TECHNICAL FIELD

[0001] Various embodiments described in the present disclosure relate to systems, controllers and methods for a risk assessment of a pressure ulcer occurring to a living entity (e.g., a human or an animal), particularly in the area of neonatology and pediatric care, patients in an intensive care unit (ICU) or a ward, and sedentary or immobile patients at home or in a nursing facility.

BACKGROUND

[0002] Babies in a neonatal intensive care unit (NICU) and children in a pediatric intensive care unit (PICU) have demonstrated a high prevalence of pressure ulcers, such as, for example, as high as 23% of babies in NICUs have experienced pressure ulcers and as high as 27% of children in PICUs have experienced pressure ulcers. Moreover, acutely ill, sedated, paralyzed or immobilized neonates are at higher risk of pressure ulcers.

[0003] More particularly, risk factors for infants in the NICU include, but are not limited to, (1) intrinsic factors for skin breakdown, such as, for example, duration and amount of pressure, friction, shear, and moisture, and (2) extrinsic factors, such as, for example, perfusion, malnutrition, infection, anemia and immobility. The sacrum, the largest bony area, is the most common location for pressure ulcers in adults. However, in the pediatric population, the occiput is the largest bony prominence and the most common site of pressure ulcer development.

[0004] There are several stages of pressure ulcers. Stage I is non-blanchable erythematous skin that may be painful, soft, warmer or cooler than adjacent tissue. Stage II has partial dermal loss, (e.g., shallow open ulcer or an intact blister). Stage III has dermal loss wherein subdermal elements are visualized. Stage IV ulcers are full thickness tissue loss with exposed bone, tendon or muscle. Stages III and IV are serious reportable events.

[0005] Additionally, in a large study observing 741 neonatal patients, the study deduced that pressure ulcers due to devices was nearly 80% overall and over 90% in preterm infants. Hospitalized neonates are susceptible to device-related injury and the rate of stage II injury is high.

[0006] To address the problem of pressure ulcers in the NICU, there have been several scales/guidelines proposed. One of them is the Glamorgan scale, which is based on literature and expert opinions and has eleven (11) statistically significant pediatric pressure ulcer risk factors.

[0007] The first factor is an inability to move without great difficulty or deterioration in condition or having prolonged surgery.

[0008] The second factor is an inability to change position without

assistance/inability to control body movement.

[0009] The third factor is a degree of mobility, but reduced for age.

[0010] The fourth factor is equipment/objects/hard surface pressing or rubbing on skin.

[0011] The fifth factor is significant anemia (e.g., hemoglobin < 9 g/dL).

[0012] The sixth factor is persistent pyrexia (e.g.., temperature > 37.5C for more than 12 hours).

[0013] The seventh factor is poor peripheral perfusion (e.g., cold

extremities/capillary refill > 2 seconds/cool mottled skin).

[0014] The eight factor is inadequate nutrition (e.g., unable to take/not absorbing oral or enteral feeds and not supplemented with hyperalimentation).

[0015] The ninth factor is low serum albumin level (e.g., <3.5 g/dL).

[0016] The tenth factor is low weight for age (e.g., < lOth percentile).

[0017] The eleventh factor is incontinence (if inappropriate for age).

[0018] At a cutoff score of 15, the Glamorgan Scale has been found to be 98.4% sensitive and have a specificity of 67.4%.

[0019] Applying the Glamorgan scale or the like in an NICU or PICU requires nurses to check the babies quite often. This an added effort and workflow for NICU nurses to calculate some of the scales for pressure ulcer risk, yet the nurses may miss important risk factors, especially if the susceptible part of the body is occluded by clothes. Also, assessment of some factors on the Glamorgan scale and accurate tracking them over time is difficult, such as, for example, (1) NICU nurses having no means of judging how long body part(s)/area(s) have been in certain positions, raising the risk of pressure ulcers, (2) NICU nurses having no objective measure of peripheral versus central perfusion and its change over time, which is a risk factor, and (3) NICU nurses unwittingly failing to notice areas around wires/devices showing poor blood flow or inflammation.

[0020] The inventions of the present disclosure address an ideal of providing technology to objectively assess many elements of a pressure ulcer risk scale (e.g., the Glamorgan scale), which lead to an unburdensome, meticulous, continuous monitoring of the risk of pressure ulcers.

SUMMARY

[0021] To objectively assess many elements of the Glamorgan scale or any other pressure ulcer risk scale, the inventions of the present disclosure are premised on monitoring primary risks factors for a pressure ulcer by a pressure ulcer candidate including, but not limited to, (1) stationary body part(s) of the pressure ulcer candidate for a significant time period, (2) equipment/wires affecting irritation of the pressure ulcer candidate and (3) poor peripheral perfusion of the pressure ulcer candidate. These primary risk factors may be combined with secondary risk factors including, but not limited to, demographics of the pressure ulcer candidate (e.g., age and weight), laboratory/clinical testing results (e.g., anemia, pryrexia and serum albumin) and diagnostic/clinical observations (e.g., incontinence).

[0022] For purposes of describing and claiming the inventions of the present disclosure, the term "pressure ulcer candidate" broadly encompasses a live person/animal, a physical simulation of a person/animal (e.g., a doll or a mannequin) or a virtual simulation of a person/animal (e.g., a computer simulation of a human model) in a physical environment or a virtual environment susceptible to pressure ulcers (e.g., a NICU, a PICU, a ICU, a ward, a nursing facility or otherwise any environment whereby the pressure ulcer candidate is sedentary/immobilized).

[0023] To monitor such primary risk factors, the inventions of the present disclosure incorporate one or more cameras (e.g., digital camera(s), an infrared camera(s) and/or video camera(s)) and/or one or more physiological sensors (e.g., accelerometer(s), pressure sensor(s) and temperature sensor(s)). [0024] The camera(s) are strategically positioned within the physical environment or the virtual environment (e.g., located and oriented to focus on particular areas of the pressure ulcer candidate) to provide (1) an imagery detection of stationary body part(s) of the pressure ulcer candidate for a significant time period, (2) an imagery detection of equipment/wires affecting irritation of the pressure ulcer candidate and/or (3) an imagery detection of poor peripheral perfusion of the pressure ulcer candidate.

[0025] The physiological sensor(s) are practically positioned relative to the pressure ulcer candidate (e.g., within a mattress or a chair, or within textiles) to provide (1) a sensory detection of stationary body part(s) of the pressure ulcer candidate for a significant time period and/or (2) a sensory detection of poor peripheral perfusion of the pressure ulcer candidate.

[0026] The inventions of the present disclosure further incorporate a clinical decision engine for inputting images from the camera(s) and/or data from the

physiological sensor(s) to thereby render a prediction and/or a classification of the pressure ulcer risk of the pressure ulcer candidate. The clinical decision support engine may also identify pressure ulcer risk area(s) of the pressure ulcer candidate (e.g., a baby’s head is in position for three (3) hours, wires irritating a baby's arm or poor peripheral perfusion of the soles of a baby's feet) whereby the identified pressure ulcer risk area(s) of the pressure ulcer candidate may be highlighted in a displayed image of the pressure ulcer candidate.

[0027] One embodiment of the inventions of the present disclosure is a pressure ulcer risk assessment system employing one or more cameras for generating imaging data of the pressure ulcer candidate, and/or one or more physiological sensors for generating physiological data of the pressure ulcer candidate. The system further employs a pressure ulcer risk assessment controller for ascertaining a pressure ulcer risk of the pressure ulcer candidate by inputting the imaging data and/or the physiological data into a clinical decision support engine (e.g., a deep learning model based software/firmware program stored on a non-transitory machine readable storage medium and/or a deep learning model based electronic circuit) to render a prediction and/or a classification of the pressure ulcer risk of the pressure ulcer candidate. The clinical decision support engine may further identify one or more body part(s)/area(s) of the pressure ulcer candidate having the highest risk of a pressure ulcer, which may be superimposed on a displayed camera image. The clinical decision support engine may further ascertain a confidence level of the prediction and/or the classification of the pressure ulcer risk of the pressure ulcer candidate.

[0028] A second embodiment of the present disclosure is a non-transitory machine-readable storage medium with instructions for execution by one or more processors for ascertaining a pressure ulcer risk of a pressure ulcer candidate. The non- transitory machine-readable storage medium comprising instructions to receive imaging data of the pressure ulcer candidate generated by one or more cameras and/or to receive physiological data of the pressure ulcer candidate generated by one or more physiological sensors. The non-transitory machine-readable storage medium further comprises instructions to input the imaging data and/or the physiological data into a clinical decision support engine (e.g., a deep learning model based software/firmware program stored on a non-transitory machine readable storage medium and/or a deep learning based electronic circuit) to render a prediction and/or a classification of the pressure ulcer risk of the pressure ulcer candidate. The non-transitory machine -readable storage medium may further comprise instructions to identify one or more body part(s)/area(s) of the pressure ulcer candidate having the highest risk of a pressure ulcer, which may be superimposed on a displayed camera image. The non-transitory machine -readable storage medium may further comprises instructions to ascertain a confidence level of the prediction and/or the classification of the pressure ulcer risk of the pressure ulcer candidate.

[0029] A third embodiment of the inventions of the present disclosure is a pressure ulcer risk assessment method for ascertaining a pressure ulcer risk of a pressure ulcer candidate. The pressure ulcer risk assessment method involves a generation, by one or more cameras, of imaging data of the pressure ulcer candidate and/or a generation, by one or more physiological sensors, of physiological data of the pressure ulcer candidate. The pressure ulcer risk assessment method further involves inputting, by a pressure ulcer risk assessment controller, the imaging data and/or the physiological data into a clinical decision support engine e.g., a deep learning model based software/firmware program stored on a non-transitory machine readable storage medium and/or a deep learning based electronic circuit) to render a prediction and/or a classification of the pressure ulcer risk of the pressure ulcer candidate. The pressure ulcer risk assessment method may further involve an identification, by the clinical decision support engine, of one or more body part(s)/area(s) of the pressure ulcer candidate having the highest risk of a pressure ulcer, which may be superimposed on a displayed camera image. The pressure ulcer risk assessment method may further involve an ascertaining, by the clinical decision support engine, of a confidence level of the prediction and/or the classification of the pressure ulcer risk of the pressure ulcer candidate.

[0030] Also for purposes of describing and claiming the inventions of the present disclosure:

[0031] (1) the terms of the art of the present disclosure including, but not limited to, "camera", "physiological sensor", "graphical user interface", "database", "database manager", "deep learning model", "neural network", "support vector machine", "logic circuit", "display manager" and "communication manager" are to be broadly interpreted as known in the art of the present disclosure and exemplary described in the present disclosure;

[0032] (2) the term "clinical decision support engine" broadly encompasses any type engine, known in the art of the present disclosure or hereinafter conceived, that is configured in accordance with the inventive principles of the present disclosure for a deep learning model based risk assessment of pressure ulcer candidates as exemplarily described in the present disclosure;

[0033] (3) the term“controller” broadly encompasses all structural configurations, as understood in the art of the present disclosure and as exemplary described in the present disclosure, of an application specific main board or an application specific integrated circuit for controlling an application of various inventive principles of the present disclosure as subsequently described in the present disclosure. The structural configuration of the controller may include, but is not limited to, processor(s), computer- usable/computer readable storage medium(s), an operating system, application module(s), peripheral device controller(s), slot(s) and port(s);

[0034] (4) the term“module” broadly encompasses electronic

circuitry/hardware and/or an executable program (e.g., executable software stored on non- transitory computer readable medium(s) and/or firmware) incorporated within or accessible by a controller for executing a specific application;

[0035] (5) the descriptive label(s) for term“module” herein facilitates a distinction between modules as described and claimed herein without specifying or implying any additional limitation to the term“module”; and

[0036] (6) “data” may be embodied in all forms of a detectable physical quantity or impulse (e.g., voltage, current, magnetic field strength, impedance, color) as understood in the art of the present disclosure and as exemplary described in the present disclosure for transmitting information and/or instructions in support of applying various inventive principles of the present disclosure as subsequently described in the present disclosure. Data communication encompassed by the inventions of the present disclosure may involve any communication method as known in the art of the present disclosure including, but not limited to, data transmission/reception over any type of wired or wireless datalink and a reading of data uploaded to a computer-usable/computer readable storage medium.

[0037] The foregoing embodiments and other embodiments of the inventions of the present disclosure as well as various features and advantages of the present disclosure will become further apparent from the following detailed description of various embodiments of the inventions of the present disclosure read in conjunction with the accompanying drawings. The detailed description and drawings are merely illustrative of the inventions of the present disclosure rather than limiting, the scope of the inventions of present disclosure being defined by the appended claims and equivalents thereof.

BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to better understand various example embodiments, reference is made to the accompanying drawings, wherein:

[0039] FIG. 1 illustrates exemplary embodiments of a pressure ulcer risk assessment method in accordance with the present disclosure;

[0040] FIG. 2 illustrate exemplary embodiments of a pressure ulcer risk assessment system in accordance with the present disclosure;

[0041] FIG. 3 illustrates an exemplary embodiment of a pressure ulcer risk assessment controller in accordance with the present disclosure;

[0042] FIG. 4A illustrates a first exemplary embodiment of a clinical support decision engine in accordance with the present disclosure; [0043] FIG. 4B illustrates a first exemplary embodiment of a clinical support decision engine in accordance with the present disclosure;

[0044] FIG. 4C illustrates a first exemplary embodiment of a clinical support decision engine in accordance with the present disclosure; and

[0045] FIG. 4D illustrates a first exemplary embodiment of a clinical support decision engine in accordance with the present disclosure.

DETAILED DESCRIPTION

[0046] The description and drawings presented herein illustrate various principles.

It will be appreciated that those skilled in the art will be able to devise various

arrangements that, although not explicitly described or shown herein, embody these principles and are included within the scope of this disclosure. As used herein, the term, “or,” as used herein, refers to a non-exclusive or (i.e., and/or), unless otherwise indicated (e.g.,“or else” or“or in the alternative”). Additionally, the various embodiments described in the present disclosure are not necessarily mutually exclusive and may be combined to render additional embodiments that incorporate the principles described in the present disclosure.

[0047] To facilitate an understanding of the inventions of the present disclosure, the following descriptions of FIG. 1 teaches exemplary pressure ulcer risk assessment methods of the present disclosure, FIG. 2 teaches exemplary pressure ulcer risk assessment systems of the present disclosure and FIG. 3 teaches exemplary pressure ulcer risk assessment controllers of the present disclosure. From the description of FIGS. 1-3, those having ordinary skill in the art of the present disclosure will appreciate how to apply the present disclosure for making and using numerous and various additional embodiments of pressure ulcer risk assessment methods, systems and controllers of the present disclosure.

[0048] Referring to FIG. 1, a flowchart 10 represents exemplary pressure ulcer risk assessment methods of the present disclosure for ascertaining a pressure ulcer risk of a pressure ulcer candidate (e.g., a live person/animal, a physical simulation of a person/animal or a virtual simulation of a person/animal).

[0049] A stage S12 of flowchart 10 encompasses an imaging by one or more cameras 20 (e.g., photographic camera(s) 21, such as, for example, a digital camera and/or an infra-red camera or picture camera(s) 22, such as, for example, a video camera) and/or a sensing by one or more physiological sensors 30 (e.g., accelerometer(s), pressure sensor(s) and/or temperature sensor(s)) of primary risks factors for a pressure ulcer by a pressure ulcer candidate including, but not limited to, (1) stationary body part(s) of the pressure ulcer candidate for a significant time period, (2) equipment/wires affecting irritation of the pressure ulcer candidate and/or (3) poor peripheral perfusion of the pressure ulcer candidate.

[0050] In practice, camera(s) 20 are strategically positioned within a physical environment or a virtual environment (e.g., located and oriented to focus on particular areas of the pressure ulcer candidate) to provide (1) an imagery detection of stationary body part(s) of the pressure ulcer candidate for a significant time period, (2) an imagery detection of equipment/wires affecting irritation of the pressure ulcer candidate an imagery detection equipment/wires affecting irritation of the pressure ulcer candidate and/or (3) an imagery detection of poor peripheral perfusion of the pressure ulcer candidate.

[0051] Physiological sensor(s) 30 are practically positioned relative to the pressure ulcer candidate (e.g., within a mattress or a chair, or within textiles) to provide (1) a sensory detection of stationary body part(s) of the pressure ulcer candidate for a significant time period and/or (2) a sensory detection of poor peripheral perfusion of the pressure ulcer candidate.

[0052] For example, as shown in FIG. 2, a pressure ulcer candidate lOOa having equipment and wires attached thereto (e.g., a feeding tube, IV(s), LEDs, etc.) is lying on a mattress 10 la embodying pressure sensors and/or temperature sensors, as symbolized by the dots on the top surface of mattress lOla, for generating physiological data 31 (FIG. 1) of pressure ulcer candidate lOOa (e.g., pressure data indicative of pressure applied to various points of the pressure ulcer candidate or temperature data indicative of a temperature of various points of the pressure ulcer candidate). A video camera 20a and/or a digital/infrared camera 20b are positioned above a foot of mattress 10 la for generating imaging data 23 (FIG. 1) of pressure ulcer candidate lOOa. For embodiments utilizing both cameras, the video camera 20a may utilized to image the entire body of pressure ulcer candidate lOOa, while the digital camera 20b may be utilized to image a specific body part of pressure ulcer candidate lOOa.

[0053] By further example, also shown in FIG. 2, a pressure ulcer candidate lOOb is lying on a mattress 10 lb incorporating pressure sensors and/or temperature sensors as symbolized by the dots on the top surface of mattress 10 lb for generating physiological data 31 (FIG. 1) of pressure ulcer candidate lOOb and wearing a hat and clothing incorporating accelerometers, pressure sensors and/or temperature sensors as symbolized by the dots on the hat and clothing for generating physiological data 31 (FIG. 1) of pressure ulcer candidate lOOb . A video camera 20c and/or a digital camera 20d (or an infrared camera) are positioned above a foot of mattress 10 lb for generating imaging data 23 (FIG. 1) of pressure ulcer candidate lOOb (e.g., pressure data indicative of pressure applied to various points of the pressure ulcer candidate or temperature data indicative of a temperature of various points of the pressure ulcer candidate). Similarly, for embodiments utilizing both cameras, the video camera 20c may utilized to image the entire body of pressure ulcer candidate lOOb, while the digital camera 20d may be utilized to image a specific body part of pressure ulcer candidate lOOb.

[0054] Referring back to FIG. 1, stage S12 of flowchart 10 may further encompass an acquisition by one or more graphical user interfaces 40 and/or a database manager 50 controlling a database 51 of secondary risk factors for a pressure ulcer by the pressure ulcer candidate including, but not limited to, demographics of the pressure ulcer candidate (e.g., age and weight), laboratory/clinical testing results (e.g., anemia, pryrexia and serum albumin) and diagnostic/clinical observations (e.g., incontinence).

[0055] Still referring to FIG. 1, a stage S14 of flowchart 10 encompasses an input of imaging data 23, physiological data 31 and candidate data 41 (i.e., the secondary risk factors acquired by GUI(s) 40 and/or database manager 50) into a clinical decision support engine 60 to render a pressure ulcer prediction 62a (e.g., a risk score of a presence of a pressure ulcer in the pressure ulcer candidate) and/or a pressure ulcer classification 62b (e.g., a risk classification of the pressure ulcer candidate as having a pressure ulcer, more than likely having a pressure ulcer, more than likely not having a pressure ulcer and not having a pressure ulcer). To render pressure ulcer prediction 62a and/or pressure ulcer classification 62b, clinical decision support engine 60 employs an X number of deep learning models 61 (e.g., deep learning neural networks, neural networks, boosting classifiers or supervised vector machines), X > 1. [0056] In one embodiment of a deep learning model 61, a deep learning neural network as known in the art of the present disclosure encompass multiple layers of activation neurons interconnected and trained for feature extraction and transformation to thereby calculate complex mappings between input data (i.e., 2D pixel or 3D voxel imaging data 23, physiological data 31 and/or candidate data 41) and one or more output neurons (i.e., a risk score or risk classifications). Each activation neuron applies a nonlinear activation function to a weighted linear combination of inputs (e.g., input data, an upstream activation neuron output or a downstream activation neuron output). As such, the parameters of importance to a deep neural network are the structure of the connective neuron network, the nonlinear activation functions and weights of the activation neurons. The deep learning neural network may be trained on data derived from pressure ulcer patients and/or data derived from non-pressure ulcer patients.

[0057] In a second embodiment of a deep learning model 61, a support vector machine as known in the art of the present disclosure encompasses support vectors trained for delineating an optimal hyperplane for linearly or nonlinearly separating feature vectors of input data (i.e., 2D pixel or 3D voxel imaging data 23, physiological data 31 and/or candidate data 41) into one of two classes (e.g., a high risk of pressure ulcer and a low risk of pressure ulcer). The support vector machine may be trained on data derived from pressure ulcer patients and/or data derived from non-pressure ulcer patients.

[0058] In practice, imaging data 23, physiological data 31 and candidate data 41 may be pre-processed if necessary and (1) collectively inputted into a single deep learning model 61 to render pressure ulcer prediction 62a and/or pressure ulcer classification 62b, (2) collectively inputted into a network of deep learning models 61 with the outputs of the Siamese deep learning models 61 being inputted into a logical circuit or another deep learning model 61 to render pressure ulcer prediction 62a and/or pressure ulcer classification 62b (e.g., a Siamese architecture of one deep learning model 61 trained on data derived from pressure ulcer patients and another deep learning model 61 trained on data derived from non-pressure ulcer patients) or (3) individually inputted into different deep learning models 61 to render pressure ulcer prediction 62a and/or pressure ulcer classification 62b with the output of each deep learning model 61 being inputted into a logical circuit or another deep learning model to render pressure ulcer prediction 62a and/or pressure ulcer classification 62b as will be further described in the present disclosure with reference to FIGS. 4A-4D.

[0059] Additionally, in practice, clinical decision support engine 60 may further produce pressure ulcer prediction 62a and/or pressure ulcer classification 62b with an identification of one or more body part(s)/area(s) of the pressure ulcer candidate having the highest risk of a pressure ulcer (e.g., an occiput of the pressure ulcer candidate has the highest risk of a pressure ulcer), which may be superimposed on a displayed camera image, and/or with a confidence level of the prediction and/or the classification of the pressure ulcer risk of the pressure ulcer candidate (e.g., the pressure ulcer candidate has a 80% risk of pressure ulcer with 60% confidence).

[0060] Clinical decision support engine 60 may also issue a monitoring alert 63 concerning a particular primary risk factor, particularly prior to an indication by pressure ulcer prediction 62a and/or pressure ulcer classification 62b of a definitive risk or a likely risk of pressure ulcer of the pressure ulcer candidate. For example, a monitoring alert 63 as issued by clinical decision support engine 60 may represent a warning of a stationary body part of the pressure ulcer candidate for a significant time period, equipment/wires affecting irritation of the pressure ulcer candidate and/or poor peripheral perfusion of the pressure ulcer candidate prior to any indication by pressure ulcer prediction 62a and/or pressure ulcer classification 62b of a definitive risk or a likely risk of pressure ulcer of the pressure ulcer candidate. Also by example, a monitoring alert 63 as issued by clinical decision support engine 60 may represent an output of physiological sensor 30 or outputs of a collective grouping of physiological sensors is(are) greater than or less than a threshold related to a primary factor of a pressure ulcer risk of the pressure risk candidate (e.g., a warning of a stationary body part of the pressure ulcer candidate for a significant time period exceeding a time threshold or a warning of a level of pressure being applied to the pressure ulcer candidate exceeding a pressure threshold).

[0061] Still referring to FIG. 1, a stage S16 of flowchart 10 encompasses a reporting by one or more display managers 70 and/or one or more communication managers 80 of the pressure ulcer prediction 62a and/or the pressure ulcer classification 62b, and a reporting by display manager(s) 70 and/or a communication manager(s) 80 of a monitoring alert 63 if applicable. [0062] In practice, a display manager 70 may control a visual reporting (e.g., a displayed camera image 71), a textual reporting (e.g., a textual description of pressure ulcer prediction 62a and/or the pressure ulcer classification 62b) and/or an audible reporting (e.g., a pre-recorded audible description of pressure ulcer prediction 62a and/or the pressure ulcer classification 62b) via one or more displays (not shown) (e.g., a patient monitor or a central station). In one embodiment, clinical decision support engine 60 identifies one or more body part(s)/area(s) of the pressure ulcer candidate having a highest risk of a pressure ulcer, and a display manager 70 superimposes the highest risk body part(s)/area(s) on a displayed camera image 71 of the pressure ulcer candidate. To this end, display manager 70 may be a display controller as known in the art of the present disclosure or an application module of a pressure ulcer risk assessment controller as will be further described in the present disclosure with reference to FIG. 3.

[0063] Further in practice, a communication manager 80 may control a printing, a textual transmission, an emailing and/or a filing pressure ulcer prediction 62a and/or the pressure ulcer classification 62b, and a monitoring alert 63 if applicable. To this end, a communication manager 80 may be a printer controller, an email controller, a mobile device controller or a file manager controller as known in the art of the present disclosure or an application module of a pressure ulcer risk assessment controller as will be further described in the present disclosure with reference to FIG. 3.

[0064] Also in practice, stages S12-S16 of flowchart 10 may be performed in a continuous basis or on regular/irregular intermittent intervals.

[0065] Still referring to FIG 1, in practice, clinical decision support engine 60 is embodied on one or more pressure ulcer risk assessment controllers as either a software/firmware program stored on a non-transitory machine readable storage medium and/or as an electronic circuit for implementing deep learning model(s) 61.

[0066] For example, as shown in FIG. 2, pressure ulcer risk assessment controller(s) 120 of the present disclosure is(are) installed within an application server 110 accessible by a plurality of clients (e.g., a client 111 and a client 112 as shown) and/or is(are) installed within a workstation 113 employing a monitor 114, a keyboard 115 and a computer 116. In practice, application server 110 may be a stand-alone server or a server incorporated with a broader monitoring system, for example, a patient monitoring system or a central monitoring station. Also in practice, workstation 113 may be a stand-alone workstation or a workstation incorporated with a broader monitoring system, for example, a patient monitoring system or a central monitoring station.

[0067] In operation as shown in FIG. 2, a pressure ulcer risk assessment controller

120 receives imaging data 23 from cameras 20a-20d and/or physiological data 31 from the physiological sensors. The pressure ulcer risk assessment controller 120 inputs imaging data 23 and physiological data 31 into the clinical decision support engine 60 (FIG. 2) to render pressure ulcer prediction 62, pressure ulcer classification 62b and/or monitoring alert 63 as previously described in the present disclosure, which is(are) communicated by the controller 120 to a variety of reporting sources 130 including, but not limited to, a printer 131, a mobile phone 133, a tablet 133, a patient monitor 134, a central monitoring database management 135, an email server (not shown) and a file server (not shown).

[0068] In practice, a pressure ulcer risk assessment controller(s) 120 may be implemented as hardware/circuity/software/firmware.

[0069] In one embodiment as shown in FIG. 3, a , pressure ulcer risk assessment controller l20a includes a processor 121, a memory 122, a user interface 123, a network interface 124, and a storage 125 interconnected via one or more system bus(es) 126. In practice, the actual organization of the components 121-125 of controller l20a may be more complex than illustrated.

[0070] The processor 121 may be any hardware device capable of executing instructions stored in memory or storage or otherwise processing data. As such, the processor 121 may include a microprocessor, field programmable gate array (FPGA), application-specific integrated circuit (ASIC), or other similar devices.

[0071] The memory 122 may include various memories such as, for example Ll,

L2, or L3 cache or system memory. As such, the memory 122 may include static random access memory (SRAM), dynamic RAM (DRAM), flash memory, read only memory (ROM), or other similar memory devices.

[0072] The user interface 123 may include one or more devices for enabling communication with a user such as an administrator. For example, the user interface 123 may include a display, a mouse, and a keyboard for receiving user commands. In some embodiments, the user interface 123 may include a command line interface or graphical user interface that may be presented to a remote terminal via the network interface 124. [0073] The network interface 124 may include one or more devices for enabling communication with other hardware devices. For example, the network interface 124 may include a network interface card (NIC) configured to communicate according to the Ethernet protocol. Additionally, the network interface 124 may implement a TCP/IP stack for communication according to the TCP/IP protocols. Various alternative or additional hardware or configurations for the network interface will be apparent.

[0074] The storage 125 may include one or more machine -readable storage media such as read-only memory (ROM), random-access memory (RAM), magnetic disk storage media, optical storage media, flash-memory devices, or similar storage media. In various embodiments, the storage 125 may store instructions for execution by the processor 121 or data upon with the processor 121 may operate. For example, the storage 125 store a base operating system (not shown) for controlling various basic operations of the hardware.

[0075] More particular to the present disclosure, storage 125 further stores control modules 127 including an embodiment of clinical decision report engine 60, and may further include, as shown, an embodiment of graphical user interface 40, an embodiment of database manager 50, an embodiments of display manager 70 and an embodiment of communication manager 80.

[0076] To facilitate a further understanding of the inventions of the present disclosure, the following descriptions of FIGS. 4A-4D teaches exemplary embodiments of a clinical decision support engine of the present disclosure. From the description of FIGS. 4A-4D, those having ordinary skill in the art of the present disclosure will appreciate how to apply the present disclosure for making and using numerous and various additional embodiments of clinical decision support engines of the present disclosure.

[0077] Referring to FIG. 4A, one embodiment of a clinical decision support engine of the present disclosure employs data preprocessors 64a-64d (which may be integrated as one data preprocessor), a sampled neural network 6 la, a sampled neural network 6 lb, an encoded neural network 6lc, an embedded neural network 6ld and a logic circuit 67a.

[0078] In operation, data preprocessor 64a is a module having an architecture for sampling imaging data 23 to render sampled data 65a. Imaging data 23 may be sampled as a time series of images generated by the camera(s) or in a time series of differential data between sequential image generated by the camera(s).

[0079] Data preprocessor 64b is a module having an architecture for sampling physiological data 31 to render sampled data 65b. Physiological data 31 may be sampled as a time series of data point(s) generated by the physiological sensor(s) or in a time series of differential data between sequential data points generated by the physiological sensor(s).

[0080] Data preprocessor 64c is a module having an architecture for encoding candidate data 4la related to demographics and diagnostic/clinical observations of the pressure ulcer candidate to render encoded data 65c (e.g., one-hot coded binary coded or autoencoding).

[0081] Data preprocessor 64d is a module having an architecture for embedding candidate data 4la related to (e.g., word embedding) laboratory/clinical testing of the pressure ulcer candidate to render embedded data 65d.

[0082] Sampled neural network 6 la is a module having a neural architecture trained for analyzing sampled data 65a to learn predictive features as related to any imagery detection of a pressure ulcer and inputs sampled data 65a to render a pressure ulcer risk assessment 66a (e.g., a pressure ulcer prediction or a pressure ulcer

classification based on imaging data 23), which is communicated to a logic circuit 67a and display/communication managers 70/80.

[0083] A nonlimiting exemplary neural architecture of sampled neural network

6la is a two-stage convolutional network (e.g., a recurrent neural network), which may provide a confidence level with the pressure ulcer risk assessment 66a.

[0084] Sampled neural network 6 lb is a module having a neural architecture trained for analyzing sampled data 65b to leam predictive features as related to any sensory detection of the a pressure ulcer and inputs sampled data 65b to render a pressure ulcer risk assessment 66b (e.g., a pressure ulcer prediction or a pressure ulcer

classification based on physiological data 31), which is communicated to a logic circuit 67 a and display/communication managers 70/80.

[0085] A nonlimiting exemplary neural architecture of sampled neural network

6 lb is a two-stage convolutional network (e.g. recurrent neural network), which may provide a confidence level with the pressure ulcer risk assessment 66b. [0086] Encoded neural network 6lc is a module having a neural architecture trained for analyzing encoded data 65c to leam predictive features as related to

demographics and diagnostic/clinical observations of the pressure ulcer candidate and inputs encoded data 65c to render a pressure ulcer risk assessment 66c (e.g., a pressure ulcer prediction or a pressure ulcer classification based on candidate data 4la), which is communicated to a logic circuit 67a and display/communication managers 70/80.

[0087] A non-limiting neural architecture of encoded neural network 6lc is a deep learning network (e.g., multilayer perceptrons), which may provide a confidence level with the pressure ulcer risk assessment 66c.

[0088] Embedded neural network 6ld is a module having a neural architecture trained for analyzing embedded data 65d to leam predictive features laboratory/clinical testing of the pressure ulcer candidate and inputs embedded data 65d to render a pressure ulcer risk assessment 66d (e.g., a pressure ulcer prediction or a pressure ulcer

classification based on candidate data 4 lb), which is communicated to a logic circuit 67a and display/communication managers 70/80.

[0089] A nonlimiting exemplary neural architecture of embedded neural network

6ld is a one-stage convolutional network (e.g. inception architecture), which may provide a confidence level with the pressure ulcer risk assessment 66d.

[0090] In practice, the neural architectures of sampled neural network 6 la, sampled neural network 6 lb, encoded neural network 6lc and/or embedded neural network 6ld may include an attention module as known in the art of the present disclosure.

[0091] Logic circuit 67a is a module implementing a circuit for logically producing pressure ulcer prediction 62a (FIG. 1) or pressure ulcer classification 62b (FIG. 1) from the pressure ulcer assessments 66a-66d. In a risk score embodiment, logic circuit 67a inputs press pressure ulcer assessments 66a-66d into an average logic circuit. In a risk classification embodiment, logic circuit 67a inputs pressure ulcer assessments 66a- 66d into an AND gate, which serves to select an output of a multiplexor between a high risk of pressure ulcer input and a low risk of pressure ulcer input.

[0092] Furthermore, each pressure ulcer assessments 66a-66d may individually serve to initiate a monitoring alert 63 (FIG. 1) when predicting a pressure ulcer of the corresponding input data or classifying the corresponding data as representing a pressure ulcer. For example, while pressure ulcer assessments 66a-66d collectively are indicating a low risk of a pressure ulcer of the pressure ulcer candidate, pressure ulcer assessments 66a may be indicating a high risk of a pressure ulcer of the pressure ulcer candidate whereby pressure ulcer assessment 66a is used to initiate the monitoring alert 63.

[0093] Referring to FIG. 4B, a second embodiment of a clinical decision support engine of the present disclosure, alternative to the embodiment of FIG. 4A, employs a sampled neural network 6le, a sampled neural network 6 lf, an encoded neural network 6lg, an embedded neural network 6lh and a convolutional neural network 68a.

[0094] Sampled neural network 6le is a module having a neural architecture trained for analyzing sampled data 65a to learn predictive features as related to any imagery detection of a pressure ulcers and inputs sampled data 65a to render a sampled feature vector 66e representative of the predictive features of sampled data 65a, which is communicated to a convolutional neural network 68a.

[0095] A nonlimiting exemplary neural architecture of sampled neural network

6le is a two-stage convolutional network (e.g., a recurrent neural network), which may provide a confidence level with the pressure ulcer risk assessment 66e.

[0096] Sampled neural network 6 lf is a module having a neural architecture trained for analyzing sampled data 65b to leam predictive features as related to any sensory detection of the a pressure ulcer and inputs sampled data 65b to render a sampled feature vector 66f representative of the predictive features of sampled data 65b, which is communicated to a convolutional neural network 68a.

[0097] A nonlimiting exemplary neural architecture of sampled neural network

6 lf is a two-stage convolutional network (e.g., a recurrent neural network), which may provide a confidence level with the pressure ulcer risk assessment 66f.

[0098] Encoded neural network 6lg is a module having a neural architecture trained for analyzing encoded data 65c to leam predictive features as related to demographics and diagnostic/clinical observations of the pressure ulcer candidate and inputs encoded data 65c to render an encoded feature vector 66g representative of the predictive features of sampled data 65c, which is communicated to a convolutional neural network 68a. [0099] A non-limiting neural architecture of encoded neural network 6lg is a deep learning network (e.g. multilayer perceptrons), which may provide a confidence level with the pressure ulcer risk assessment 66g.

[00100] Embedded neural network 6lh is a module having a neural architecture trained for analyzing embedded data 65d to leam predictive features laboratory/clinical testing of the pressure ulcer candidate and inputs embedded data 65d to render a sampled feature vector 66h representative of the predictive features of embedded data 65h, which is communicated to a convolutional neural network 68a.

[00101] A nonlimiting exemplary neural architecture of embedded neural network

6lh is a one-stage convolutional network (e.g., an inception architecture), which may provide a confidence level with the pressure ulcer risk assessment 66h.

[00102] In practice, the neural architectures of sampled neural network 6le, sampled neural network 6 lf, encoded neural network 6lg and/or embedded neural network 6lh may include an attention module as known in the art of the present disclosure.

[00103] Convolutional neural network 68a is a module having a neural architecture trained for combining feature vectors 66e-66h to render pressure ulcer prediction 62a (FIG. 1) or pressure ulcer classification 62b (FIG. 1).

[00104] A non-limiting neural architectures of convolutional neural network 68a is a sigmoid-based convolutional neural network (e.g., multilayer perceptrons), which may provide a confidence level with pressure ulcer prediction 62a (FIG. 1) or pressure ulcer classification 62b (FIG. 1).

[00105] Referring to FIG. 4C, a third embodiment of a clinical decision support engine of the present disclosure employs data preprocessors 64e-64h (which may be integrated as one data preprocessor), a sampled support vector machine 6 li, a sampled support vector machine 6 lj, an encoded support vector machine 6lk, an embedded support vector machine 611 and a logic circuit 67b.

[00106] In operation, data preprocessor 64e is a module having an architecture for sampling imaging data 23 to render sampled data 65e. Imaging data 23 may be sampled as a time series of images generated by the camera(s) or in a time series of differential data between sequential image generated by the camera(s). [00107] Data preprocessor 64f is a module having an architecture for sampling physiological data 31 to render sampled data 65f . Physiological data 31 may be sampled as a time series of data point(s) generated by the physiological sensor(s) or in a time series of differential data between sequential data points generated by the physiological sensor(s).

[00108] Data preprocessor 64g is a module having an architecture for encoding candidate data 4la related to demographics and diagnostic/clinical observations of the pressure ulcer candidate to render encoded data 65g (e.g., one -hot coded binary coded or autoencoding).

[00109] Data preprocessor 64h is a module having an architecture for embedding candidate data 4la related to (e.g., word embedding) laboratory/clinical testing of the pressure ulcer candidate to render embedded data 65h.

[00110] Sampled support vector machine 6 li is a module having an optimal hyperplane trained for discriminately classifying sampled data 65e as an imagery detection or an imagery non-detection of a pressure ulcer and inputs sampled data 65e to render a pressure ulcer risk assessment 66i (e.g., a pressure ulcer prediction or a pressure ulcer classification based on imaging data 23), which is communicated to a logic circuit 67b and display/communication managers 70/80.

[00111] Sampled support vector machine 6 lj is a module having an optimal hyperplane trained for discriminating classifying sampled data 65f as an sensory detection or an sensory non-detection of a pressure ulcer and inputs sampled data 65f to render a pressure ulcer risk assessment 66j (e.g., a pressure ulcer prediction or a pressure ulcer classification based on physiological data 31), which is communicated to a logic circuit 67b and display/communication managers 70/80.

[00112] Encoded support vector machine 6lk is a module having an optimal hyperplane trained for discriminating classifying encoded data 65g as demographics and diagnostic/clinical observations of the pressure ulcer candidate indicating a high risk or a low risk of a pressure ulcer and inputs encoded data 65g to render a pressure ulcer risk assessment 66k (e.g., a pressure ulcer prediction or a pressure ulcer classification based on candidate data 4 la), which is communicated to a logic circuit 67b and

display/communication managers 70/80. [00113] Embedded support vector machine 611 is a module having an optimal hyperplane trained for discriminating classifying embedded data 65i as laboratory/clinical testing of the pressure ulcer candidate indicating a high risk or a low risk of a pressure ulcer and inputs embedded data 65h to render a pressure ulcer risk assessment 661 (e.g., a pressure ulcer prediction or a pressure ulcer classification based on candidate data 4lb), which is communicated to a logic circuit 67b and display/communication managers 70/80.

[00114] Logic circuit 67b is a module implementing a circuit for logically producing pressure ulcer prediction 62a (FIG. 1) or pressure ulcer classification 62b (FIG. 1) from the pressure ulcer assessments 66i-66l. In a risk score embodiment, logic circuit 67b inputs press pressure ulcer assessments 66i-66l into an average logic circuit. In a risk classification embodiment, logic circuit 67b inputs pressure ulcer assessments 66i-66l into an AND gate, which serves to select an output of a multiplexor between a high risk of pressure ulcer input and a low risk of pressure ulcer input.

[00115] Furthermore, each pressure ulcer assessments 66i-66l may individually serve to initiate a monitoring alert 63 (FIG. 1) when predicting a pressure ulcer of the corresponding input data or classifying the corresponding data as representing a pressure ulcer. For example, while pressure ulcer assessments 66i-66l collectively are indicating a low risk of a pressure ulcer of the pressure ulcer candidate, pressure ulcer assessments 66i may be indicating a high risk of a pressure ulcer of the pressure ulcer candidate whereby pressure ulcer assessment 66i is used to initiate the monitoring alert 63.

[00116] Referring to FIG. 4D, a fourth embodiment of a clinical decision support engine of the present disclosure employs data preprocessors 64e-64h (which may be integrated as one data preprocessor), a sampled support vector machine 6lm, a sampled support vector machine 6ln, an encoded support vector machine 6lo, an embedded support vector machine 6lp and a convolutional neural network 68b.

[00117] In operation, data preprocessor 64e is a module having an architecture for sampling imaging data 23 to render sampled data 65e. Imaging data 23 may be sampled as a time series of images generated by the camera(s) or in a time series of differential data between sequential image generated by the camera(s).

[00118] Data preprocessor 64f is a module having an architecture for sampling physiological data 31 to render sampled data 65f . Physiological data 31 may be sampled as a time series of data point(s) generated by the physiological sensor(s) or in a time series of differential data between sequential data points generated by the physiological sensor(s).

[00119] Data preprocessor 64g is a module having an architecture for encoding candidate data 4la related to demographics and diagnostic/clinical observations of the pressure ulcer candidate to render encoded data 65g (e.g., one -hot coded binary coded or autoencoding).

[00120] Data preprocessor 64h is a module having an architecture for embedding candidate data 4la related to (e.g., word embedding) laboratory/clinical testing of the pressure ulcer candidate to render embedded data 65h.

[00121] Sampled support vector machine 6lm is a module having an optimal hyperplane trained for discriminately classifying sampled data 65e as an imagery detection or an imagery non-detection of a pressure ulcer and inputs sampled data 65e to render a pressure ulcer risk assessment 66m (e.g., a pressure ulcer prediction or a pressure ulcer classification based on imaging data 23), which is communicated to a convolutional neural network 68b.

[00122] Sampled support vector machine 6ln is a module having an optimal hyperplane trained for discriminating classifying sampled data 65f as an sensory detection or an sensory non-detection of a pressure ulcer and inputs sampled data 65f to render a pressure ulcer risk assessment 66n (e.g., a pressure ulcer prediction or a pressure ulcer classification based on physiological data 31), which is communicated to convolutional neural network 68b.

[00123] Encoded support vector machine 6lo is a module having an optimal hyperplane trained for discriminating classifying encoded data 65g as demographics and diagnostic/clinical observations of the pressure ulcer candidate indicating a high risk or a low risk of a pressure ulcer and inputs encoded data 65g to render a pressure ulcer risk assessment 66o (e.g., a pressure ulcer prediction or a pressure ulcer classification based on candidate data 4la), which is communicated to convolutional neural network 68b and display/communication managers 70/80.

[00124] Embedded support vector machine 6lp is a module having an optimal hyperplane trained for discriminating classifying embedded data 65i as laboratory/clinical testing of the pressure ulcer candidate indicating a high risk or a low risk of a pressure ulcer and inputs embedded data 65h to render a pressure ulcer risk assessment 66p (e.g., a pressure ulcer prediction or a pressure ulcer classification based on candidate data 4lb), which is communicated to convolutional neural network 68b.

[00125] Convolutional neural network 68b is a module having a neural architecture trained for combining data 65e-65h and pressure ulcer risk assessments 66m-66p to render pressure ulcer prediction 62a (FIG. 1) or pressure ulcer classification 62b (FIG. 1).

[00126] A non-limiting neural architectures of convolutional neural network 68b is a sigmoid-based convolutional neural network (e.g., multilayer perceptrons), which may provide a confidence level with pressure ulcer prediction 62a (FIG. 1) or pressure ulcer classification 62b (FIG. 1).

[00127] Referring to FIGS. 1-4, those having ordinary skill in the art will appreciate the many benefits of the inventions of the present disclosure including, but not limited to, an unburdensome, meticulous, continuous monitoring of the risk of pressure ulcers.

[00128] Furthermore, it will be apparent that various information described as stored in the storage may be additionally or alternatively stored in the memory. In this respect, the memory may also be considered to constitute a“storage device” and the storage may be considered a“memory.” Various other arrangements will be apparent. Further, the memory and storage may both be considered to be“non-transitory machine- readable media.” As used herein, the term“non-transitory” will be understood to exclude transitory signals but to include all forms of storage, including both volatile and non volatile memories.

[00129] While the device is shown as including one of each described component, the various components may be duplicated in various embodiments. For example, the processor may include multiple microprocessors that are configured to independently execute the methods described in the present disclosure or are configured to perform steps or subroutines of the methods described in the present disclosure such that the multiple processors cooperate to achieve the functionality described in the present disclosure. Further, where the device is implemented in a cloud computing system, the various hardware components may belong to separate physical systems. For example, the processor may include a first processor in a first server and a second processor in a second server. [00130] It should be apparent from the foregoing description that various example embodiments of the invention may be implemented in hardware or firmware.

Furthermore, various exemplary embodiments may be implemented as instructions stored on a machine-readable storage medium, which may be read and executed by at least one processor to perform the operations described in detail herein. A machine-readable storage medium may include any mechanism for storing information in a form readable by a machine, such as a personal or laptop computer, a server, or other computing device. Thus, a machine-readable storage medium may include read-only memory (ROM), random-access memory (RAM), magnetic disk storage media, optical storage media, flash-memory devices, and similar storage media.

[00131] It should be appreciated by those skilled in the art that any block diagrams herein represent conceptual views of illustrative circuitry embodying the principles of the invention. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in machine readable media and so executed by a computer or processor, whether or not such computer or processor is explicitly shown.

[00132] Although the various exemplary embodiments have been described in detail with particular reference to certain exemplary aspects thereof, it should be understood that the invention is capable of other embodiments and its details are capable of modifications in various obvious respects. As is readily apparent to those skilled in the art, variations and modifications can be affected while remaining within the spirit and scope of the invention. Accordingly, the foregoing disclosure, description, and figures are for illustrative purposes only and do not in any way limit the invention, which is defined only by the claims.