cap 140 serves to prevent fluid leakage from the proximal end of the bridging portion 130. The catheter 108 may be a flexible port manufactured by Smith+Nephew. As mentioned, the negative pressure wound therapy system 100 may include a negative pressure source, such as a wound treatment device 110, capable of supplying negative pressure to the wound 104 through the catheter 108. Although not required, the wound treatment device 110 may also include a tank or other container for storing wound exudate and other fluids that can be removed from the wound.

[0051] The wound treatment device 110 can be connected to the connector 134 via a conduit or tube 142. In use, the applicator 132 can be placed over an opening formed in a wound covering 106, which is placed over a properly prepared wound or wound 104. Subsequently, with the wound treatment device 110 connected to the connector 134 via the tube 142, the wound treatment device 110 can be activated to supply negative pressure to the wound. The negative pressure can be applied until the desired healing level of the wound is achieved.

[0052] The bridging portion 130 may include an upper channel material or layer positioned between the upper and intermediate layers, wherein a lower channel material or layer is positioned between the intermediate and bottom layers. The upper, intermediate, and bottom layers may have elongated portions extending between proximal and distal ends and may include fluid-impermeable materials, such as polymers like polyurethane. It should be appreciated that the upper, intermediate, and bottom layers may each freely comprise different material compositions including semi-permeable materials. In some cases, one or more of the upper, intermediate, and bottom layers may be at least partially transparent. In some cases, the upper and bottom layers may be curved, rounded, or convex outwards along their majority length.

[0053] The upper and lower channel layers may be elongated layers extending from the proximal end to the distal end of the bridging portion 130, and may each preferably comprise a porous material, including, for example, open-cell foam, such as polyethylene or polyurethane. In some cases, one or more of the upper and lower channel layers may be composed of fabric (e.g., knitted or woven spacer fabrics such as knitted polyester 3D fabric, Baltex 7970.RTM, or Gehring 879.RTM)) or nonwoven materials or terry-knitted or terry-pile materials. The fibers may not necessarily be woven and may include felted and flocked (including materials such as Flotex.RTM) fibrous materials. The selected materials are preferably adapted to guide wound exudate away from the wound through the channels and to deliver negative pressure or exhaust air to the wound site, and may also impart a degree of kink resistance or clogging resistance to the channel layers. In one example, the upper channel layer may comprise an open-cell foam, such as polyurethane, and the lower channel layer may comprise a fabric. In another example, the upper channel layer is optional, and the system may be modified to have an open upper channel. The upper channel layer may have a curved, rounded, or upwardly convex upper surface and a substantially flat lower surface, and the lower channel layer may have a curved, rounded, or downwardly convex lower surface and a substantially flat upper surface.

[0054] The fabric or material of any component of the bridging portion 130 may have a three-dimensional (3D) structure, wherein one or more types of fibers form a structure in which fibers extend in all three dimensions. In some cases, such fabrics may aid in wicking, fluid delivery, or the transmission of negative pressure. In some cases, the material of the fabric or channel may comprise several layers of material stacked or laminated on top of each other, which may be used in some cases to prevent the channel from collapsing under the application of negative pressure. The material used in some embodiments of the catheter 108 may be conformal and flexible, which in some cases may help avoid pressure ulcers and other complications that can be caused by the wound management system pressing against the patient's skin.

[0055] The distal ends of the upper, middle, and lower layers, as well as the channel layer, may be enlarged (to be positioned above the wound site) and may form a "teardrop" or other enlarged shape. At least the distal ends of the upper, middle, and lower layers, as well as the channel layer, may also be provided with at least one through-hole. This hole can be used not only to drain wound exudate and apply negative pressure to the wound, but also during the manufacture of the device, as these holes can be used to properly align these respective layers.

[0056] In some embodiments, a controlled gas leak 146 (sometimes referred to as a gas leak, air leak, or controlled air leak) may be provided on the bridging portion 130, for example, at its proximal end. This air leak 146 may include an opening or channel extending through the upper layer of the bridging portion 130, such that the air leak 146 is in fluid communication with the upper channel of the bridging portion 130. When suction is applied to the conduit 108, gas (such as air) may enter through the gas leak 146 and move along the upper channel of the bridging portion 130 from the proximal end to the distal end of the bridging portion 130. The gas can then be drawn into the lower channel of the bridging portion 130 through an orifice penetrating the distal ends of the upper, middle, and lower layers.

[0057] Air vent 146 may include a filter. Preferably, air vent 146 is located near the proximal end of bridging portion 130 to minimize the possibility of wound exudate or other fluids coming into contact with and potentially clogging or interfering with air vent 146 or the filter. In some cases, the filter may be a microporous membrane capable of excluding microorganisms and bacteria and potentially filtering out particles larger than 45 μm. Preferably, the filter excludes particles larger than 1.0 μm, and more preferably, particles larger than 0.2 μm. Advantageously, some embodiments may provide a filter that is at least partially chemically resistant to, for example, water, common household liquids such as shampoo, and other surfactants. In some cases, applying a vacuum again to the suction adapter or wiping the exposed external portion of the filter may be sufficient to remove any foreign matter clogging the filter. The filter may be made of a suitably resistant polymer such as acrylic, polyethersulfone, or polytetrafluoroethylene, and may be oleophobic or hydrophobic. In some cases, gas vent 146 may supply a relatively constant gas flow that does not significantly increase with the application of additional negative pressure to conduit 108. In an example of the negative pressure wound therapy system 100, the gas flow through the gas leak 146 increases with the application of additional negative pressure. Preferably, this increased gas flow is minimized and does not increase proportionally to the negative pressure applied thereto. Further descriptions of such bridging elements, conduits, air leaks, and other components, features, and details that may be used with any embodiment of the negative pressure wound therapy system disclosed herein can be found in U.S. Patent No. 8,801,685, which is incorporated herein by reference in its entirety as if fully set forth herein.

[0058] Any wound treatment device disclosed herein (e.g., device 110 or 110') can provide continuous or intermittent negative pressure therapy. Continuous therapy can be delivered at pressures above 0 mmHg, -25 mmHg, -40 mmHg, -50 mmHg, -60 mmHg, -70 mmHg, -80 mmHg, -90 mmHg, -100 mmHg, -120 mmHg, -125 mmHg, -140 mmHg, -160 mmHg, -180 mmHg, -200 mmHg, or below -200 mmHg. Intermittent therapy can be delivered between low negative pressure setpoints and high negative pressure setpoints (sometimes referred to as setpoints). The low setpoint can be set to above 0 mmHg, -25 mmHg, -40 mmHg, -50 mmHg, -60 mmHg, -70 mmHg, -80 mmHg, -90 mmHg, -100 mmHg, -120 mmHg, -125 mmHg, -140 mmHg, -160 mmHg, -180 mmHg, or below -180 mmHg. The high setpoint can be set to above -25 mmHg, -40 mmHg, -50 mmHg, -60 mmHg, -70 mmHg, -80 mmHg, -90 mmHg, -100 mmHg, -120 mmHg, -125 mmHg, -140 mmHg, -160 mmHg, -180 mmHg, -200 mmHg, or below -200 mmHg. During intermittent treatment, negative pressure at a low set point can be delivered for a first duration, and after the first duration expires, negative pressure at a high set point can be delivered for a second duration. After the second duration expires, negative pressure at a low set point can be delivered. The first and second durations can be the same or different values.

[0059] In operation, wound packing 102 can be inserted into the cavity of wound 104, and wound covering 106 can be placed to seal wound 104. Wound treatment device 110' can provide negative pressure to wound covering 106, which can be transmitted to wound 104 via wound packing 102. Fluids (such as wound exudate) can be drawn through conduit 108' and stored in a container. In some cases, the fluid is absorbed by wound packing 102 or one or more absorbent layers (not shown).

[0060] Wound dressings that can be used with the pump assembly and system of this application include Renasy-F™, Renasy-G™, Renasy AB™, and Pico™ dressings available from Smith+Nephew. In some cases, foam or gauze can be used to pack the wound, particularly when packing the oral cavity is important. Further descriptions of such wound dressings and other components of negative pressure wound therapy systems that can be used with the pump assembly and system of this application can be found in U.S. Patent Publications 2012 / 0116334, 2011 / 0213287, 2011 / 0282309, 2012 / 0136325, and U.S. Patent No. 9,084,845, and International Application Publication WO2021 / 069642, each of which is incorporated herein by reference in its entirety as if fully set forth herein. In some cases, other suitable wound dressings may be used.

[0061] More generally, any of the wound dressings described herein can be used without applying negative pressure. For example, the wound dressing can be ALLEVYN™, OPSITE™, ACTICOAT™, or DURAFIBER™ dressings manufactured by Smith+Nephew. For instance, ALLEVYN™ dressings are moist wound environment dressings designed for treating wounds without the use of negative pressure.

[0062] In some cases, wound dressings can deliver an effective dose of nitric oxide (NO) to wounds. Nitric oxide affects vasodilation, stimulates angiogenesis, influences the host immune response, and exhibits potent broad-spectrum antimicrobial and antibiofilm activity. Due to these multiple effects, nitric oxide exhibits a potent effect on tissues, and increasing the amount of NO can support accelerated wound healing, particularly of chronic wounds. Under normal conditions, nitric oxide is a free radical with a short lifespan, converting into a more stable chemical substance within seconds of its generation. Therefore, for example, if gaseous nitric oxide comes into contact with air, it will be rapidly oxidized to generate nitrogen dioxide (NO2). Consequently, it may be difficult to maintain high concentrations of nitric oxide within wound dressings or other similar structures for extended periods. Therefore, devices or wound dressings having one or more layers containing a more stable composition can efficiently generate nitric oxide over time after activation for the stable and sustained delivery of nitric oxide to biological tissues.

[0063] Wound dressings for delivering nitric oxide may include: a covering layer configured to form a seal around the wound; an activator layer; a dried nitric oxide source layer that is free of or relatively free of liquid; and a collection and distribution layer. The wound dressing may also include a masking layer configured to at least partially restrict the visibility of the wound. The dried nitric oxide source layer may include nitrite. Nitrite may include sodium nitrite. The activator layer may be positioned above the nitric oxide source layer. In some embodiments, the nitric oxide source layer may be positioned above the activator layer. The collection and distribution layer may be positioned between the activator layer and the dried nitric oxide source layer. The activator layer may include a hydrogel or a dry gel. The wound dressing may include a second dried nitric oxide source layer. The wound dressing may be configured to generate nitric oxide when placed over the wound. In some cases, the wound dressing may be configured not to generate nitric oxide before being placed on the wound. The wound dressing may be used in conjunction with providing negative pressure wound therapy.

[0064] Figures 2A to 2C A negative pressure wound therapy device 110 is shown. As shown, a pump assembly 160 and a tank 162 can be connected to form the wound therapy device 110. (Reference) Figure 2C The pump assembly 160 may include an interface panel 170 having a display 172, one or more indicators 174, or one or more controls or buttons, including, for example, but not limited to, treatment start and pause buttons 180 or warning / alarm mute buttons 182. The interface panel 170 may have one or more input controls or buttons 184 (three are shown) that can be used to control any function of the pump assembly 160 or the interface panel 170. For example, but not limited to, one or more buttons of the buttons 184 can be used to turn the pump assembly 160 on or off, start or pause treatment, operate and monitor the operation of the pump assembly 160, scroll through menus displayed on the display 172, or control or perform other functions. In some cases, the command buttons 184 may be programmable and may be made of soft-touch rubber.

[0065] Additionally, the interface panel 170 may have a visual indicator 186 that indicates which of one or more buttons 184 is active. The interface panel 170 may also have a lock / unlock control or button 188 that can be configured to selectively lock or unlock the functions of various buttons (e.g., button 184) or displays 172. For example, treatment settings adjustments can be locked / unlocked via the lock / unlock control 188. When the lock / unlock button 188 is locked, pressing one or more of the various other buttons or displays will not cause the pump assembly 160 to change any display or performance functions of the device. In this way, the interface panel 170 is protected against accidental bumps or touches to the various buttons or displays. The interface panel 170 may be located on the upper portion of the pump assembly 160, such as, but not limited to, the upward-facing surface of the pump assembly 160.

[0066] Display 172 may be a screen such as an LCD screen, and may be mounted in the middle portion of interface panel 170. Display 172 may be a touch screen display. Display 172 may support playback of audiovisual (AV) content (such as instructional videos) and present multiple screens or a graphical user interface (GUI) for configuring, controlling, and monitoring the operation of pump assembly 160.

[0067] One or more indicators 174 may be lights (e.g., LEDs) and may be configured to provide visual indications of warning conditions and / or the status of the pump. For example, but not limited to, one or more indicators 174 may be configured to provide visual indications of the status of the pump assembly 160 or other components of the negative pressure wound treatment system 100, including but not limited to the conduit 108 or wound covering 106 (e.g., to provide indications of normal operation, low battery, leak, full tank, blockage, overpressure, etc.). Any one or more suitable indicators, such as visual, audio, tactile indicators, etc., may be used additionally or alternatively.

[0068] Figure 2B It shows Figure 2A The image shows a rear or back view of the wound treatment device 110. As shown, the pump assembly 160 may include a speaker 192 for generating sound. For example, but not limited to, the speaker 192 may generate an audible alert in response to deviations in treatment delivery, non-compliance in treatment delivery, or any other similar or suitable conditions or combinations thereof. The speaker 192 may provide audio to accompany one or more instructional videos that may be displayed on the display 172.

[0069] Pump assembly 160 may be configured to provide easy access (e.g., an access door on the housing of the pump assembly) to one or more filters (e.g., antibacterial filters) of the pump assembly 160. This allows a user (e.g., a healthcare provider or patient) to more easily access, inspect, or replace such filters. Pump assembly 160 may also include a power socket 196 for supplying power to pump assembly 160 or for charging and recharging an internal power source (e.g., a battery). Some embodiments of pump assembly 160 may include a disposable or renewable power source, such as one or more batteries, eliminating the need for a power socket. The capacity, health status, charging current, etc., of the power source can be monitored. Pump assembly 160 may have a recess 198 formed therein to facilitate gripping of pump assembly 160.

[0070] Canister 162 holds the fluid aspirated from wound 104. For example, canister 162 may have a capacity of 800 mL (or approximately 800 mL), or a capacity from 300 mL or less to 1000 mL or more, or any capacity level within that range. Canister 162 may include tubing for connection to conduit 108 to form a fluid flow path. For example, when canister 162 is full of fluid, it can be replaced with another canister. Reference Figure 2A The wound treatment device 110 may include a can inlet tube 142 (also referred to herein as a dressing port connector) in fluid communication with the can 162. For example, but not limited to, the can inlet tube 142 may be used to connect to the catheter 108.

[0071] Tank 162 can be selectively coupled and removed from pump assembly 160. (See reference) Figure 2A In some cases, the can release button 202 can be configured to selectively release the can 162 from the pump assembly 160. One or more microphones (located on the pump assembly 160 or externally, such as in the remote computing device 334) can detect the click of the can release button 202 to facilitate the detection of can placement, removal, or replacement. Reference Figure 2B The tank 162 may have one or more fill lines or scales 204 to indicate to the user the amount of fluid or exudate stored in the tank 162.

[0072] The wound treatment device 110 may have a handle 208 for lifting or carrying the wound treatment device 110. The handle 208 may be coupled to the pump assembly 160 and is rotatable relative to the wound treatment device 110, such that the handle can be rotated upwards for lifting or carrying the wound treatment device 110 or the pump assembly 160, or rotated to a lower profile in a more compact position when the handle is not in use. In some cases, the handle 208 may be coupled to the pump assembly 160 in a fixed position. The handle 208 may be coupled to the upper portion of the pump assembly 160 or may be removed from the wound treatment device 110.

[0073] Figure 3 A schematic diagram of a control system 300 is illustrated, which can be used in any wound treatment device described herein, such as wound treatment device 110. Electrical components are operable to accept user input, provide output to the user, operate a pressure source, provide connectivity, etc. A first processor (e.g., main controller 310) may be responsible for user activities, and a second processor (e.g., pump controller 370) may be responsible for controlling another device, such as pump 390.

[0074] Input / output (I / O) module 320 can be used to control inputs and / or outputs to another component or device, such as pump 390, one or more sensors (e.g., one or more pressure sensors 325 configured to monitor pressure at one or more locations along a fluid flow path), etc. For example, the I / O module can receive data from one or more sensors through one or more ports, such as serial (e.g., I2C), parallel, mixed ports, etc. Any of the pressure sensors can be part of a wound treatment device or canister. In some cases, any of the pressure sensors 325 can be located remotely from the wound treatment device, such as at or near a wound (e.g., in a dressing or in a catheter connecting the dressing to the wound treatment device). In such embodiments, any of the remote pressure sensors can communicate with the I / O module via a wired connection or with one or more transceivers 340 via a wireless connection.

[0075] The main controller 310 can receive data from and provide data to one or more expansion modules 360, such as one or more USB ports, SD ports, optical disc (CD) drives, DVD drives, FireWire ports, Thunderbolt ports, PCI Express ports, etc. The main controller 310, together with other controllers or processors, can store data in memory 350 (e.g., one or more memory modules), which can be internal or external to the main controller 310. Any suitable type of memory can be used, including volatile or non-volatile memory, such as RAM, ROM, magnetic storage, solid-state memory, magnetoresistive random access memory (MRAM), etc.

[0076] The main controller 310 can be a general-purpose controller, such as a low-power processor or a dedicated processor. The main controller 310 can be configured as the "central" processor in the electronic architecture of the control system 300, and it can coordinate the activities of other processors, such as the pump controller 370, one or more communication controllers 330, and one or more additional processors 380. The main controller 310 can run a suitable operating system, such as Linux, Windows CE, VxWorks, etc.

[0077] Pump controller 370 controls the operation of pump 390, which can generate negative pressure or reduce pressure. Pump 390 can be a suitable pump, such as a diaphragm pump, peristaltic pump, rotary pump, rotary vane pump, rolling pump, screw pump, liquid ring pump, diaphragm pump operated by a piezoelectric transducer, voice coil pump, etc. Pump controller 370 can use data received from one or more pressure sensors 325 to measure the pressure in the fluid flow path, calculate the fluid flow rate, and control the pump. Pump controller 370 can control the pump actuator (e.g., a motor) to achieve a desired negative pressure level in wound 104. The desired negative pressure level can be a pressure set or selected by the user. Pump controller 370 can use pulse width modulation (PWM) or pulse control to control the pump (e.g., a pump motor). The control signal used to drive the pump can be a 0-100% duty cycle PWM signal. Pump controller 370 can perform flow rate calculations and detect warnings. Pump controller 370 can transmit information to main controller 310. Pump controller 370 can be a low-power processor.

[0078] Pump controller 370 can monitor the activity of pump 390. For example, it can monitor the speed of pump actuators (such as motors), the duty cycle of pump actuators, etc. Activity can indicate flow rate and can be used to detect blockages, leaks, tank overfilling, etc.

[0079] Any one of the one or more communication controllers 330 can provide connectivity (e.g., wired or wireless connection 332). The one or more communication controllers 330 can utilize one or more transceivers 340 to send and receive data. The one or more transceivers 340 may include one or more antennas, optical sensors, optical transmitters, vibration motors or transducers, vibration sensors, acoustic sensors, ultrasonic sensors, etc. Any one of the one or more transceivers 340 can function as a communication controller. In this case, one or more communication controllers 330 may be omitted. Any one of the one or more transceivers 340 can be connected to one or more antennas that facilitate wireless communication. The one or more communication controllers 330 can provide one or more of the following types of connectivity: Global Positioning System (GPS), cellular connectivity (e.g., 2G, 3G, LTE, 4G, 5G, etc.), NFC, Bluetooth connectivity (or BLE), Radio Frequency Identification (RFID), Wireless Local Area Network (WLAN), Wireless Personal Area Network (WPAN), WiFi connectivity, Internet connectivity, optical connectivity (e.g., using infrared light, barcodes, such as QR codes, etc.), acoustic connectivity, ultrasonic connectivity, etc. The connection can be used for a variety of activities, such as pump component location tracking, asset tracking, compliance monitoring, remote selection, log uploading, alerts and other operational data, as well as treatment setting adjustments, software or firmware upgrades, pairing, etc.

[0080] Any one of the one or more communication controllers 330 may provide dual GPS / cellular functionality. Cellular functionality may be, for example, 3G, 4G, or 5G. The one or more communication controllers 330 may transmit information to the main controller 310. Any one of the one or more communication controllers 330 may include internal memory or available memory 350. Any one of the one or more communication controllers 330 may be a low-power processor.

[0081] The control system 300 can store data, such as GPS data, treatment data, device data, and event data. This data can be stored, for example, in a memory 350. This data may include patient data collected by one or more sensors. The control system 300 can track and record treatment and other operational data. Such data can be stored, for example, in a memory 350.

[0082] Using a connection provided by one or more communication controllers 330, the control system 300 can upload any data stored, maintained, or tracked by the control system 300 to a remote computing device, such as device 334. The control system 300 can also (e.g., via a connection to device 334) download various operational data, such as treatment selections and parameters, firmware and software patches and upgrades, etc. One or more additional processors 380 may be used, such as processors for controlling one or more user interfaces (e.g., one or more displays). In some cases, depending on the implementation of a wound monitoring or treatment system in which the control system 300 is used, any of the components illustrated or described in the control system 300 may be omitted.

[0083] Motion sensor 328 can monitor the movement of the wound treatment device. Motion sensor 328 can be one or more accelerometers of a gyroscope.

[0084] Any negative pressure wound therapy device described herein may include one or more features disclosed in U.S. Patent No. 9,737,649 or U.S. Patent Publication No. 2017 / 0216501, each of which is incorporated herein by reference in its entirety.

[0085] Multiple dressing negative pressure wound therapy

[0086] Figure 4Another negative pressure wound treatment system 400 is illustrated. System 400 may include a wound treatment device, such as wound treatment device 110, capable of supplying negative pressure to one or more wound sites. Wound treatment device 110 may be in fluid communication with one or more wound dressings 406a, 406b (collectively referred to as 406) to supply negative pressure to one or more wounds, such as wounds 104a and 104b. A first fluid flow path may include components providing a fluid connection from wound treatment device 110 to first wound dressing 406a. As a non-limiting example, the first fluid flow path may include a path from wound dressing 406a to wound treatment device 110 or a path from first wound dressing 406a to inlet 446 of branch attachment (or connector) 444 fluidly connected to wound treatment device 110. Similarly, a second fluid flow path may include components providing a fluid connection from wound treatment device 110 to second wound dressing 406b.

[0087] Apart from the fact that multiple wounds 104a and 140b are being treated by system 400, system 400 may be similar to system 100. System 400 may include any one or more components of system 100, which in Figure 4 The additional letters "a" or "b" are used to distinguish between the first and second wounds (e.g., wounds 104a and 104b, dressings 106a and 106b). As shown, system 400 may include multiple wound dressings 406a, 406b (and corresponding fluid flow paths) in fluid communication with wound treatment device 110 via multiple suction adapters (e.g., adapter 108). The suction adapter may include any one or more components of adapter 108, which in Figure 4 The additional letters “a” or “b” are used to distinguish the first wound and the second wound (e.g., bridging portions 130a and 130b, connectors 134a and 134b, and caps 140a and 140b).

[0088] Wound treatment device 110 can be fluidly connected via tube 142 to inlet 446 of connector 444. Connector 444 can be fluidly connected via branches 445a, 445b and tubes or conduits 442a, 442b to connectors 134a, 134b, which can be fluidly connected to tubes or conduits 130a, 130b. Tubes or conduits 130a, 130b can be fluidly connected to wound dressings 406a, 406b. Once all conduit and dressing components are connected and operatively positioned, wound treatment device 110 can be activated, thereby supplying negative pressure to wounds 104a, 104b via fluid flow paths. Negative pressure can be applied until the desired healing level of wounds 104a, 104b is achieved. Although Figure 4The illustration shows two wounds and a wound dressing, but some embodiments of the wound treatment device 110 can provide treatment to a single wound (e.g., by closing unused branches 445a or 445b of connector 444) or to more than two wounds (e.g., by adding branches to connector 444).

[0089] System 400 may include one or more features disclosed in U.S. Patent Publication No. 2020 / 0069850, International Publication No. WO2018 / 167199, International Publication No. WO2018 / 167199 or International Publication No. WO2023 / 072704, each of which is incorporated herein by reference in its entirety.

[0090] Tankless pump assembly

[0091] Figure 5A , Figure 5B and Figure 5C Perspective, front, and rear views of a TNP or negative pressure wound therapy (NPWT) device 500 (sometimes referred to as a wound treatment device) are illustrated. The wound treatment device 500 may include a housing 502 and a mounting component 510 (such as an accessory). The mounting component 510 may be removably attached to the housing 502, allowing the wound treatment device 500 to be used with or without the mounting component 510. For example, Figure 5C An example of a wound treatment device 500 without mounting component 510 is shown. Mounting component 510 may be designed to allow the wound treatment device 500 to be mounted on another object (such as, but not limited to, a user's body). Mounting component 510 may include a clamp 504 designed to hold the mounting component 510 on the user's outer garment, such as in a pocket, pouch, belt, lapels, or other location.

[0092] The housing 502 (sometimes referred to as the “outer housing”) may contain or support components of the wound treatment device 500. The housing 502 may be formed from one or more portions, such as a front portion 502A and a rear portion 502B, which may be removably attached to form the housing 502.

[0093] The housing 502 may include a user interface 512, which may be designed to provide information to the user (e.g., information about the operational status of the wound treatment device 500). The user interface 512 may include one or more indicators, such as icon 514, which may alert the user to one or more operational or malfunction conditions of the decompression wound treatment system.

[0094] The wound treatment device 500 may include one or more user input features, such as a button 516, which are designed to receive input from a user to control the operation of the wound treatment device 500. A single button may be present, which can be used to activate and deactivate the decompression wound treatment device or to control other operating parameters of the wound treatment device 500.

[0095] The wound treatment device 500 may include a connector 530 for connecting a tube or catheter to the wound treatment device 500. The connector 530 may be used to connect the wound treatment device 500 to a wound dressing.

[0096] The wound treatment device 500 may be a canisterless device. The wound dressing may retain fluids (such as exudate) aspirated from the wound. Such dressing may include a filter, such as a hydrophobic filter, that prevents liquid downstream of the wound dressing from passing through (towards the wound treatment device 500).

[0097] The wound treatment device 500 may include a covering 518, such as Figure 5C As illustrated, the cover may be removable. Cover 518 may cover a cavity (not shown) in which one or more power sources (such as batteries) for powering the wound treatment device 500 are located.

[0098] The wound treatment device 500 may include one or more controllers or other electronic components described herein. The wound treatment device 500 may be similar to the Pico™ negative pressure wound treatment device manufactured by Smith+Nephew. In some cases, the pump of the wound treatment device 500 may be periodically activated and deactivated to save power. Leakage may be detected based on one or more of monitoring the pressure in the fluid flow path or the pump's duty cycle, and the pump may be automatically deactivated (or stopped) in response to the detection of a leak of a specific intensity. The wound treatment device 500 may be configured to operate for a limited duration (or lifespan) (such as 7 days or less, 10 days or 14 days or longer).

[0099] Any negative pressure wound therapy device described herein may include one or more features disclosed in the following patents, including leak detection, blockage detection, full-fill detection, power monitoring, and operation for a limited duration: U.S. Patent No. 8,843,327 entitled "CANISTER STATUS DETERMINATION" and published September 23, 2014; U.S. Patent No. 9,408,954 entitled "SYSTEMS AND METHODS FOR CONTROLLING OPERATION OF NEGATIVE PRESSURE WOUND THERAPY APPARATUS" and published August 9, 2016; U.S. Patent No. 9,737,649 entitled "SYSTEMS AND METHODS FOR APPLYING REDUCED PRESSURE THERAPY" and published August 22, 2017; and U.S. Patent No. CANISTER FLUID LEVELDETECTION IN REDUCED PRESSURE THERAPY. U.S. Patent No. 10,912,870, published February 9, 2021, entitled “SYSTEMS”; U.S. Patent Publication No. 2019 / 0231939; U.S. Patent No. 8,734,425, entitled “PRESSURE CONTROLAPPARATUS”, published May 27, 2014; U.S. Patent No. 8,905,985, entitled “SYSTEMS AND METHODS FOR CONTROLLING OPERATION OF A REDUCED PRESSURE THERAPY SYSTEM”, published December 9, 2014; U.S. Patent No. 9,084,845, entitled “REDUCED PRESSURE THERAPYAPPARATUSES AND METHODS OF USING SAME”, published July 21, 2015; and U.S. Patent No. 9,084,845, entitled “NEGATIVE PRESSURE WOUND THERAPY”. U.S. Patent No. 9,427,505, entitled “APPARATUS”, was published on August 30, 2016; and U.S. Patent No. 10,737,002, entitled “PRESSURE SAMPLING SYSTEMS AND METHODSFOR NEGATIVE PRESSURE WOUND THERAPY”, was published on August 11, 2020.International patent publication No. WO 2022 / 223645, entitled "CANISTER STATUS DETERMINATION FOR NEGATIVE PRESSURE WOUND THERAPY DEVICES"; and international patent publication No. WO 2023 / 110376, entitled "SMART BATTERY PACKWITH POWER SAVING MODES FOR NEGATIVE PRESSURE WOUND THERAPY DEVICES", are incorporated herein by reference in their entirety.

[0100] Any negative pressure wound therapy device described herein can perform calibration and self-testing as described in U.S. Patent Publication No. 2023 / 0037943 entitled “SELF-TESTING FOR NEGATIVE PRESSURE WOUND THERAPY DEVICES”, which is incorporated herein by reference in its entirety.

[0101] Natural Language Interface

[0102] The use of TNP devices can be complex, confusing, and cumbersome for users, such as healthcare providers or patients. Existing TNP devices provide users with user guides that anticipate certain scenarios and list instructions on what to do next. Such guides can be voluminous and difficult to interpret, especially by non-experts. It could be advantageous to streamline procedures and reduce the need for expert intervention and product support interactions by providing systems that allow for more autonomous and accurate treatment duration. Such TNP systems can leverage generative AI, such as Natural Language Processing (NLP), to automatically monitor device-related sensors and user input in real time and respond by alerting users to abnormal readings, responding to user queries, or warning of changes in system status or the need for intervention. Additionally, the system can correlate case data with patient outcomes by collecting and storing treatment data for further analysis.

[0103] NLP models (such as one or more large language models (LLMs)) can be trained using regulatory medical device information (which may be stored in a database) regarding the use and operation of treatment systems (such as any negative pressure wound therapy system described herein). Therefore, responses to user inquiries can be based on and limited to such regulatory information to comply with various medical device or healthcare laws. Regulatory information may be provided by the manufacturer of the TNP system (such as Smith+Nephew) and may include one or more of the following: user manuals, repair manuals, functional descriptions of system operation, catalogs of system accessories (such as dressings), catalogs of spare parts, instructions for use (such as instructional videos), lists of indications and contraindications, information on other compatible products, or information on follow-up treatment. Regulatory information may include the level of safety associated with the provision of treatment. The NLP training database may be provided solely by the manufacturer and regulated by the manufacturer through controls to prevent alterations or additions to the data, such as cyberattacks or malicious actors. In some cases, the database may be digitally locked to a specific version and encrypted. Any changes to the NLP training database will require a rigorous approval process by subject matter experts. For example, a change may require approval from a multidisciplinary team of experts before implementation. In cases where a change may affect the safe use of the product, it may first require approval from a government regulatory agency or notified body. Therefore, all information in the NLP training database will be verified or validated by the manufacturer to ensure compliance with all regulatory requirements. In addition to such regulatory information, NLP models may provide responses based on specific data captured during treatment delivery, such as image data, sensor data, or historical data that may be captured over one or more durations, as described in this paper regarding context windows.

[0104] As described herein, users can provide voice prompts, text, or images (such as photos, drawings, or videos) for NLP processing, which in turn provides one or more responses. For example, one or more responses may include instructions for changing one or more treatment settings, selecting and applying a specific dressing, or changing a canister. These instructions may be provided directly to the negative pressure wound therapy device and may be executed by one or more controllers of the device. As another example, one or more responses may include an interactive set of instructions that may require further responses from the user.

[0105] Figure 6AAn example is illustrated using a TNP system 600 to apply negative pressure wound therapy to a wound covered by a wound dressing 640. This TNP system includes an NPWT device 500 (which may be a Pico™ device manufactured by Smith+Nephew). The NPWT device 500 can apply negative pressure wound therapy to a wound of a patient 610 covered by the wound dressing 640. The NPWT device 500 can deliver negative pressure to the wound via a fluid flow path that may include a conduit 630, a port 660 (such as a SoftPort manufactured by Smith+Nephew), and the wound dressing 640. The TNP system 600 may include a telecomputing device 334 communicating with a cloud 690. The cloud 690 can execute one or more NLP models to assist in the application of negative pressure wound therapy. The remote computing device 334 can be operated by a user and can allow the user to request information about the operation or treatment of the NPWT device 500 based on one or more data sensed by one or more sensors of the TNP system 600 (such as pressure, motion (e.g., accelerometer or gyroscope), status data, image data, verbal prompts, etc.

[0106] The remote computing device 334 can provide a user interface (UI) that facilitates interaction with the user. The UI can be a web interface, an application, etc. Advantageously, the UI can be designed to be easy to use, ensuring that users can easily type, speak, select or transmit information, and view, listen to, or read responses. Different user accounts can be supported, and user authentication can be implemented to control access to different characteristics or personal data.

[0107] refer to Figure 6A The UI (which may be referred to as "Pico AI") may include options 672 for determining the status or treatment of the NPWT device 500, options 674 for taking photos and uploading them to the cloud 690, and options 676 for asking questions. These options can be selected using voice commands, touch, etc.

[0108] The executable code implementing the UI can interface with one or more NLP models implemented by the Cloud 690. This can be achieved using an Application Programming Interface (API) that communicates with the Cloud 690. The API can be executed by a remote computing device 334. The API can issue HTTP requests and receive response data. The API can implement rate limiting and error responses. Appropriate error handling mechanisms for the API can ensure that any issues within the API or UI are properly managed. Logging can be implemented for debugging or performance monitoring. The API can implement data authentication (e.g., by using API keys).

[0109] Since communication will take place over the internet, implementing robust security measures is likely important. These could include HTTPS for secure communication, responsible handling of user data, and ensuring compliance with data privacy regulations (such as GDPR).

[0110] Cloud 690, such as one or more servers 692, can execute application logic that processes user requests (or queries), communicates with one or more NLP models, and transmits responses to remote computing devices 334. The application logic can be built using various programming languages ​​and frameworks, such as Node.js, Python, etc. The NLP models can execute on one or more servers provided by the TNP system 600 provider or a third party.

[0111] One or more databases can be implemented in Cloud 690 to store user data, conversation history, or any other relevant information. Databases can ensure data integrity and security.

[0112] The architecture of the TNP System 600 can be scalable, especially when a large number of users are expected. This can be achieved through load balancing, efficient caching strategies, or optimized API call patterns to avoid unnecessary costs or latency.

[0113] As described in this paper, NLP models can be trained on information related to specific proprietary processing devices, processing methods, and condition-if-then data developed by the manufacturers of the processing devices. NLP models can be version controlled using traceability of training epochs and training datasets, system configurations and auxiliary software and hardware, and black-box validation testing. Product control data can be traced back to the manufacturer, for example, via specific serial numbers and documentation stored by the manufacturer.

[0114] Users can request information about the status of the NPWT device 500. The NPWT device 500 can automatically monitor device-related sensors and user input in real time and generate indications or alarms as described herein. For example, the NPWT device 500 can provide indications for normal operation, low battery, blockage, leakage, full tank, overpressure, etc. For example, in response to such an indication, a user can request information related to the status of the NPWT device 500. To request information, the user can interact with the UI on the remote computing device 334 as described herein.

[0115] Users can request the status of NPWT device 500 via the UI, and this request can be transmitted to cloud 690 for processing. A response regarding the status can be transmitted from cloud 690 and provided to the user (e.g., via the UI). The request can be received as a natural language voice command, text, or image. The response can be displayed or provided in an audio format (e.g., via telecomputing device 334). The submitted request can be associated with patient 610 (each patient can be assigned a unique patient identifier). The request can include the patient identifier and is processed by cloud 690 based on the specific patient identifier. Furthermore, cloud 690 can respond to the request by considering previous interactions associated with patient 610 (e.g., providing a context window as described herein). Therefore, patient data can be aggregated in a database that is only accessible to the NLP model when a request relates to a specific patient or device processing event. When a device is associated with a new patient, data from previous patients can be deleted or rendered inaccessible by the NLP model.

[0116] The response may include one or more instructions or information regarding the operation or treatment of the NPWT device 500, its status, modifications to treatment settings, etc. The NPWT device 500 may determine the level of wound treatment based on determinations of various sensor information collected by the NPWT device 500, wherein the information may be transmitted to a remote computing device 334 or a cloud-based server 692.

[0117] Table 1 provides an example of interaction with the TNP system 600. Table 1 provides columns for requests made by the user, responses provided by the NLP model, inputs used by the NLP model to provide responses, and actions transmitted by the cloud server 690 to the device or a third party. Table 1 assumes that negative pressure wound therapy is provided by a canisterless device (such as device 500).

[0118] Table 1: Example Interactions with the Tankless TNP System

[0119]

[0120]

[0121] As noted in conjunction with Example #2, Cloud 690 can directly issue instructions or commands to NPWT device 500, for example, to shut down (or deactivate) the negative pressure source when a minor leak is being repaired. As noted in conjunction with Example #3, Cloud 690 can communicate with third parties such as clinicians.

[0122] In some cases, the NPWT device 500 may override commands issued by the cloud 690, or user action may be required to confirm a command before the NPWT device 500 executes it. This may be done for security or compliance reasons. For example, user confirmation may be required, such as closing a negative pressure source by pressing a button (e.g., button 180 or 516). As another example, the NPWT device 500 may override the command if the warning continues to be triggered again.

[0123] Table 2 provides examples of interaction with TNP system 600, which includes devices (such as device 110) having canisters for providing negative pressure wound therapy. Table 2 provides columns for requests made by the user, responses provided by an NLP model, inputs used by the NLP model to provide responses, and actions transmitted to the device or a third party by cloud server 690. As described herein, messages can take various formats, such as voice prompts in a selected language, text or icons on the device's screen, or any text in a selected language on a separate device.

[0124] Table 2: Example Interactions with a TNP System Utilizing Tanks

[0125]

[0126]

[0127]

[0128]

[0129]

[0130] As illustrated in Example #4, Cloud 690 can directly issue instructions or commands to NPWT device 500, such as turning off (or deactivating) the negative pressure source when the can is being replaced, and turning it on (or activating) to restart treatment after the can has been replaced. Examples #2 and #3 illustrate Cloud 690 instructing NPWT device 500 to set alarms for can replacement and for battery charging, respectively. Referring to Example #3, Cloud 690 can instruct NPWT device 500 to modify the battery warning threshold indicating a low battery condition. Cloud 690 can determine the required duration for battery charging based on one or more of the following: current charge level, temperature, battery condition, and usage. Example #7 illustrates Cloud 690 instructing device 500 to generate an alarm when the battery charge has 2 hours remaining. As illustrated in Example #6, Cloud 690 can communicate with third parties such as healthcare providers or service and repair providers.

[0131] In response to issuing an alarm or providing another type of output, Cloud 690 can (e.g., via telephone, email, SMS, etc.) contact the patient's healthcare provider to convey the alarm or other type of output. Cloud 690 can transmit the alarm or other type of output to the computing devices of one or more of the healthcare providers or other parties. Cloud 690 facilitates two-way communication with the healthcare provider or other party and can relay information to the user of device 500. Cloud 690 facilitates communication between the healthcare provider or other party and third parties, for example, to order additional medical equipment, schedule medical tests, etc.

[0132] System 600 can be configured to detect the type of user interacting with the system, such as a healthcare provider or a patient. System 600 can perform this detection based on the requests made by the user. For example, a healthcare provider might make more complex requests than a patient. Additionally, the level of complexity of the healthcare provider (such as a specialist clinician, a non-specialist clinician, a nurse, etc.) can be detected. System 600 can adjust its response based on the detection of the user's type. Users can be identified by their education level or proficiency with the system, and the system can respond using language meaningful to that specific user. For example, a user might be identified as non-technical, and therefore the response can be adjusted to provide more explanation using simpler, less technical language.

[0133] Figures 6B to 6E This is an example illustration of the interaction between a user and the TNP system 600. (Example:) Figure 6B As shown, an example of user input is illustrated using a user input dialog box 670A. The user can request (e.g., using a voice command or a text command) the status of the dressing 640. The user can provide the request to a remote computing device 334, which can transmit the request to a cloud 690 for processing. The cloud 690 can provide a response, as illustrated by an output dialog box 680A. The response can be based on data collected by one or more sensors (such as pressure) and one or more photographs of the dressing provided to the cloud 690. The output dialog box 680A can provide a response related to the status of the dressing, such as the dressing being approximately 40% full. The output dialog box 680A can include step-by-step instructions from the user requesting further information.

[0134] exist Figure 6CAnother example interaction is illustrated below. A user can request an explanation of what it means for all visual indicators of device 500 to be off, as shown in input dialog box 670B. A response can indicate that power has been depleted, as illustrated in output dialog box 680B. The response can provide additional instructions for changing the power supply. One or more NLP models can provide information about the process steps for performing a specific task. As illustrated in output dialog box 680B, additional actions can be suggested, and suggestions can be made to guide the user in performing such additional actions.

[0135] Figure 6D An example interaction involving an alarm generated by the NPWT device 500 is illustrated. Information about the alarm can be provided to the user without a prior request from the user (as indicated by the empty input dialog box 670C). As described herein, the cloud 690 can receive information (such as sensor status, indications, or alarms) from the NPWT device 500. The cloud 690 can process the information and provide an interpretation to the user. As illustrated in the output dialog box 680D, the information may indicate that a dressing is full and needs to be changed, and a proposal to schedule an appointment with a healthcare provider.

[0136] Figure 7A and Figure 7B Examples of responses to user requests for state (such as combined with...) Figure 6B (as explained) or in response to an alarm (such as in combination) Figure 6D and Figure 6E (As explained) to provide the status or treatment of the NPWT device 500. The status can be provided on the UI of the remote computing device 334. The UI status screen 734 can display indicators 736 of the NPWT device 500, in particular providing a highlighted indicator 738 for alarms, or providing a graph 740 of the pressure at the wound site over time. The control 742 can provide the opportunity to type text to communicate with the cloud 690.

[0137] Users can interact with the UI status screen 734. For example, a user can select (e.g., touch) an indicator 738 and receive information from the cloud 690 about alarms and how to resolve them. For example, a user can select indicator 738 and ask, "Why is it red?" A user can select any status indicator in status indicators 736 and receive information from the cloud 690. Similarly, a user can select a pressure icon 740 (or select any specific area) and receive information about applying pressure to a wound. The cloud-based server 692 determines the status indication regarding the identification of a specific event trigger based on previously processed information.

[0138] As described herein, an NLP model can be configured to understand and process the context in which the TNP system 600 (or any other device described herein) operates. For example, the example interactions in Tables 1 and 2 illustrate the context awareness of an NLP model. Answers can be tailored for specific users (such as patients or healthcare providers) and provided within the context of the interaction with the user. This can be referred to as providing a context window that can be customized for the user. An NLP model can be configured to provide a context window for each user. To provide a context window, the NLP model can be configured to track the passage of time.

[0139] When the TNP system 600 (or any other device described herein) is set up to deliver treatment to a patient, a context window can be created for a specific user. Previous context windows can be deleted, and new context windows can be created using supervised device information, as described herein. In some cases, instead of deleting context windows for previous users, data associated with such context windows can be saved and reloaded when starting another treatment session for a previous patient (or restarting a previous treatment session). Advantageously, context windows can be created and customized for specific patient treatment sessions (or care events). Starting with the supervised information at the time of context window creation, the dataset used by the NLP model can grow based on interactions with the user. That is, the supervised information can be used as baseline information at the initiation of the context window, and additional information can be learned as the patient care event progresses. The responses or recommendations provided by the NLP model can be adjusted, taking into account such context and the level of safety associated with the delivery of treatment (which may be part of the baseline information).

[0140] Baseline information can be separated from additional information learned during patient care events to maintain the integrity of the regulatory information used as baseline information. Regulatory information can be updated (e.g., by the manufacturer), which can lead to the formation of new baseline information. Instead of deleting activity context windows created before an update, this separation allows such activity context windows to be maintained by updating their baseline information. Advantageously, the additional baseline information of the context window associated with a specific patient can "grow" independently of any updates or changes to the baseline information.

[0141] Context windows can be configured based on the user's education or knowledge level, which can be provided by the user or a third party (e.g., in response to prompts). Interactions with the user (such as responses or questions) can be customized to match the user's education or knowledge level. Additional details are provided in conjunction with publicly available information on user types. In some cases, context windows can be configured based on the patient's medical condition, which can be provided by the patient or a third party or determined using one or more sensors, including, for example, one or more pressure sensors, motion sensors (e.g., accelerometers or gyroscopes), etc. Medical conditions can include one or more of the following: patient mobility (e.g., moving or sitting), exudate levels (e.g., high or low), vital signs (e.g., good or poor), etc.

[0142] The context window can suggest a transition from cupped negative pressure wound therapy (e.g., using the Renasys negative pressure wound therapy device manufactured by Smith+Nephew) to cupless negative pressure wound therapy (e.g., using the Pico™ negative pressure wound therapy device manufactured by Smith+Nephew) in response to a determination of a decrease in exudate levels. Such a determination can be based, for example, on directly monitoring the flow rate in the fluid flow path (or the fluid level in the cup) and detecting a decrease in the flow rate over a period of time (or detecting the static level of the fluid in the cup over a period of time). Additionally or alternatively, the determination can be made by indirectly monitoring the flow rate (e.g., by monitoring the number of cup changes and detecting a decrease in the number of cup changes over a period of time). Additional details regarding the transition to cupless therapy are disclosed in International Patent Publication No. WO 2023 / 072704, which is incorporated herein by reference in its entirety.

[0143] Context windows can suggest additional or alternative adjustments to treatment. For example, a context window can recommend adjusting the negative pressure setpoint or transitioning to intermittent (or continuous) negative pressure wound therapy. As an example, a context window can suggest increasing (or decreasing) the negative pressure setpoint in response to determining that the user has previously increased (or decreased) it. The context window can ask whether the user experienced discomfort or pain before increasing the negative pressure setpoint. In another example, if the user is experiencing pain, the context window can recommend decreasing the negative pressure setpoint. In some cases, such a decrease can be automatic as a safety mechanism to protect the patient's health (e.g., if blood is detected in the fluid flow path, if the patient's vital signs deteriorate, etc.).

[0144] In some cases, NLP models can assign different weights to information during query processing and response determination. One or more of the regulatory information or prescribed treatment settings (such as negative pressure setpoint, dressing type, dressing size, etc.) can be assigned a higher weight than information obtained via a context window during a treatment session for a specific user. This can be advantageous for maintaining treatment delivery within the constraints of the regulatory information. For example, regulatory information may include the golden rule for dressing fit, namely that the wound should be completely covered by the dressing, including overlapping the dressing around the wound perimeter for a good seal. Consistent with this rule, for Pico™ dressings manufactured by Smith+Nephew, it might be desirable to select a dressing size slightly larger than the wound size so that the benefits of negative pressure therapy extend to the area around the wound. Treatment settings may include information that Pico™ dressings have been prescribed for negative pressure wound therapy, and the aforementioned dressing size rule can be assigned a higher weight and utilized by the NLP model (e.g., in response to queries related to how to place the dressing on the wound).

[0145] As another example, suppose a user issues the following query: “Give me a simple explanation of how to change the canister.” The response to this query can be weighted more heavily on using precise wording, images, or videos from regulatory information, such as instructions for use, than on using simplified language that may have been determined to be appropriate for a particular user during a care event (e.g., due to the user’s education or level of knowledge).

[0146] As another example, a user can request a wireless compliance license number for the device. The user can open the conversation in a specific foreign language and be asked if they want a license number for a region associated with that language. Upon confirming the region the user desires, NLP can provide the exact legal information provided in the instructions for use and labels related to the wireless licensing compliance of the device. This information can be verbal (e.g., spoken through a speaker), text or image data on a graphic display, or text or image data sent to a printer and printed on paper for the user. Therefore, a user may request an accurate description of the instructions for use and the legal labeling information required for such a device to comply with the laws of the region of use.

[0147] As another example, suppose a user asks a question that contradicts regulatory information. The user might ask, “What does the orange light on top of the device mean?” According to the regulatory information, the device may not have an orange status light. However, it's possible that the user previously asked about the orange light (e.g., because the user is colorblind) through a context window during the interaction. The response generated by the NLP model can be weighted according to the regulatory information to interpret the status light. For example, the response could ask the user to confirm that the color statement does not include an orange light.

[0148] As another example, regulatory information might state a battery life of 12 hours. Based on information obtained via a context window during a specific user's care event (such as current draw, battery health, or operating temperature), it could be determined that the battery could last for 14 hours. When responding to user inquiries about battery life, the weighting of the response could be biased towards a higher limit, such as the 12-hour limit specified by the regulatory information. Therefore, an NLP model could provide a response like: "Expected battery life is 12 hours. Under your current usage conditions, this could extend to up to X minutes" (where X could be 30 minutes or less, 1 hour or less, etc.). In other words, the answer could be biased towards the regulatory information.

[0149] In yet another example, the TNP device can also irrigate a wound by controlling the flow of an irrigation solution (such as saline or other physiological therapeutic solutions) into the dressing. Therefore, the user can inquire how long the current irrigation supply will last before it runs out. Based on information obtained via a context window during a specific user's care event (such as treatment settings including irrigation flow rate, previous irrigation reservoir changes, fluid changes, or treatment pauses), it can be determined that the current irrigation supply can last for 24 hours. When responding to a user's question about irrigation lifespan, the weighting of the response can be biased towards a higher limit, such as 22 hours as specified by regulatory information. This difference can account for tolerance or operational parameter effects that may affect the flow rate already determined by the manufacturer during the device's development and testing. Thus, an NLP model could provide a response like: "The expected irrigation supply lifespan is 22 hours. Under your current usage conditions, this may extend to up to X minutes" (where X could be 30 minutes or less, 1 hour or less, etc.). That is, the answer can be biased towards regulatory information.

[0150] Figures 8A to 8DExample illustrations are provided for selecting appropriate wound dressings for applying negative pressure wound therapy. The Cloud 690 can receive one or more photographs of a wound (such as an incision) or data related to the wound's anatomy or topology to recommend the appropriate type, form, and size of dressing for wound management. These interactions can occur after the wound has formed (e.g., after surgery) or before the wound has formed (e.g., before surgery). Although Figures 8A to 8D The examples in the text refer to incision sites (before surgery) or incisions (after surgery), but similar methods can be used to treat any wound site before or after surgery. The term wound site should be interpreted broadly and encompass any tissue area between or after surgeries, including closed or open wounds (which cover any incision site or incision). In some implementations, the wound site can be treated using negative pressure wound therapy prior to surgery. For example, an NPWT device 500, such as... Figures 6A to 6E or Figure 7A As illustrated herein, Pico™ dressings can be used to apply negative pressure preoperatively. The advantages of using negative pressure wound therapy to treat wound sites include, in particular, increased irrigation, reduced edema, and disinfection or decontamination. In some cases, positive pressure may be applied preoperatively in addition to or instead of negative pressure, and any of the examples described herein can be used to apply positive pressure alone or in combination with negative pressure.

[0151] Such preoperative wound preparation (sometimes referred to as tissue conditioning) can improve overall wound healing outcomes. For example, most surgical infections result from the patient's own skin microbiota contaminating the wound. By increasing blood flow before surgery, tissues can be prepared to rapidly supply cells with the body's own defensive biochemicals and nutrients. By reducing the bioburden on the patient's skin, the chance of sufficient microbiota entering the wound (e.g., incision) and causing infection is reduced. Optimizing moisture content to maintain tissue elasticity and flexibility reduces stress on wound closure methods (such as sutures, splints, or adhesives), thus reducing the chance of dehiscence.

[0152] Any negative pressure wound therapy device (such as the NPWT device 500) can treat not only the wound site area but also the surrounding area to achieve a wider treatment zone. Therefore, a suitable dressing can be larger than the wound site area and its size should be correctly determined.

[0153] refer to Figure 8ATo select a suitable wound dressing for the incision site 820 (or more generally, the wound site), the user can ask, "What size dressing do I need for the incision site?" The system can instruct the user to obtain one or more images of the wound (or the area that will form a wound) and its surroundings. For example, the instruction could be, "Hold the camera above the incision (or possible incision) so that the entire incision (or possible incision) and the anatomical location are in the field of view, and wait for the dressing outline to appear, then press the button to take an image." The user can obtain images of the incision site 820 and the surrounding anatomical structures 810. In the preoperative scenario, the user can obtain images of possible incision lines, which can be physical or virtual. Images can be uploaded to the cloud 690, and an NLP model can process the images to determine the size and shape of the wound site and its surroundings, and recommend a suitable dressing size and shape. The NLP model can utilize parameters provided by the wound dressing manufacturer (such as those used to train the NLP model) to determine the optimal fit of the dressing to the wound site. For example, parameters can indicate the minimum width of the adhesive in contact with the skin surrounding the wound site, less than 1.0 mm, less than 5.0 mm, less than 10.0 mm, etc. These parameters can depend on the type of pressure-sensitive adhesive (e.g., silicone or acrylic), reactive curing adhesive (e.g., cyanoacrylate, epoxy, polyurethane, etc.), dressing type (NPWT, conventional adhesive, primary and secondary dressings), or anatomical location (joints, trunk, head, etc.). For wound dressings that deliver negative pressure to a wider treatment area (e.g., Pico™ dressings), these parameters can be correlated with the optimal width of the intact skin surrounding the wound site to be treated.

[0154] NLP models that determine the size and shape of a wound site and its surroundings and recommend appropriate dressing sizes and shapes can perform image processing and recognition to identify and classify one or more patterns or features in an image. For example, an NLP model can identify possible incision lines (which may be physically drawn on a patient's skin or drawn in an image, as described herein). NLP models can include one or more of convolutional neural networks (CNNs), deep residual networks (ResNet), Inception-v3 (GoogleNet), VGG16 (Visual Geometry Group), or YOLO (You Only Look Once). In some implementations, NLP models can utilize Euler video magnification, as described in U.S. Patent No. 11,791,030, which is appended to Appendix A and incorporated herein by reference in its entirety. As described herein, NLP models can be trained using only data provided by manufacturers of negative pressure wound therapy devices. Such data can include various correspondences between different wound sites and dressing types and sizes.

[0155] In some implementations, the images may include infrared (IR) images, which can facilitate the identification of the size and shape of the wound site. The NLP model can use IR images to determine, for example, the shape and size of the wound site. In some cases, the size and shape of the wound site can be determined by a user using a scale 950 as described herein or a utility mounted on a remote computing device. This information can be provided to the NLP model verbally, for example, by the user.

[0156] While a treatment zone at least 50 mm from the edge of the wound (using negative pressure) can positively impact blood flow, tissue hydration, and mechanical stress on the tissue, such a zone may not be achievable for some anatomical structures (e.g., in the case of anatomical openings or the overall size of body parts). In such cases, the shape and size of the treatment zone (and therefore the recommended dressing) can be adjusted by an NLP model. For example, while an NLP model can be trained on data indicating that 50 mm is the minimum target, image data can drive a trade-off between optimal treatment and the type or size of dressing available in the manufacturer's catalog or the user's own supplies. For instance, a user could provide verbal feedback to the NLP model that only certain sizes of dressings are available, and the NLP model can process the image data to determine the most suitable dressing among those available.

[0157] In some cases, preoperative planning and management may include the following steps:

[0158] • The surgeon (or another healthcare provider) marks the incisions the surgeon intends to make on the patient's entire skin line (physically or virtually).

[0159] • The surgeon takes images of the complete incision site, including the lines.

[0160] • The image (or a description of the complete cut area) is provided to the NLP model alone or together with information about the size of the line.

[0161] The NLP model identifies and provides recommendations for appropriate dressings (such as dressing type and size, or presenting a stock unit (SKU) code or product code for a specific dressing), optionally along with an enhanced image showing the dressing positioned above the incision site with lines.

[0162] • The surgeon selects a dressing and applies it to the incision site as instructed.

[0163] • Initiate preoperative treatment of the incision site.

[0164] Alternatively, in some cases, preoperative planning and management may include the following steps:

[0165] • The surgeon (or another healthcare provider) takes an image of the patient’s proposed incision site.

[0166] • The surgeon draws the proposed incision line on the image (physically or virtually).

[0167] • The image (or a description of the complete cut area) is provided to the NLP model alone or together with information about the size of the line.

[0168] The NLP model identifies and provides recommendations for appropriate dressings (such as dressing type and size), optionally along with an enhanced image showing the dressing positioned above the incision site with lines.

[0169] • The surgeon selects a dressing and applies it to the incision site as instructed.

[0170] • Initiate preoperative treatment of the incision site.

[0171] In some cases, anatomical or topological data collected by sensors (such as light detection and ranging (LIDAR) sensors or scanners) can be uploaded to the cloud 690 to facilitate dressing selection. Such anatomical or topological data can be used by NLP models. The remote computing device 334 can be equipped with various sensors (such as image sensors, LIDAR sensors, or scanners) for determining wound site dimensions.

[0172] NLP models can utilize one or more images, anatomical or topological data, and manufacturer data regarding dressing size and shape (such as catalog data) to provide appropriate dressing recommendations. For example, a dressing recommendation could be in the form of "use the 10cm x 30cm PICO dressing with product code 66802013." Figure 8B As illustrated, a specific recommended dressing can be overlaid on the incision site 820 in real time to provide the user with a visual representation of how the dressing should be placed above the incision site. The overlay 830 can be illustrated in the UI of the remote computing device 334. The system can provide instructions for dressing placement (such as video instructions).

[0173] The NLP model can generate or recommend dressings with optimal shapes. These shapes can match the shape of the wound site. The generated shapes can be overlaid on the dressing in real time (e.g., on a display of a remote computing device 334). This allows the user to cut the dressing into the optimal shape, as described herein.

[0174] In some cases, shapes can be printed to scale on a standard desktop printer and then superimposed on dressings as templates for clinicians to draw or cut. Thus, physical templates can be created to help users shape dressings. In another example, the NLP model can generate G-code or other suitable code to operate the digital controls of manufacturing machines (such as laser cutters, milling machines, or 3D printers) to cut, shape, or produce dressings according to the shapes defined by the NLP model. In some situations, a single dressing may not be sufficient to cover the required area, and in such cases, multiple shaped dressings can be combined to cover the treatment site on the patient.

[0175] Figure 8C and Figure 8D Another example illustrates the selection of appropriate dressings for incision sites 850 in different parts of the body. In this case, a specific recommended dressing size could be 15x30cm (product code 66802017).

[0176] Wound fillers or dressings can be similarly selected for use alone or in conjunction with different forms of treatment. Examples of such wound fillers and dressings are described in this article. Wound fillers can be used when it is important to achieve a good fit with the wound site itself, allowing the dressing to seal the wound site and remain in place as the patient moves, rather than sticking the dressing to the periphery of the wound site.

[0177] Figure 9A An example of using enhancement to select an appropriate dressing is illustrated. A user can upload an image of wound site 940 to cloud 690, which can determine a segmented region 930 for placing the wound dressing. As described herein, additional data, such as data related to anatomical features, can be provided. The segmented region 930 can be determined based on features identifying wound site 940. The segmented region 930 can include wound site 940 and the area surrounding the wound site, as described herein. The segmented region 930 can be overlaid in real-time on the UI of telecomputing device 920 to provide an enhanced view. The user can adjust the size, shape, or positioning of the segmented region 930 in the enhanced view. As described herein, recommendations regarding dressing size, shape, and form can be provided to match the segmented region. Instructions for placing the dressing can also be provided.

[0178] Figure 9B Examples of the same Figure 9AA similar environment is provided, in which a ruler 950 is added to help determine the segmentation region 930. One or more NLP models executed in the cloud 690 can perform image processing based on measurements provided by the ruler 950 to accurately determine the segmentation region 930. The ruler 950 may include markers to help determine the size of the wound site 940. The ruler 950 may include a color grid to verify the color of the wound site 940. A support 960 for securing a telecomputing device during the process of selecting an appropriate dressing for the wound site 940 is also illustrated. The device support 960 may be adjustable (e.g., pivotable) to properly position the telecomputing device 920. The device support 960 may include one or more of a bracket, retainer, mount, arm, etc. The device support 960 and the telecomputing device 920 may be manipulated by a user (e.g., a healthcare professional).

[0179] Figure 9C A further example illustrates user 970 interacting with cloud 690 using voice commands. For instance, voice commands can be directed to remote computing device 920 and relayed to cloud 690. User 970 can upload an image of wound site 940 to cloud 690. User 970 can provide additional information about the condition of the wound site.

[0180] Figure 10 An example of using augmented reality (AR) to place a suitable wound dressing is illustrated. A user 1070 wearing an AR headset 1030 (or another AR device) can determine the size of the wound dressing 1020 based on instructions received from the cloud 690. The instructions can be generated based on image 1010. The user 1070 can be guided in an augmented reality space, in which, for example, a dressing template or wound site can be displayed to the user to facilitate the preparation of the wound dressing 1020. The user 1070 can utilize an augmented head-up display with actual visibility of reality, which overlays information showing where to cut the wound dressing 1020 (e.g., foam or gauze) and how to orient and position the dressing on the wound. Guidance can be provided hands-free. The wound dressing 1020 being prepared or positioned can be overlaid on one or more templates or wound sites in the view provided by the headset 1030. This allows the user 1070 to follow the template or wound site to more accurately prepare and place the wound dressing 1020.

[0181] Figure 11 An example is shown using AR or augmentation to place an appropriate wound dressing on the wound site of patient 1120. (As in combination) Figures 9A to 9C As described, user 970 can use a remote computing device supported by support 1160 to prepare wound dressing 1020 and place it on the wound site. Alternatively or additionally, such as in combination Figure 10As described, a user can use the AR headset 1030 to prepare a wound dressing 1020 and place it on the wound site. The support 1160 (shown as a bracket) may include an articulated arm for positioning the telecomputing device 920.

[0182] Figure 12 The preparation of wound dressing 1020 is illustrated. Segmentation markers (or guide markers) 1210 may be presented to the user to aid in the preparation of wound dressing 1020. Guide markers 1210 may be virtual guide markers superimposed on wound dressing 1020. In some cases, guide markers 1210 may be physical guide markers or a combination of physical and virtual guide markers. Cloud 690 may provide guide markers 1210 to help the user cut the wound dressing into a defined shape and size. Additional or different visual, auditory, or tactile aids may be provided to the user to prepare and place wound dressing 1020.

[0183] Figure 13A A support 1160 for positioning a telecomputing device 920 is illustrated. The support 1160 is illustrated as a movable support capable of rolling to different positions. The support 1160 may include one or more articulated arms for supporting the telecomputing device 920. A user can use the support 1160 to utilize the remote mobile device hands-free.

[0184] Figure 13B Another support 1260 is illustrated. Similar to support 1160, support 1260 can be a movable support with an articulated arm. Multiple devices or tools can be positioned on the arm, such as a first electronic device 1330A, a magnifying glass 1330B, a second electronic device 1330C, and a measuring device 1330D (such as a ruler). Additional or alternative devices or tools can be supported by support 1260.

[0185] Figure 13C Another support member 1360 is illustrated. Support member 1360 may be a movable support capable of actuating itself. Support member 1360 may actuate itself based on one or more commands received from an external device (e.g., wirelessly). Support member 1360 may include an electric motor and a battery pack, enabling it to actuate itself. Support member 1360 may operate autonomously. Support member 1360 may include one or more motorized articulated arms for supporting remote computing device 920. Support member 1360 may include one or more sensors 1340, such as cameras. Support member 1360 may include one or more sensors for assisting in the selection of wound dressings, such as one or more LiDAR sensors or scanners.

[0186] Figure 14Another support 1460 is illustrated. While support 1460 may be similar to support 1360, support 1460 may be equipped with features for printing a wound dressing 1410 of suitable size, shape, and form. For example, support 1460 may include one or more additive printers (or 3D printers). Support 1460 can create the wound dressing 1410 from scratch and determine the size and shape of the wound dressing template to match the defined dimensions and shape of the wound dressing or wound site. Support 1460 may receive instructions from cloud 690 (or from remote computing device 920) to prepare a suitable wound dressing. Support 1460 may utilize one or more additive printers to accurately cut the wound dressing to the provided size and shape. Support 1460 may then dispense the wound dressing.

[0187] In some implementations, AI systems trained using information from regulatory medical devices to provide NLP interfaces can be used to communicate and interact with waterjet and plasma ablation debridement systems that enable clinicians to remove dead and necrotic tissue and debris from wounds. When using such systems, the success of the procedure often depends on the user's skill. The user must make expert judgments to select appropriate disposable handpieces and power settings on the device to ensure good results.

[0188] Therefore, users can request the AI ​​system to suggest appropriate settings for the type of surgery, anatomical location, or type of debris to be removed. For example, a user could ask, “I have a patient with a small diabetic ulcer on their foot that needs debridement, with some very dry eschar that needs to be removed—which handheld device and machine settings should I use?” In response, based on technical performance and operational data, the NLP model can provide a response for a specific handheld part number and device settings (which can be confirmed verbally and on the device's display). The NLP model can also provide clinicians with the option to accept device settings via a user interface. Thus, in this specific example, based on information about the surgery provided by the clinician, from Figure 15 The six options for the handheld device illustrated can be interpreted in the language of a clinician: “We recommend the VERSAJET II Plus disposable handheld device 66800045, and suggest starting with setting 5 on the device and fine-tuning the settings based on performance.”

[0189] In some cases, such as by Figure 15As illustrated in 1502, an image of the product code or recommended handheld device can be displayed on the device's screen or on another device running the associated application (such as a smartphone running an application the clinician is logged into). The clinician can then request a description of that particular handheld device. An example response from an NLP model could be: "This handheld device is ideal for removing non-viable tissue from wounds as described, it has a 45-degree angled cutting tip and an 8mm wide cutting window, and it is ideal for debridement of small wounds."

[0190] During surgery, clinicians can verbally interact with the device without taking their attention away from the procedure. For example, a clinician could ask, “How long will it take for the saline solution to run out?” Based on a database containing technical data, the elapsed time of the current saline supply, and the device controller settings, the NLP model can determine the expected time for the saline solution to run out and provide a response (such as verbal notification or a visual response).

[0191] Wound diagnosis or wound status devices can interact in a similar manner using AI systems trained on information from monitored medical devices, providing users with an NLP interface. Such devices can use image capture to determine the status or clinical diagnosis of a wound's condition. For example, UV fluorescence can be used to determine the bacterial load in a wound. Users can request help interpreting data or images, such as, "How much of the wound is colonized by bacteria and what types of bacteria are present?" Based on technical data contained in a database and analysis of wound images, the NLP model will provide responses such as a verbal description of the types of bacteria present and a display of both, along with the percentage area of ​​the wound associated with each type of bacteria. The NLP model can provide and enhance images of the wound periphery filled with tinted areas, which indicate different bacterial types via different colors, and also provide the percentage area for each bacterial type, such as... Figure 16 As illustrated. Alternatively or concurrently, the user may request that the image be sent to a printer, transmitted via email to an email address, or associated with the patient's electronic medical record (or otherwise transmitted electronically).

[0192] In some implementations, an AI system trained using information from a monitored medical device can be used to interact with wound prevention devices in a similar manner, providing an NLP interface for the user. For example, such devices can monitor patient positioning or pressure on a surface over time to provide caregivers with information on when patient positioning should be changed to prevent pressure injuries. Many patients in a ward can send their data to a central monitoring or care station. With other tasks pending, clinicians might ask questions like, “Which patients will need to be turned over in the next 10 minutes?” Based on technical data contained in a database and analysis of individual patient data, an NLP model can provide a list of patients (e.g., verbally and on a graphical display). Clinicians can also request the NLP model to send the list to a phone or portable tablet or print it out, making it available during ward visits and ensuring no patient is missed. Alternatively, clinicians can request individual patient identification by triggering alarms on patient monitoring devices (e.g., providing audible alarms on each patient monitoring device).

[0193] Other variations

[0194] While some embodiments describe negative pressure wound therapy, the systems, devices, and / or methods disclosed herein can be applied to other types of treatment, either alone or in addition to TNP therapy. The systems, devices, and / or methods disclosed herein are extendable to any medical device, and specifically, to any wound monitoring and / or treatment device. For example, the systems, devices, and / or methods disclosed herein can be used with devices that provide one or more of ultrasound therapy, oxygen therapy, nerve stimulation, microwave therapy, active agents, antibiotics, antimicrobial agents, etc. Additionally, such devices can provide TNP therapy. As another example, the systems, devices, and / or methods disclosed herein can be used with wound debridement systems, patient monitoring systems, etc. The systems and methods disclosed herein are not limited to medical devices and can be utilized by any electronic device.

[0195] Any data transfer described herein can be performed securely. For example, one or more of the following can be used: encryption, HTTPS protocol, secure VPN connection, error checking, delivery confirmation, etc.

[0196] While some examples involve the use of NLP models or LLM, other types of AI or machine learning models may be used additionally or alternatively.

[0197] Any values ​​for thresholds, limits, durations, etc., provided herein are not intended to be absolute and are therefore approximate. Furthermore, any thresholds, limits, durations, etc., provided herein may be fixed or automatically or by the user. Additionally, related terms relative to reference values, such as exceeding, greater than, less than, etc., as used herein, are intended to also cover being equal to the reference value. For example, exceeding a positive reference value may cover being equal to or greater than the reference value. Furthermore, relative terms relative to reference values, such as exceeding, greater than, less than, etc., as used herein, are also intended to cover the opposite relationships disclosed, such as being below, less than, greater than, etc., relative to the reference value.

[0198] Features, materials, characteristics, or groups described in connection with a particular aspect, embodiment, or example shall be construed as applicable to any other aspect, embodiment, or example described herein, unless incompatible therewith. All features disclosed in this specification (including any appended claims, abstract, and drawings) and / or all steps of any method or process so disclosed may be combined in any combination, unless at least some of these features and / or steps are mutually exclusive combinations. Protection is not limited to the details of any of the foregoing embodiments. Protection extends to any novel feature or any novel combination of features disclosed in this specification (including any appended claims, abstract, and drawings), or any novel feature or any novel combination of steps of any method or process so disclosed.

[0199] While certain embodiments have been described, these embodiments are presented by way of example only and are not intended to limit the scope of protection. In fact, the novel methods and systems described herein can be embodied in a variety of other forms. Furthermore, various omissions, substitutions, and changes to the form of the methods and systems described herein are possible. Those skilled in the art will recognize that in some embodiments, the actual steps taken in the illustrated and / or disclosed processes may differ from those shown in the figures. Depending on the embodiment, some steps described above may be removed, and other steps may be added. For example, the actual steps and / or the order of steps taken in the disclosed processes may differ from those shown in the figures. The various components illustrated in the figures or described herein can be implemented as software and / or firmware on a processor, controller, ASIC, FPGA, and / or dedicated hardware. The software or firmware may include instructions stored in a non-transitory computer-readable storage medium. These instructions may be executed by a processor, controller, ASIC, FPGA, or dedicated hardware. Hardware components such as controllers, processors, ASICs, FPGAs, etc., may include logic circuit systems. Furthermore, the features and attributes of the specific embodiments disclosed above may be combined in different ways to form additional embodiments, all of which fall within the scope of this disclosure.

[0200] Furthermore, the various exemplary logic blocks and modules described in conjunction with the embodiments disclosed herein can be implemented or executed by a machine, such as a machine learning service server, DSP, ASIC, FPGA or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A machine learning service server may be or include a microprocessor, but alternatively, it may be or include a controller, microcontroller, or state machine, or a combination thereof, configured to generate and publish machine learning services supported by machine learning models. A machine learning service server may include circuitry configured to process computer-executable instructions. Although this document primarily describes digital technologies, a machine learning service server may also primarily include analog components. For example, some or all of the modeling, simulation, or servicing algorithms described herein may be implemented in analog circuitry or mixed analog and digital circuitry. The computing environment may include any type of computer system, including but not limited to microprocessor-based computer systems, host computers, digital signal processors, portable computing devices, device controllers, or computing engines within appliances, etc.

[0201] Elements of the methods, processes, routines, or algorithms described in conjunction with the embodiments disclosed herein can be directly embodied in hardware, in software modules executed by a machine learning service server, or a combination of both. Software modules can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disks, removable disks, CD-ROMs, or any other form of non-transitory computer-readable storage medium. An exemplary storage medium can be coupled to the machine learning service server, enabling the machine learning service server to read information from and write information to the storage medium. Alternatively, the storage medium can be integrated into the machine learning service server. The machine learning service server and storage medium can reside in an ASIC. The ASIC can reside in a user terminal. Alternatively, the machine learning service server and storage medium can reside as discrete components in a user terminal (e.g., an access device or a network service client device).

[0202] The user interface screens illustrated and described herein may include additional and / or optional components. These components may include menus, lists, buttons, text boxes, labels, radio buttons, scroll bars, sliders, check boxes, combo boxes, status bars, dialog boxes, windows, etc. User interface screens may include additional and / or alternative information. Components may be set up, grouped, and displayed in any suitable order.

[0203] The conditional language used herein, such as “can,” “may,” “can,” “may,” “e.g.,” etc., unless otherwise expressly stated or otherwise understood in the context in which they are used, is generally intended to express that certain embodiments include, while other embodiments do not, certain features, elements, and / or states. Therefore, this conditional language is not generally intended to imply that one or more embodiments require features, elements, and / or states in any way, or that one or more embodiments must include logic for determining, with or without user input or prompting, whether such features, elements, and / or states are included in any particular embodiment or will be performed in any particular embodiment. The terms “including,” “comprising,” “having,” etc., are synonymous and used in an open-ended manner, and do not exclude additional elements, features, actions, operations, etc. Additionally, the term “or” is used in its inclusive sense (rather than in its proprietary sense) such that, when used, for example, to connect lists of elements, the term “or” means one, some, or all of the elements in the list. Furthermore, as used herein, the term “each,” in addition to having its ordinary meaning, can also mean any subset of a set of elements to which the term “each” is applied. Additionally, when used in this application, the terms “in this text,” “above,” “below,” and similar terms refer to the application as a whole, and not to any particular part of the application.

[0204] Unless otherwise specifically stated, connective language such as the phrase "at least one of X, Y, and Z" should be understood in conjunction with the commonly used context to express that an item, term, etc., can be X, Y, or Z, or a combination thereof. Therefore, such connective language is not generally intended to imply that certain embodiments require the separate presence of at least one of X, at least one of Y, and at least one of Z.

[0205] The degree language used herein, such as the terms “about,” “approximately,” “generally,” and “roughly,” indicates a value, quantity, or characteristic that is close to a specified value, quantity, or characteristic, which still performs the desired function or achieves the desired result. For example, the terms “about,” “approximately,” “generally,” and “substantially” can refer to a quantity that is less than 10%, less than 5%, less than 1%, less than 0.1%, and less than 0.01% of a specified quantity. As another example, in some implementations, the terms “generally parallel” and “roughly parallel” refer to a value, quantity, or characteristic that deviates from exact parallelism by less than or equal to 15 degrees, 10 degrees, 5 degrees, 3 degrees, 1 degree, or 0.1 degrees.

[0206] Unless otherwise explicitly stated, articles such as “a” or “an” should generally be interpreted as including one or more of the described items. Thus, phrases such as “devices configured as…” are intended to include one or more of the described devices. Such devices of one or more descriptions may also be collectively configured to perform the description.

[0207] While this disclosure includes certain embodiments, examples, and applications, those skilled in the art will understand that this disclosure extends beyond the specific disclosed embodiments to other alternative embodiments and / or uses, as well as their obvious modifications and equivalents, including embodiments that do not provide all the features and advantages set forth herein. Therefore, the scope of this disclosure is not intended to be limited by the specific disclosure of preferred embodiments herein, and may be defined by the claims set forth herein or any claims to be filed hereafter.

Claims

1. A negative pressure and / or positive pressure therapy system, said negative pressure and / or positive pressure therapy system comprising: Treatment device, the treatment device comprising: A pressure source configured to provide pressure therapy to a wound site covered by a wound dressing; A controller, configured to control the operation of the pressure source; and Communication circuitry, configured to facilitate wireless communication with the treatment device; and A storage medium storing executable instructions that, when executed by one or more processors, cause the one or more processors to: Receive information associated with providing the pressure therapy from the communication circuit; Receive natural language queries associated with providing the pressure therapy, the natural language queries being issued by the user of the therapy device; The natural language query is processed and a response is determined using one or more machine learning models trained solely on data provided by the manufacturer of the treatment device; and The response is provided to one or more of the user or the controller.

2. The system as claimed in any of the preceding claims, wherein the response comprises a set of instructions for adjusting the operation of the pressure source.

3. The system as claimed in any of the preceding claims, wherein providing the response to the controller causes the controller to adjust the operation of the pressure source.

4. The system of claim 3, wherein the controller deactivates or activates the pressure source.

5. The system as claimed in any of the preceding claims, wherein the natural language query relates to one or more of the state of the treatment device or the state of the wound dressing.

6. The system as claimed in any of the preceding claims, wherein the controller is configured to provide an alarm associated with providing the stress treatment, wherein the natural language query is associated with the alarm, and wherein the response includes instructions for remedying the alarm.

7. The system of claim 1, wherein the pressure source includes a negative pressure source configured to provide negative pressure therapy to the wound site.

8. The system as described in any of the preceding claims, wherein: The natural language query is associated with selecting the wound dressing from a plurality of wound dressings suitable for the treatment device; The information includes one or more images of the wound, which optionally include the outline of an incision to be made at the wound site; and The response includes one or more of the type, shape, or size of the wound dressing suitable for providing the pressure treatment to the wound site, and optionally, the response is determined based on the contour of the incision to be made at the wound site.

9. The system of claim 8, wherein: Determining the response also includes generating a wound dressing shape that matches the shape of the wound site; and Providing the response includes displaying the shape of the wound dressing superimposed on the wound site.

10. The system of claim 8 or 9, wherein providing the response includes displaying the wound dressing superimposed on the wound site.

11. The system as claimed in any of the preceding claims, wherein providing the response includes displaying the response.

12. The system as claimed in any of the preceding claims, wherein the one or more processors are located in one or more servers.

13. The system as claimed in any of the preceding claims, wherein the natural language query comprises one or more of speech or text.

14. The system as claimed in any of the preceding claims, wherein the information includes one or more of image data or data collected by one or more sensors.

15. A method for controlling a negative pressure and / or positive pressure therapy system, the method comprising: Pressure therapy is provided from a pressure source to a wound site covered by a wound dressing by a treatment device, the pressure source being controlled by a controller; as well as One or more processors located remotely from the treatment device: Receive natural language queries associated with providing the pressure therapy, the natural language queries being issued by the user of the therapy device; The natural language query is processed and the response is determined using one or more machine learning models trained solely on data provided by the manufacturer of the treatment device; as well as The response is provided to one or more of the user or the controller.

16. The method of any of the preceding claims, wherein the response comprises a set of instructions for adjusting the operation of the pressure source.

17. The method of any of the preceding claims, wherein providing the response to the controller causes an operation to adjust the pressure source, and optionally, wherein adjusting the pressure source includes deactivating or activating the pressure source.

18. The method of any of the preceding claims, wherein the natural language query relates to one or more of the state of the treatment device or the state of the wound dressing.

19. The method of any of the preceding claims, further comprising providing an alarm by the treatment device in association with providing the pressure treatment, wherein the natural language query is associated with the alarm, and optionally, wherein the response includes instructions for remedying the alarm.

20. The method as claimed in any of the preceding claims, wherein: The method includes processing the natural language query and determining a response based on one or more images of the wound; The natural language query is associated with selecting a wound dressing from a plurality of wound dressings suitable for the treatment device, and the one or more images optionally include the outline of an incision to be made at the wound site; and The response includes one or more of the type, shape, or size of the wound dressing suitable for providing the pressure treatment to the wound site; optionally, the response is determined based on the contour of the incision to be made at the wound site; and Providing the response includes displaying the wound dressing superimposed on the wound site.

21. The method of claim 20, wherein: Determining the response also includes generating a wound dressing shape that matches the shape of the wound site; and Providing the response includes displaying the shape of the wound dressing superimposed on the wound site.

22. A negative pressure and / or positive pressure therapy system, said negative pressure and / or positive pressure therapy system comprising: Treatment device, the treatment device comprising: A pressure source configured to provide pressure therapy to a wound site covered by a wound dressing; A controller, configured to control the operation of the pressure source; and Communication circuitry, configured to facilitate wireless communication with the treatment device; and A storage medium storing executable instructions that, when executed by one or more processors, cause the one or more processors to: One or more machine learning models are used to initiate an interactive session in a baseline state, the interactive session in the baseline state utilizing the one or more machine learning models trained only with data provided by the manufacturer of the treatment device to provide a response to natural language queries issued by the user of the treatment device; Receive information associated with providing the pressure therapy from the communication circuit; Update the state of the interactive session, the updated state of the interactive session including the baseline state and also including the information associated with providing stress therapy; Receive a natural language query associated with providing the pressure therapy, the natural language query being issued by the user of the therapy device; and The natural language query is processed and a response is determined using one or more machine learning models trained solely on data provided by the manufacturer of the therapeutic device and using the information associated with providing stress therapy; and The response is provided to one or more of the user or the controller.

23. The system of claim 22, wherein the information associated with providing the stress therapy includes at least one of the following: the user's level of knowledge, data collected by one or more sensors, or changes in the parameters of the stress therapy over a period of time.

24. The system of any one of claims 22 or 23, wherein determining the response comprises assigning a higher priority to information associated with the baseline state than to information associated with providing the stress treatment.

25. A medical system, the medical system comprising: Medical device, the medical device comprising: A processing module configured to provide treatment to a patient; A controller, configured to control the operation of the processing module; and Communication circuitry, configured to facilitate wireless communication with the medical device; and A storage medium storing executable instructions that, when executed by one or more processors, cause the one or more processors to: Receive information associated with providing the processing from the communication circuit; Receive and provide natural language queries associated with the processing, the natural language queries being issued by the user of the medical device; The natural language query is processed and a response is determined using one or more machine learning models trained solely on data provided by the manufacturer of the medical device; and The response is provided to one or more of the user or the controller.

Images