Apparatus and method for monitoring peripheral diabetic neuropathy and / or peripheral arterial disease

Abstract

The plantar surface of the patient is in contact with the receiving area of the device body, causing a set of temperature sensors to produce a current set of temperature values. Four or more earlier sets of temperature values produced earlier in time for the plantar surface of the patient are accessed. Next, after setting a normalization reference for the earlier sets of temperature values and the current set of temperature values to produce normalized data, the normalized data is translated into model information representative of PDN or PAD progression. The model information is used to determine a trajectory of the PDN or PAD of the patient.

Description

[0001] priority

[0002] This patent application claims priority to U.S. Provisional Patent Application No. 62 / 958,858, filed January 9, 2020, entitled “Apparatus and Method for Identifying and Monitoring Progression of Peripheral Diabetic Neuropathy and / or Peripheral Artery Disease”, and designates Brian Petersen, Katherine Wood, David Linders, and Min Zhou as inventors, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0003] The illustrative embodiments of the present invention generally relate to peripheral diabetic neuropathy and / or peripheral artery disease, and more specifically, various embodiments of the present invention relate to monitoring peripheral diabetic neuropathy and / or peripheral artery disease. Background Technology

[0004] Poorly managed diabetes can lead to serious health complications affecting the body's limbs. Two such complications are diabetic peripheral neuropathy and diabetic peripheral artery disease. Peripheral neuropathy causes loss of protective sensation in the limbs, thus increasing the likelihood of wounds in the feet, ankles, and legs. These wounds can become infected and chronic, potentially leading to gangrene, cellulitis, amputation of the foot or leg, and in some cases, death. Peripheral artery disease is characterized by impaired circulatory pathways and reduced blood flow to the limbs. Peripheral artery disease hinders wound healing and can lead to gangrene, claudication, ischemia, amputation, and death.

[0005] Therefore, healthcare professionals recommend routine evaluations for diabetic patients to determine the presence of peripheral artery disease and peripheral neuropathy, as well as other complications. Diagnostic tests identify these complications and are typically performed during outpatient clinical examinations. However, peripheral neuropathy and peripheral artery disease are both progressive conditions, and many diabetic patients do not undergo the recommended frequency of evaluation for these complications. Summary of the Invention

[0006] According to one embodiment of the invention, a method and / or device monitors peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in a patient with a foot having a sole. For this purpose, the method provides a body having a base with a top surface having a receiving area configured to receive the sole of the foot. The base forms an open platform or a closed platform. Furthermore, the body has a set of temperature sensors in communication with the top surface of the receiving area. This set of temperature sensors is preferably located within the receiving area and is configured to activate upon receiving stimulation applied to one or both of the open or closed platform and the set of temperature sensors. The set of temperature sensors is also configured to be in thermal communication with the sole of the foot within the receiving area to determine the current temperature at each of a set of different locations on the sole of the foot. Additionally, the set of temperature sensors is configured to generate a set of temperature values, where each location has a corresponding temperature value.

[0007] The method and / or device also brings the patient's sole surface into contact with the receiving area, causing the temperature sensor array to generate a current set of temperature values ​​and accessing four, five, or more earlier sets of temperature values ​​generated at a previous time on the patient's sole surface. Next, after setting standardized references for each of the earlier and current temperature value sets to generate standardized data, the method / device transforms the standardized data into model information representing the progression of the PDN or PAD. Using this model information, the method / device determines the trajectory of the patient's PDN or PAD.

[0008] Each of the earlier temperature value groups is separated from the other earlier temperature data value groups by at least one day in time. Furthermore, the method / device can set a normalization reference by applying a normalization function to the earlier temperature value groups and the current temperature value group using a normalization reference. Among other things, the normalization reference may include contralateral temperature values ​​(from the other foot), ipsilateral temperature values ​​(e.g., from the same foot), and / or ambient temperature.

[0009] Those skilled in the art can access significantly more than four earlier groups. For example, the method / apparatus can access from four to 10,000 earlier groups (or more, such as 20,000). The method / apparatus can perform the conversion in any of a variety of ways. For example, some embodiments perform the conversion by selecting a model that characterizes the earlier temperature value groups and the current temperature value groups as a simpler system, and then applying that model to multiple earlier temperature value groups and the current temperature value groups to generate model information.

[0010] After determining the trajectory, some implementations use model information to predict the future state of a patient's PDN or PAD relative to the current temperature value group. Among other things, the trajectory may include one or both of the rate of change of model information and the magnitude of model information. Furthermore, as mentioned above, some implementations only convert and determine the PDN, or only the PAD.

[0011] Earlier temperature value sets can be generated from any number of sources. For example, they can be generated using a set of temperature sensors, from some other sources, or from a set of temperature sources and other sources together.

[0012] In another embodiment, the monitor manages peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in a patient with a foot having a sole. For this purpose, the monitor has a body having a base with a top surface. The top surface has a receiving area configured to receive the sole of the foot, and the base forms an open or closed platform. The body has a set of temperature sensors in communication with the top surface of the receiving area. The set of temperature sensors is spaced apart within the receiving area and configured to activate upon receiving stimulation applied to one or both of the open or closed platform and the set of temperature sensors. Furthermore, the set of temperature sensors is configured to be in thermal communication with the sole of the foot within the receiving area to determine the current temperature at each of a set of spaced-apart locations on the sole of the foot. The set of temperature sensors is also configured to generate a set of temperature values, where each location has an associated temperature value, and to generate the current set of temperature values ​​upon contact between the receiving area and the patient's sole surface.

[0013] The monitor also includes an input device configured to receive four or more earlier sets of temperature values ​​of the patient's plantar surface generated at an earlier time, and a normalizer configured to set a normalization reference between the earlier and current temperature sets to generate normalized data. The monitor also has a modeler operatively connected to the normalizer. This modeler is configured to transform the normalized data into model information representing the progression of PDN or PAD, to be used to determine the trajectory of the patient's PDN or PAD.

[0014] The illustrative embodiments of the present invention are implemented in the form of a system and / or a computer program product having a computer-usable medium on which computer-readable program code is stored. This computer-readable code can be read and used by a computer system according to conventional procedures.

[0015] In other embodiments, a non-contact method and / or device monitors peripheral diabetic neuropathy ("PDN") and / or peripheral arterial disease ("PAD") in a patient with the sole of their foot. To this end, the method / device provides a thermal imager with an infrared radiation sensor and directs the infrared radiation sensor toward the patient's sole surface to generate a current set of thermal images, which is considered to represent the current temperature of the entire sole surface. The method / device accesses four or more earlier sets of thermal images of the patient's sole surface generated at an earlier time and sets a normalized reference for the earlier and current sets of thermal images to generate normalized data. Next, the method / device transforms the normalized data into model information representing the progression of PDN or PAD; the model information is then used to determine the trajectory of the patient's PDN or PAD. Attached Figure Description

[0016] Those skilled in the art will more fully appreciate the advantages of various embodiments of the invention from the following "Detailed Description" discussed with reference to the accompanying drawings.

[0017] Figure 1 An instance of a foot with a PAD and a PDN is shown schematically.

[0018] Figure 2A The illustration schematically shows one use and form factor that can be implemented according to an illustrative embodiment of the invention.

[0019] Figure 2B An open platform that can be configured according to an illustrative embodiment of the invention is schematically shown. The figure also illustrates its use in the case of a single-leg amputee.

[0020] Figure 3A A cross-sectional view of an open platform that can be configured according to an illustrative embodiment of the present invention is shown schematically.

[0021] Figure 3B A schematic close-up view of the platform's details regarding the pad and temperature sensor in the foot receiving area, shown in an illustrative embodiment.

[0022] Figure 4 A network for implementing illustrative embodiments of the present invention is shown schematically.

[0023] Figure 5 A schematic diagram of various components of an illustrative embodiment of the present invention is shown.

[0024] Figure 6A Details of the data processing module according to an illustrative embodiment of the present invention are shown schematically.

[0025] Figure 6B Details of the additional functions of the data processing module according to an illustrative embodiment of the present invention are shown schematically.

[0026] Figure 7 The process of monitoring a patient's foot according to an illustrative embodiment of the present invention is shown.

[0027] Figure 8A The following is a graphical representation of an example of the foot temperature and a model of these temperatures in one embodiment of the invention.

[0028] Figure 8B An example of temperature values ​​standardized by environmental standardization reference and a model of these standardized temperature values ​​is illustrated in the figure according to one embodiment of the present invention.

[0029] Figure 8C An example of a model of temperature values ​​standardized with a contralateral standardized reference and these standardized temperature values ​​is illustrated in the figure according to an embodiment of the present invention.

[0030] Figure 8D An example of a model of temperature values ​​standardized using a same-side standardized reference and these standardized temperature values ​​is illustrated in the figure according to an embodiment of the present invention. Detailed Implementation

[0031] In an illustrative embodiment, a method (and / or device) effectively monitors the progression of one or both peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD"). Understanding this progression enables the patient to take earlier action when necessary to improve treatment outcomes. To this end, the method (and / or device) collects several earlier sets of temperature values ​​of the patient's foot (e.g., temperature values ​​for each day of the past 4 days), as well as current sets of temperature values ​​of the same patient's foot (e.g., most recently acquired sets of temperature values). All sets of temperature data preferably relate to the same location on the patient's foot.

[0032] The method (and / or device) then standardizes the temperature value groups and transforms this standardized data into model information for identifying health trends. Therefore, using this model information, the illustrative implementation can determine the trajectory of a patient's PDN or PAD, effectively enabling earlier medical intervention than is permitted by prior art monitoring techniques known to the inventors. Details of the illustrative implementation will be discussed below.

[0033] Peripheral arterial disease (PAD) is a complication characterized by impaired circulatory pathways and reduced blood flow. PAD cases can range from moderate to severe; in many cases, PAD leads to limited physical capabilities and significantly reduces quality of life. Furthermore, infection and impaired wound healing are far more common due to poor macrocirculation and microcirculation. PAD often progresses to severe limb ischemia (CLI), a serious vascular complication, often accompanied by gangrene, frequently requiring amputation, and associated with high mortality. PAD is typically progressive, affecting the most distal parts of the limbs (e.g., toes) and then the more proximal parts (e.g., heels). Other complications associated with vascular disease of the extremities include thrombosis, embolism, claudication, and arteriosclerosis.

[0034] As PAD progresses, the arteries supplying blood to the limbs are damaged due to narrowing, partial occlusion, or complete blockage. Consequently, oxygenated blood flow to the feet decreases over time. The progression of this disease is associated with the worsening of these conditions. Without timely treatment, this leads to insufficient oxygen supply to tissues, preventing the healing of diseased or damaged tissues and ultimately causing cell death. In many cases, timely clinical intervention is required to prevent further tissue damage, infection, or amputation. This progression can occur slowly over months or years, as in cases of arteriosclerosis, or rapidly over weeks or days, as in cases of thrombosis or embolism. If the damaged vessel is large and proximal, the resulting changes in blood flow may affect the entire limb. Alternatively, if the damaged vessel is small and distal, blood flow may be reduced to only a portion of the limb.

[0035] This leads to changes in the health of the blood vessels flowing to the limbs or limb parts, effectively reducing the volumetric flow of warm, oxygenated blood reaching the distal tissues supplied by the affected arteries. Consequently, these distal tissues tend to have lower oxygen levels and are less thermoregulated than healthy tissues. In cold environments, this may manifest as colder extremities than normal. Or, in hot environments, it may manifest as warmer extremities than normal. In extreme cases, this impaired thermoregulation can cause tissue overheating, breakdown, and acute injuries such as foot ulcers.

[0036] Traditionally, the diagnosis of PAD (Patient Ankle-Brachial Pressure Index) has been based on the ankle-brachial pressure index (ABPI), determined by the ratio of blood pressure measured at the ankle to that measured in the upper arm. However, ABPI can be unreliable due to arterial calcification and sclerosis, a lack of standardized measurement and calculation methods, patient position during measurement, cuff placement at the ankle, bilateral or contralateral measurement, and other influencing factors. It is also impossible in certain situations, such as with trauma or unhealed wounds. If arterial calcification is present, pressure can be measured in the toes, but this is not feasible for patients with a history of toe amputation. Other diagnostic methods include Doppler ultrasound and fluoroscopy, as well as laser Doppler flowmetry; both of these methods are complex, expensive, and unstable. MRI and angiography can also be used, but they produce static images of blood flow.

[0037] Furthermore, some vascular interventions aimed at permanently improving limb blood flow, such as angioplasty and stenting, sometimes fail or worsen over time. Therefore, limb vascular flow may improve briefly and then decline over time. During this period following intervention, but prior to the next clinical evaluation, there is currently no known technology or device to assess limb vascular health to determine if the intervention has been sustained. If the intervention is not maintained, this could lead to modifications or additional vascular surgery, resulting in greater costs and morbidity.

[0038] Peripheral neuropathy ("PN", also known as "peripheral diabetic neuropathy" or "PDN") is a complication of diabetes that causes damage to sensory nerve function, typically involving impaired nerve pathways in the hands and feet. Nerve damage affects the direct nerve contacts from the limbs through the spinal cord to the brain. PDN is often caused by diabetes, where poor blood sugar control leads to high levels of sugar and fat in the blood, causing neuronal degeneration. This complication can lead to diabetic foot ulcers and other foot complications due to severe loss of sensation in the foot. Patients also report extreme heat or coldness, tingling, or pain in the foot as symptoms of PDN. Patients with PDN generally have a higher risk of developing the condition. Like PAD, PDN is usually progressive, first affecting the most distal parts of the limbs, such as the toes, and then the more proximal parts, such as the heel.

[0039] With the development of vasomotor nerve distension (PDN), the autonomic nervous system mediating vasodilation and vasoconstriction in the limbs becomes dysfunctional. Overactivity of the sympathetic nervous system, responsible for mediating blood vessels, can ultimately lead to vascular damage. Other neurological disorders can reduce the system's ability to dilate or constrict blood vessels in response to bodily needs, such as changes in ambient temperature. Therefore, limbs affected by PDN will not regulate their temperature as effectively as healthy limbs. In other cases, limbs with PDN may not effectively regulate inflammation, leading to widespread inflammation throughout the limb. Abnormal temperatures can be observed in both cases, both at rest and in response to changes in ambient temperature.

[0040] Furthermore, PDN reduces the function of motor neurons, which may lead to decreased foot muscle tone, resulting in deformities and gait problems. If left uncorrected, these can create high-pressure points in the foot, potentially causing further damage to underlying tissues.

[0041] PDN typically progresses slowly, taking months or years, making accurate diagnosis or monitoring difficult. Regular assessment of changes in neuropathy is essential to help patients and healthcare providers implement footwear and lifestyle changes to prevent foot injuries.

[0042] Several methods are currently used to diagnose PDN, most of which involve the patient's ability to sense stimuli to varying degrees. One method uses 10 grams of Semim' Weinstein monofilaments to determine the loss of protective sensation by applying blunt stimulation to different locations on the patient's foot. Each monofilament requires a certain amount of force to bend, which the patient cannot feel if PDN is present. Some medical institutions use biovibration threshold meters to quantitatively measure a patient's vibration perception threshold. Similarly, doctors sometimes use a 128Hz vibrating tuning fork for PDN assessment. The ability to distinguish between warm and cold temperatures of the sole of the foot is also used to identify PDN.

[0043] Diagnosing PDN and PAD can be challenging because they are often asymptomatic, and many diabetic patients do not receive adequate and routine screening for these complications. In many cases, the devices currently known to the inventors are generally insufficient for adequate diagnosis. However, the illustrative embodiments have been modified technically from the basic device to provide better screening and monitoring of PDN and PAD.

[0044] Furthermore, there are no standardized tests for PDN or PAD assessment, and there is significant variation among the different diagnostic methods used by healthcare professionals. Finally, existing diagnostic methods are typically performed by healthcare professionals during clinical examinations and are designed to diagnose complications, rather than predict them or monitor their progression and severity.

[0045] Therefore, patients diagnosed with these complications still face the risk of neurological and vascular changes that can lead to adverse and costly outcomes such as foot ulcers, gangrene, critical limb ischemia, and amputation.

[0046] Therefore, these patients subsequently suffer from a significant problem that the prior art, as known to the inventors, cannot address: they cannot rely on conventional devices and methods to identify common complications of diabetes, such as PDN and PAD, to monitor disease progression. This deficiency, among other reasons, lies in the design of existing devices and technologies, which are configured for clinic use and administered by trained medical professionals. The illustrative implementation addresses these problems by producing devices and applying techniques for analyzing lower limb temperature to predict, identify, and monitor the progression of PDN and PAD.

[0047] Specifically, the illustrative implementation analyzes a patient's foot to predict or determine the presence and progression of a PDN or PAD. This allows for earlier intervention for patients, their healthcare providers, and / or their caregivers, reducing the risk of more serious complications. For this purpose, a temperature detection modality (e.g., an open or closed platform measuring surface temperature) receives the patient's foot and generates temperature data, which is processed to determine whether a PDN or PAD has occurred, and / or the progression of a previously diagnosed PAD or PDN. This modality can use any of a variety of different methods, such as comparing one or more parts of the foot or leg with some other prescribed value, such as ambient / circumferential temperature or the temperature of another part of the body.

[0048] Using this comparison, if the pattern determines that the limb exhibits at least one of several prescribed patterns, then the output information generated by various embodiments indicates whether or not a PDN or PAD has appeared, and / or the movement trajectory of the PDN or PAD. This output information can also indicate whether a known PDN or PDN has been developed. Details of the illustrative embodiments will be discussed below.

[0049] To analyze one or more limbs, illustrative implementations may use patterns and techniques similar to those discussed in US 9,271,672, the disclosure of which is incorporated herein by reference in its entirety. For example, Figure 1 The patient's foot is schematically shown 10, adversely affected by diabetic peripheral neuropathy (PDN) 12 and peripheral artery disease (PAD) 14. These complications may have associated sequelae such as thrombosis, foot ulcers, ischemia, and many other problems known to those skilled in the art.

[0050] Figure 2A and 2BThe diagram schematically illustrates one form of factor in which a patient / user steps on an open platform 16 that collects data about the user's foot (or foot 10). In this particular example, the open platform 16 has a mat-like body placed in areas where the patient frequently stands, such as in front of a bathroom sink, beside a bed, in front of a shower, on a footrest, or integrated into a mattress. As an open platform 16, the patient initiates the process simply by stepping onto the receiving area of ​​the sensing surface on top of the platform 16 (e.g., where a prosthesis would normally be, or where it is supported by some object). Therefore, this and other forms of factors often do not require the patient to explicitly decide to interact with the platform 16. Instead, many anticipated forms of factors are configured for use in areas where the patient frequently stands throughout the day without foot coverage. Additionally, the open platform 16 can be moved to directly contact the foot 10 of a patient who cannot stand. For example, if the patient is bedridden, the platform 16 can contact the patient's foot 10 while in bed.

[0051] Bathroom mats or rugs are just two of a variety of potential form factors. Others may include something like a scale, stand, foot pedal, console, a platform built into the floor tile 16, or a more portable mechanism that can receive at least one foot 10. Figure 2A and 2B The illustrated implementation has a top surface area larger than the surface area of ​​one or both of the patient's feet 10. This allows caregivers to obtain a complete view of the patient's entire sole, providing a more complete view of the foot 10.

[0052] The open platform 16 may also have indicators or monitors 18 on its top surface, which can have any of many functions. For example, indicators can change color or emit an alarm sound after a reading is completed, indicate the progress of a process, or display the result of a process. Of course, indicators or monitors 18 can be located anywhere other than the top surface of the open platform 16, such as on the side, or as separate components communicating with the open platform 16. In fact, in addition to using visual or auditory indicators, or in lieu of visual or auditory indicators, the platform 16 can have other types of indicators, such as tactile indicators / feedback, and our thermal indicators.

[0053] In addition to using the open platform 16, alternative implementations can be carried out using a closed platform 16, such as shoes, insoles, soles, or socks that the patient wears regularly, or as needed. For example, the insoles of the patient's shoes or boots can have the function of detecting the presence of PADs or PDNs or predicting their occurrence, and / or monitoring the progression of PADs or PDNs. Some implementations may also have the ability to monitor the presence and / or occurrence of ulcers and / or pre-ulcer stages.

[0054] To monitor for complications in the patient's foot (discussed in more detail below), Figure 2A and 2B The platform 16 collects temperature data from multiple different locations on the sole of the foot 10. This temperature data provides core information ultimately used to determine the health status of the foot 10. Figure 3A A cross-sectional view of an open platform 16 configured and arranged according to one embodiment of the present invention is shown schematically. Of course, this embodiment is only one of many potential embodiments, and like other features, it is discussed only by way of example.

[0055] As shown in the figure, the platform 16 is formed as a stack of functional layers sandwiched between the cover plate 20 and the rigid base 22. For safety reasons, the base preferably has a rubberized bottom surface or other anti-slip features. Figure 3A One embodiment of this anti-slip feature is shown, namely the anti-slip base 24. The platform 16 preferably has a relatively thin profile to avoid tripping the patient and to make it easy to use.

[0056] To measure foot temperature, platform 16 has an array or matrix of temperature sensors 26 fixed directly below cover 20. These temperature sensors are preferably located in a receiving area on the top surface of platform 16—that is, the top surface or area of ​​platform 16—to receive foot 10. Preferably, the temperature sensors 26 are mounted on a relatively large printed circuit board 28 and communicate directly with the receiving area.

[0057] The sensors 26 are preferably arranged on the printed circuit board 28 in a two-dimensional array / matrix of fixed contact sensors. Their spacing or distance is preferably relatively small, allowing for a greater number of temperature sensors 26 on the array. Among other things, the temperature sensors 26 may include temperature-sensing resistors (e.g., printed or discrete elements mounted on the circuit board 28), thermocouples, fiber optic temperature sensors, or thermochromic films. Therefore, when used with temperature sensors 26 requiring direct contact, the illustrative embodiment forms the cover plate 20 with a thin material having relatively high thermal conductivity. The platform 16 may also use temperature sensors 26 that can still detect temperature through the patient's socks.

[0058] Other implementations may use non-contact temperature sensors 26, such as infrared detectors. In fact, in this case, the cover plate 20 may have openings to provide a line of sight from the sensor 26 to the sole of the foot 10. Therefore, the discussion of contact sensors is merely illustrative and is not intended to limit the various implementations. As discussed and noted in more detail below, regardless of their specific type, multiple sensors 26 generate multiple corresponding temperature data values ​​for multiple portions / points on the patient's foot 10 to monitor the health of the foot 10.

[0059] Some implementations may also use pressure sensors to achieve various functions, such as determining the orientation of foot 10 and / or automatically initiating the measurement process. Among other things, pressure sensors may include piezoelectric, resistive, capacitive, or fiber optic pressure sensors. This layer of platform 16 may also have other sensor modes besides temperature sensor 26 and pressure sensor, such as positioning sensors, GPS sensors, accelerometers, gyroscopes, and other sensors known to those skilled in the art.

[0060] An illustrative embodiment of thermal analysis of the foot 10 can obtain temperature input values ​​from various sensor types, including thermal imagers and open or closed platforms 16 with contact or non-contact temperature sensors. Some such platforms 16 may include shoes, insoles, bandages, and wrappings. Some embodiments may allow for manual point temperature measurements. Temperature sensors may include infrared photodiodes, phototransistors, resistive temperature detectors, thermistors, thermocouples, fiber optics, and thermochromic sensors. Those skilled in the art will understand that these temperature sensing methods and sensor types are examples of alternatives, and some or all of the analytical methods described below are not dependent on the sensor methods used in the system.

[0061] To reduce the time required to sense the temperature at a specific point, the illustrative embodiment places an array of thermal pads 30 above an array of temperature sensors 26. To illustrate this, Figure 3B A small portion of the temperature sensor array 26 is schematically shown, depicting four temperature sensors 26 and their pads 30. The temperature sensors 26 are depicted as phantoms because they are preferably covered by the pads 30. However, some embodiments do not cover the sensors 26, but simply thermally connect the sensors 26 to the pads 30.

[0062] Therefore, each temperature sensor 26 in this embodiment has an associated thermal pad 30 that directs heat directly from a two-dimensional portion of the foot 10 (although the foot may have some depth dimension, it is considered a two-dimensional region) to its exposed surface. The array of thermal pads 30 preferably occupies the majority of the total surface area of ​​the printed circuit board 28. The distance between the pads 30 thermally isolates them from each other, thereby eliminating thermal short circuits.

[0063] For example, each pad 30 can be square in shape, with each side approximately 0.1 to 1.0 inches long. Therefore, the spacing between the pads 30 is less than this amount. Thus, as a further detailed example, some embodiments may use 0.25-inch (per side) square pads 30 to space the temperature sensors 26 approximately 0.4 inches apart, with the square pads 30 oriented such that each sensor 26 is centered within the square pad 30. This leaves an open area (i.e., spacing) of approximately 0.15 inches between the square pads 30. Among other things, the pads 30 can be formed from a film of a thermally conductive metal, such as copper.

[0064] Alternative implementations do not require pad 30.

[0065] As described above, some embodiments do not use an array of temperature sensors 26. Instead, such embodiments may use a single temperature sensor 26, which can obtain most or all of the temperature readings of the sole of the foot. For example, a monolithic thermosensitive material, such as a thermochromic film (as described above), or a similar device should be sufficient. As is known to those in the art, thermochromic films based on liquid crystal technology have liquid crystals inside that reorient themselves with changes in temperature (typically above ambient temperature), resulting in a noticeable color change. Alternatively, one or more individual temperature sensors 26, such as thermocouples or temperature sensor resistors, may be moved to repeatedly read the temperature on the bottom of the foot 10.

[0066] For effective operation, the open platform 16 should be configured so that its top surface substantially contacts the entire sole 10 of the patient's foot within the receiving area. To this end, the platform 16 has a flexible and movable layer 32 of foam or other material that adapts to the user's foot 10. For example, this layer should conform to the arch of the foot 10. Naturally, the sensor 26, the printed circuit board 28, and the cover plate 20 should also be flexible yet robust, conforming to the foot 10 accordingly. Therefore, the printed circuit board 28 is preferably formed primarily of a flexible material supporting the circuitry. For example, the printed circuit board 28 may be formed primarily of flexible circuitry supporting the temperature sensor 26, or it may be formed of a strip of material that bends separately when the foot is received. Alternative implementations may lack such flexibility (e.g., formed of conventional printed circuit board materials such as FR-4), and therefore may produce less effective data.

[0067] The (integral) rigid base 22, located between foam 32 and anti-slip base 24, provides rigidity to the overall structure. Furthermore, the outline of the rigid base 22 can accommodate the motherboard 34, battery pack 36, circuit housing 38, and other circuit elements that provide further functionality. For example, the motherboard 34 may contain integrated circuits and a microprocessor that control the functions of the platform 16.

[0068] In addition, the motherboard 34 may also have the user interface / indicator monitor 18 as described above, and the communication interface 40 ( Figure 5 The communication interface 40 can connect to a larger network 44, such as the Internet, via wireless or wired connection, enabling various data communication protocols, such as Ethernet. Alternatively, the communication interface 40 can communicate via embedded Bluetooth or other short-range wireless radio frequencies, communicating with cellular phone networks 44 (e.g., 3G or 4G networks).

[0069] Platform 16 may also have edges 42 and other surface features to improve its aesthetic appearance and patient comfort. The layers can be secured together using one or more adhesives, clips, nuts, bolts, or other fastening devices.

[0070] In another embodiment, the open platform 16 allows the foot to be placed at a distance from the thermal imager to capture a thermal image of the sole. The platform 16 may have an infrared-transparent or translucent window on which the foot rests. Alternatively, the platform 16 may have an infrared-opaque layer with holes, cuts, or other discontinuities through which the thermal imager can image the foot.

[0071] Some implementations may use a remote temperature sensing device, such as a thermal imager, instead of platform 16. In this case, as described below, such an implementation may point its thermal sensor at the bottom of the patient's foot 10. For example, the thermal imager may point its infrared radiation emitter or sensor at the surface of the patient's foot to generate a current set of thermal image data. This set of thermal image data can be considered similar to the set of calorific values ​​discussed above. Therefore, using a thermal imager in some implementations may be a useful non-contact method to obtain temperature information for monitoring the PDN or PAD.

[0072] While it collects temperature and other data about the patient's feet, the illustrative implementation allows additional logic for monitoring foot health to be placed elsewhere. For example, such additional logic could be on a remote computing device. For this and other purposes, Figure 4 The diagram schematically illustrates a method in which platform 16 can communicate with a larger data network 44 according to various embodiments of the invention. As shown, platform 16 can connect to the Internet via a local router, its local area network, or directly without intervention equipment. This larger data network 44 (e.g., the Internet) can include any of many different endpoints that are also interconnected. For example, platform 16 can communicate with an analytics engine 46 that analyzes thermal data from platform 16 and determines the health status of patient foot 10. Platform 16 can also communicate directly with healthcare providers 48, such as doctors, nurses, relatives, and / or organizations responsible for managing patient care. In fact, platform 16 can also communicate with patients, such as via text message, telephone, email, or other means permitted by the system.

[0073] Figure 5A block diagram of a foot monitoring system is schematically shown, illustrating communication between platform 16 (or other means, such as a thermal imager) and remote server 60. As shown, the patient communicates with platform 16 by standing on the receiving area of ​​the main body of platform 16, which activates sensor 26. Alternatively, foot 10 can be received in some way by an array of sensors 26 (e.g., an array of temperature sensors 26, or, if not on an open or closed platform 16, by a thermal imager), which is represented in this figure as "sensor matrix 52". (Example: ...) Figure 3A The motherboard 34 and the circuitry implemented therein control the acquisition of temperature and other data for storage in a data storage device 56. Among other things, the data storage device 56 may be a volatile or non-volatile storage medium, such as a hard disk, high-speed random access memory ("RAM"), or solid-state memory. Input / output interface ports 58, also controlled by the motherboard 34 and other electronic devices on the platform 16, selectively transfer or forward acquired data from the storage device to an analysis engine 46 on a remote computing device, such as the server 60 mentioned above. The data acquisition device 54 may also control a user indicator / monitor 18, which provides feedback to the user via the aforementioned indicators (e.g., auditory, visual, or tactile).

[0074] The analytics engine 46 on the remote server 60, together with the health data analytics module 62, analyzes the data received from the platform 16. The server output interface 64 forwards the processed output information / data from the analytics engine 46 and the health data analytics module 62 to others on the network 44, such as to providers, network monitors, or to users via telephone alerts, email alerts, text alerts, or other similar means.

[0075] The output information can be provided in a relatively raw form for further processing. Alternatively, the output information can be output in a high-level format to facilitate review by automated logic or the person viewing the data. Among other things, the output information can indicate the actual presence of a PAD or PDN, the risk of a PAD or PDN appearing, the trajectory of a PAD or PDN, or simply indicate that the patient is healthy and there is no risk of a PAD or PDN. Furthermore, the output information can also include information to help end-users or healthcare providers monitor the progress of existing PADs or PDNs and / or begin monitoring the progress of newly located PADs or PDNs.

[0076] The output of this analysis can be processed to generate risk summaries and scores, which can be displayed to different users to trigger alerts and suggest the need for intervention. Among other things, state estimation models can model potential changes in a user's condition to assess the likelihood of future concurrent PADs or PDNs. Furthermore, these models can be combined with predictive models, such as linear logistic regression models and support vector machines, which can integrate large amounts of diverse current and historical data, including important patterns discovered during offline analysis. This can be used to predict whether a user is likely to experience problems within a given timeframe. Predictions of likelihood can be processed into risk scores, which can also be displayed to users and other third parties. These scores and displays will be discussed in more detail below.

[0077] Use like Figure 5 The distributed processing arrangement shown offers numerous advantages. Among other things, it allows platform 16 to have relatively simple and inexpensive components that are inconspicuous to patients. Furthermore, this allows for a "Software as a Service" ("SaaS model") business model, which, among other things, allows for more flexible functionality, generally easier patient monitoring, and faster feature updates. Additionally, the SaaS model facilitates the accumulation of patient data to enhance analytical capabilities.

[0078] Some implementations may distribute and physically locate functional components in different ways. For example, platform 16 may have an analytics engine 46 on its local motherboard 34. In fact, some implementations provide functionality entirely on platform 16 and / or within other components locally near platform 16. For example, all these functional elements (e.g., analytics engine 46 and other functional elements) may be housed within a housing formed by cover 20 and rigid base 22. Therefore, the discussion of distributed platforms is just one of many implementations that can be adapted to a particular application or purpose.

[0079] Those skilled in the art can use any of many different hardware, software, firmware, or other unknown technologies to perform the functions of the analysis engine 46. Figure 6A Several functional blocks are shown, which, together with other functional blocks, can be configured to perform the functions of the analysis engine 46. This figure only shows these blocks and illustrates one way of implementing various implementations.

[0080] In short, Figure 6AThe analysis engine 46 implements a thermogram generator 66, configured to generate a thermogram of the patient's foot 10 based on multiple temperature readings from the bottom of the foot 10 (if the thermogram is to be used in the analysis); and a pattern recognition system 68, configured to determine whether a specific temperature reading from thermal imaging (if used) and / or from thermal sensor 26 presents any of several different prescribed patterns. Pattern data and other information may be stored in local memory 76. If the thermogram and / or multiple temperature readings present any of these prescribed patterns, then the foot 10 may be unhealthy in some way (e.g., with PAD or PDN) and / or have an undesirable trajectory with PAD or PDN.

[0081] Analysis engine 46 also has analyzer 70, which is configured to generate the aforementioned output information, indicating any of a number of different conditions of foot 10. For example, the output information may indicate the risk of a PAD or PDN, the existence of a previously undiagnosed PAD or PDN, or the progress of a known PAD or PDN. Communicating through some interconnection mechanism, such as bus 72 or a network connection, these modules cooperate to determine the state of foot 10, which can be transmitted or forwarded via input / output port 74, which communicates with the aforementioned parties on the larger data network 44.

[0082] Figure 6B The diagram schematically illustrates additional components that can be part of the analysis engine 46. These components can be integrated with... Figure 6A The components mentioned earlier are used together. Specifically, the analysis engine 46 also has a normalizer 77, configured to set a normalization reference for multiple groups of temperature data values, thereby producing normalized data. As discussed in detail below, in a preferred embodiment, the normalization reference is a way of normalizing temperature data values ​​relative to a common reference, such as ambient temperature (e.g., ambient temperature), or the temperature of the same location on another foot. The normalizer 77 is operatively connected to the modeler 79 via bus 72 (or other interconnection device). In a preferred embodiment, the modeler 79 is configured to transform the normalized data from the normalizer 77 into model information representing the progress of the PDN or PAD. This model information essentially describes the multiple normalized temperature data points (in this case, normalized temperature data values) as a simpler system. In other words, the normalized temperature data points are a more complex system, and the modeler 79 transforms this data into a simpler system that can be characterized more accurately and easily (as described below).

[0083] Figure 7 The illustration shows a process for monitoring a patient's foot 10 according to an illustrative embodiment of the invention. It should be noted that this process is a simplification of a typically longer process that might be used to monitor a patient's foot 10. Therefore, Figure 7The process has additional steps that might be used by those skilled in the art. Furthermore, some steps may be performed in a different order than shown, or simultaneously. Therefore, those skilled in the art can appropriately modify the process.

[0084] Figure 7 The process begins at step 700, which collects the current temperature value of the patient's foot 10. For example, if platform 16 is used, the user can step on the receiving area, thus contacting the bottom of their foot 10 and communicating with a set (one, two, or more) of temperature sensors 26. Platform 16 may automatically detect this contact (e.g., using pressure or other sensors), or require some input, such as a button, which, when activated, effectively causes the set of temperature sensors 26 to generate a set of temperature values. Since these temperature values ​​are recently or "currently" obtained, this set is referred to as the "current temperature value set," etc. Each temperature value in this set represents a discrete temperature measurement from a specific temperature sensor (or an area of ​​a single temperature sensor).

[0085] Simultaneously, before or after generating the current temperature value set, in step 702, input / output 74 receives multiple earlier temperature value sets generated at an earlier time. Each earlier set can be retrieved from memory 76 or other location and obtained (and stored) at an earlier time. For example, the earlier temperature value sets may include five separate sets, each generated by a pattern (e.g., platform 16) from the patient's foot over 10 consecutive days (e.g., each set is obtained one day apart).

[0086] Therefore, within a week, the first group might be produced on Monday, the second on Tuesday, the third on Wednesday, the fourth on Thursday, and the fifth on Friday. Other implementations can produce different groups at different intervals, such as more than one day (e.g., every two or three days), weekly, monthly, etc. The intervals can also be irregular, such as producing one group on Monday, the next on Tuesday, the third on Friday, the fourth on Saturday, and the fifth on Sunday. Clinicians can configure the system to the desired intervals as needed.

[0087] If used as described, each temperature value in each group (preferably also has a corresponding temperature value in other groups. For example, if sensor 26 has a single sensor that collects the temperature value of the bottom of the big toe, then each group has a temperature value for the big toe. The key difference between the groups is that each temperature value is time-separated from the temperature values ​​at other corresponding / similar locations.

[0088] The inventors have discovered that in many cases, as few as four earlier temperature value groups are sufficient to provide appropriate information for the stated purpose. While four earlier temperature value groups have been discussed, those skilled in the art may choose an appropriate number of earlier groups for a particular application. For example, some embodiments may use dozens (e.g., 95), hundreds (e.g., 990), or thousands (e.g., 9950, 10000, or more) of earlier temperature value groups, or any number between four and these exemplary numbers.

[0089] Some implementations may use more data than just a specific point on temperature sensor 26. In this case, thermogram generator 66 can use the thermogram to determine the temperature of one or more other parts of the patient's foot 10 to obtain geographic temperature data of interest. Therefore, if the patient's position on platform 16 is inappropriate, or if the exact location cannot be obtained through other means of obtaining temperature values, then thermogram generator 66 can adjust the data and generate an accurate temperature value hypothesis. As noted in the incorporated patent, such indirectly obtained temperature values ​​can be obtained using interpolation and similar techniques.

[0090] Next, in step 704, the normalizer 77 sets a normalization reference for the current and earlier temperature value groups. As described above, in a preferred embodiment, the normalization reference normalizes the temperature data values ​​relative to a common reference to form normalized data. Each temperature value group is preferably normalized using the same normalization reference. In an illustrative embodiment, the normalizer 77 provides normalization reference functionality, thus normalizing the temperature value groups using one or a combination of one or more methods / techniques. Each of these techniques is preferably applied to all temperature value groups in the analysis. Furthermore, each of these techniques can be used as part of a mathematical function or algorithm to normalize the temperature values.

[0091] Exemplary Standardized Reference Method 1: Comparison between contralateral positions

[0092] Temperature at any location on one foot 10 can be compared to temperature at any location on the other foot 10. For example, due to anatomical symmetry, the temperature on the big toe of the left foot 10 may be a good reference point for comparison with the temperature on the big toe of the right foot 10. In cases where the PAD or PDN affects one limb more than the contralateral limb, the temperature difference between the two locations may increase over time as the disease progresses. This could be due to a slow trend caused by disease progression or an acute condition, such as thrombosis.

[0093] • Implementation Method A: Anatomical Matching. Measure the temperature at one location on one foot 10 and the temperature at the same location on the other foot 10, calculate the absolute value of the difference between the two locations, and compare this difference to a predetermined threshold (e.g., 2°C) to determine whether the temperature pattern indicates the presence of a complication.

[0094] • Implementation Method B: Anatomical Difference. The temperature at one location on one foot 10 and the temperature at a different location on the other foot 10 are measured, and the absolute value of the difference between the two locations is calculated. This difference is then compared to a predetermined threshold (e.g., 2°C) to determine if the temperature pattern indicates the presence of a complication. This implementation method allows patients with previously amputated limbs who may lack anatomical structures to achieve anatomically matched contralateral temperature comparisons. In this case, an area proximate to the anatomical structure of the amputation can be used. For example, if a patient's right big toe has been amputated but the left big toe is preserved, the temperature of the left big toe can be compared to the temperature of the ball of the foot 10.

[0095] Exemplary Standardized Reference Method 2: Comparison between positions on the same side

[0096] The temperature at any point on the foot 10 can be compared to another point on the same foot 10. For example, the heel can serve as a stable reference point because its temperature remains relatively stable over time compared to more distal parts of the foot 10. The distal parts of the foot may not be as effective as the proximal parts in regulating body temperature in response to changes in ambient temperature, resulting in a greater temperature difference.

[0097] • Implementation Method A: Absolute value above a certain threshold. Measure the temperature at two locations, calculate the absolute value of the difference between the two locations, and compare this difference with a predetermined threshold (e.g., 2°C) to determine whether the temperature pattern indicates the presence of a complication.

[0098] • Implementation Method B: Asymmetric Thresholds. Measure the temperature at two locations on foot 10. Subtract the temperature at location 1 from the temperature at location 2 and compare it to threshold A. Then subtract the temperature at location 2 from the temperature at location 1 and compare it to threshold B, where threshold A and threshold B are different. Then determine whether either difference exceeds two different predetermined thresholds. This implementation method can detect complications leading to abnormally warm areas and complications that may lead to abnormally cool areas.

[0099] • Implementation Method C: Unique Thresholds for Different Locations. Measure the temperature at three locations on foot 10. Subtract the temperature at location 1 from the temperature at location 2 and compare it to threshold A. Then subtract the temperature at location 3 from the temperature at location 2 and compare it to threshold B. Determine whether any difference exceeds two different pre-determined thresholds. This implementation optimizes accuracy for various anatomical locations. For example, the toes may require a higher threshold than the heel because the temperature variation is greater in the more distal regions of foot 10.

[0100] Exemplary Standardized Reference Method 3: Comparison of Location and Statistics

[0101] Instead of relying on comparisons of individual locations, it's better to compare them against statistics summarizing the temperature of the entire foot 10, as individual locations may exhibit unstable temperature patterns over time. In cases where the entire foot 10 may be affected by PAD or PDN, such as widespread inflammation, the temperature of the entire foot 10 may vary over time. Furthermore, if the location of vascular damage is unknown, methods for identifying the locations of lowest and / or highest temperatures are highly sensitive for recognizing changes in foot 10 health.

[0102] • Implementation Method A: Compare with a central tendency statistic (such as the mean or median). Measure the temperature at multiple discrete locations or continuous sections, up to 10, and calculate the mean or median temperature. Measure the temperature at another location, which may be within or outside the mean area. Then subtract the mean from the temperature at the location of interest and compare it to a threshold.

[0103] • Implementation Method B: Comparison with the minimum value. The lowest temperature is calculated from a set of discrete temperature values ​​of foot 10 or from a continuous portion. If a continuous portion of foot 10 is used, this region can exclude data within a certain boundary starting from the edge of foot 10. The temperature at another location is measured, either within or outside the region of the mean. The minimum value is then subtracted from the temperature of the location of interest and compared with a threshold.

[0104] • Implementation Method C: Percentile Comparison. Similar to Implementation Method B, but instead of calculating the lowest temperature value used for comparison, a predetermined percentile is used, such as the 10th percentile. This method avoids extreme values ​​in the low or high end of the temperature distribution, which could lead to inaccurate analysis.

[0105] • Implementation Method D: Comparison with Statistical Distribution. Calculate the statistical distribution of a set of discrete temperature values ​​over a range of 10, or within a continuous portion of the range of 10. Measure the temperature at another location, either within or outside the mean range. Then use common statistical methods to determine whether the location of interest falls within the distribution range.

[0106] Exemplary Standardized Reference Method 4: Comparison of Temperature Range and Threshold

[0107] The range of foot temperature data sets can capture both abnormally warm and abnormally cold locations, conveniently presenting them as a single statistical statistic that can be easily compared to thresholds. In a healthy foot with normal blood flow, the entire foot is expected to have good vascularity and a warm, oxygenated blood supply, resulting in a generally uniform temperature distribution, i.e., a lower temperature range. However, in feet affected by PAD or PDN, certain parts of the foot may appear significantly warmer than others.

[0108] • Implementation Method A: Range of Discrete Temperature Locations. Measure the temperature at multiple discrete temperature locations on foot 10. Calculate the range of this set of temperatures and compare it to a predetermined threshold to determine whether the temperature pattern indicates the presence of a complication.

[0109] • Implementation Method B: Range of Continuous Temperature Data. Measure the temperature of a continuous area on foot 10. If necessary, exclude data within the boundary starting from the edge of foot 10. Calculate the temperature range within this area and compare it to a predetermined threshold to determine if the temperature pattern indicates a complication.

[0110] Exemplary Standardized Reference Method 5: Variation over Time

[0111] In some cases, absolute temperature at a given time is not as informative as temperature variation over time. Chronic diseases may present with slow, prolonged changes, while acute diseases may present with rapid onset or a short-lived pattern. Changes in foot temperature over short or long periods can indicate the progression of PAD or PDN.

[0112] • Implementation Method A: Simple Threshold Above Baseline. Foot temperature is measured and stored at a baseline time reference. Then, for a later time t, the foot temperature is measured again. The temperature at time t is compared to the baseline temperature to determine if the temperature change at any location relative to the baseline exceeds a predetermined threshold. Alternatively, the temperature difference between locations on foot 10 is measured, and the spatial difference is compared to the baseline spatial difference. The advantage of this method is that it allows for personalized analysis based on an individual's specific foot temperature pattern. However, it assumes that the baseline temperature is a healthy reference location, which is not the case for individuals with recently healed wounds or other active complications.

[0113] • Implementation Method B: Moving Average Baseline. In a related implementation, the baseline temperature can be calculated as a moving average or a filtered result of time series from multiple temperature data sets at different time locations. The average can be taken from a small sample to optimize the detection of acute changes in foot temperature, or from a large sample to optimize the detection of subtle changes or chronic conditions.

[0114] • Implementation Method C: Integral Variation of Temperature over Time. In another implementation, the initial temperature can be compared to a baseline reference or a static threshold for each set of data values ​​in the sample time series. These comparisons can then be summed, integrated, or otherwise aggregated to produce a summary statistic of variation over time. The advantage of this approach is that it emphasizes the persistence of variation over time while filtering out noisy or inconsistent temperature fluctuations.

[0115] Exemplary Standardized Reference Method 6: Comparison with the Environment

[0116] Comparing foot temperature to ambient temperature (i.e., ambient temperature value) provides an opportunity to detect concurrent PAD or PDN within the foot, even when there may be no spatial variation within the foot. As mentioned above, the thermoregulation capacity of the foot affected by PAD or PDN is poorer than that of healthy limbs, resulting in a lower temperature difference between the foot and the ambient environment in cold conditions. Furthermore, systemic inflammation caused by PDN may lead to a larger temperature difference between the foot and the ambient environment.

[0117] • Implementation Method A: Compare the central tendency statistics of the temperature value with the ambient temperature. Measure the ambient temperature using the background signal from a temperature sensor (e.g., the background of a thermal image or a non-foot area of ​​a 2D temperature scan) or the background signal from a standalone temperature sensor that does not measure foot temperature. Measure the temperature of the entire foot¹⁰ and calculate the central tendency statistics (e.g., mean, median, pattern). Compare the central tendency statistics with the ambient temperature and determine if the difference exceeds a predetermined threshold.

[0118] • Implementation Method B: Comparing a specific location with ambient temperature. In a related implementation, the ambient temperature is measured, and then the foot temperature at a specific location or area on the foot 10 is measured. The temperature at that location is compared with the ambient temperature, and it is determined whether the difference exceeds a predetermined threshold. The advantage of this implementation is that it allows clinicians or researchers to select a relatively stable and consistent location on the foot 10, which is not as easily affected by environmental or other temporary disturbances as other locations.

[0119] • Implementation C: Compare the maximum value with the ambient temperature. In another related implementation, the ambient temperature is measured, and then the temperature of the entire foot 10 is measured to calculate the maximum temperature of the foot 10. The maximum temperature is compared with the ambient temperature, and it is determined whether the difference exceeds a predetermined threshold. This implementation is expected to provide good sensitivity in cases where the hottest part of the foot 10 may move from one scan to another.

[0120] Exemplary Standardized Reference Method 7: Comparison with Body Temperature

[0121] This method is similar to Exemplary Standardized Reference Method 5, but is less susceptible to intermittent or irregular fluctuations in ambient temperature caused by changes in environmental conditions. Comparing foot temperature to body temperature can provide a more accurate basis for detecting complications by taking into account external variables affecting foot temperature.

[0122] • Implementation Method A: Comparison with internal body temperature. Measure the internal body temperature at the core site or preferably on the limb closest to the surface measurement location. Then compare the surface foot temperature measurement with the internal body temperature and determine whether the difference exceeds a predetermined threshold.

[0123] • Implementation Method B: Limb Surface Temperature. The surface temperature of the limb is measured, preferably near the foot (e.g., ankle or leg). The foot surface temperature measurement is then compared with the limb surface temperature to determine if the difference exceeds a predetermined threshold. This implementation method may be easier to obtain (compared to internal body temperature) because the surface temperature sensor 26 can be adhered to the skin to collect surface temperature. An additional advantage of this method is that it limits the influence of ambient temperature, body activity, and blood vessels, which typically affect the limb and foot 10.

[0124] Exemplary Standardized Reference Method 8: Isothermal Region

[0125] For certain complications, such as monitoring wound healing, the size of the area of ​​elevated temperature may be more informative than the specific temperature of that area.

[0126] • Implementation Method A: Comparing Isothermal Zones. Select a comparison from any of the exemplary standardized reference methods described above and calculate the difference between each location in the set of temperature data and the comparison value. Then determine which locations, pixels, or regions are above a predetermined threshold. Calculate the number of points, pixels, or area (e.g., cm²) of regions exceeding the threshold. Determine whether the temperature rise in the region exceeds the predetermined threshold.

[0127] • Implementation Method B: Monitoring the isothermal region as it changes over time. Similar to Exemplary Standardized Reference Method 7, Implementation Method A, except that it determines whether the isothermal region changes over time.

[0128] In itself, the exemplary standardized reference methods 1-8 can detect a unique type of diabetic complication at least 10 minutes of age and can be optimized to detect this complication with high sensitivity and specificity. However, using only one method may not be generalizable to other types of complications. Therefore, illustrative implementations may combine two or more of the exemplary standardized reference methods 1-8, or use them individually. For example, two or more of these methods may be used with simple logical terms or linear combinations to provide more accurate predictions. For example, some implementations combine two, three, four, five, six, or seven of these methods, or combine one or more methods with another method not discussed.

[0129] In one implementation, two or more of the above methods are combined with an OR statement. For example, if exemplary standardized reference method 1 is true or exemplary standardized reference method 2 is true, then the probability of complications is high. The advantage of this combination is that it allows for the specialization of methods to detect certain types of complications and naturally increases the sensitivity of the detection system across a variety of complications. In another implementation, the methods can be combined with an AND statement. For example, if exemplary standardized reference method 1 is true and exemplary standardized reference method 2 is true, then the probability of complications is high. Therefore, this combination can create highly specific detection methods.

[0130] In another implementation, methods can be combined as a linear combination of continuous or categorical outputs. For example, if two methods are combined, each producing a continuous variable output such as degrees Celsius, the combined formula can multiply each method variable by a coefficient to obtain a final result, which can then be used to determine the probability of complication PAD or PDN. In this implementation, the formula could be of the form R = A*M1 + B*M2, where R is the risk, M1 and M2 are variables of exemplary standardized reference method 1 and exemplary standardized reference method 2, and A and B are coefficients. An additional benefit of this combination technique is that the weighting of variables is non-uniform, depending on which variable has a greater impact on the complication of interest to the researcher. Furthermore, it can simultaneously optimize all independent input variables to obtain a system that maximizes sensitivity and / or specificity according to the researcher's objectives.

[0131] Those skilled in the art will recognize that threshold optimization can be performed on a per-method basis or on a set of methods in any combination to optimize the sensitivity and specificity of the combined methods.

[0132] Furthermore, rather than applying simple thresholds (whether for a single set of foot temperature measurements or for multiple sets of temperature values) to identify risk, the magnitude of any of the indicators given in Exemplary Standardized Reference Methods 1-8 can also be informative about risk. For example, a large difference in temperature difference described in Exemplary Standardized Reference Method 1 may indicate a higher risk than a smaller difference in temperature difference.

[0133] Those skilled in the art will recognize that the temperature of certain areas of the foot 10 may be more informative in identifying the presence or progression of diabetic complications such as PAD or PDN. For example, since both diseases are progressive, starting from the most distal parts of the foot anatomy (such as the toes), the temperature of the toes may be more important for predicting, identifying, and monitoring the progression of PAD or PDN. As another example, the inventors were surprised to find, using their open-platform device, that the temperature of the mid-medial segment of the foot 10 or the arch of the foot has a disproportionate predictive value for the presence of PDN, likely due to vasodilation of the medial plantar artery, which branches near the arch of the foot 10. Illustrative embodiments use specific techniques to facilitate data access to this finding. In some embodiments, large differences when comparing the temperature of one or more toes to ambient temperature, and / or when comparing the temperature of the midfoot to ambient temperature, may indicate a problem.

[0134] Therefore, after setting the standardized reference, the process continues to step 706, where modeler 79 transforms the standardized data regarding the exemplary standardized reference method, as described above, into model information conforming to one or more models. These models represent the progress of the PDN or PAD. In various implementations, this transformation of the model typically reduces or eliminates noise in the temperature values ​​to produce a more accurate determination of the PDN or PAD trajectory. Figure 8B-8D Standardized data and simple models are illustrated using the standardized references mentioned above related to environment, opposite side, and same side references (exemplary standardized reference methods 6, 1, and 2, respectively). Figure 8A The diagram shows a model that uses temperature values ​​without standardization. In these cases, the points represent standardized data of a given geographic area changing over time. The straight line generated by modeler 79 serves as the model, simplifying the trend and details of the points representing standardized data as described above. In fact, the straight line is a simple example for illustrative purposes only. Other models can be applied by those skilled in the art, as described below.

[0135] Next, in step 708, modeler 79 uses the model information to determine the trajectory of the patient's PDN or PAD. Among other things, this trajectory may include one or both of the rate of change and the magnitude of the model information. The determination of this trajectory enables modeler 79 (which may have a predictor (not shown)) to predict the future state of the patient's PDN or PAD relative to the current set of temperature values ​​(step 710).

[0136] Some implementations can be subjected to simple statistical evaluation to determine whether any of the methods and values ​​listed above show a trajectory indicating that PAD or PDN is rising or falling over time. The magnitude of these trends and the trends within a predetermined time frame can indicate whether PAD or PDN is progressing, stabilizing, or resolving.

[0137] Other implementations can utilize machine learning and advanced filtering techniques to identify risks and predictions related to the presence or progression of PD12 or PDN using standardized data and corresponding model information as described above. More specifically, advanced statistical models can be applied to estimate the current state and health of a patient's foot 10 and predict future changes in foot health. State estimation models, such as switched Kalman filters, can be processed as model information and relevant data become available and updated in real time to their estimates of the user's current foot 10 state. Statistical models can combine expert knowledge based on clinical experience and published studies (e.g., specifying which variables and factors should be included in the model) with real data collected and analyzed from users. This allows for training and optimization of the model according to various performance metrics.

[0138] As more data is collected, models can be continuously improved and updated to reflect state-of-the-art clinical research. These models can also be designed to account for various potential confounding factors, such as physical activity (e.g., running), environmental conditions (e.g., cold floors), individual baselines, past injuries, predisposition to developmental problems, and other known complications. In addition to using these models to provide real-time analysis of users, they can also be used offline to detect significant patterns in large historical archives of data.

[0139] Exemplary reference methods 9 to 10 may extend exemplary standardized reference methods 1-8 and are illustratively, primarily, or entirely related to modeling (discussed above). For example, methods 9 and 10 may utilize previously indicated methods 1-8 during the modeling process.

[0140] Exemplary Method 9: Statistical Inference

[0141] Foot temperature data from patients processed using one or more of the exemplary standardized reference methods 1-8 can be used to predict the future development of PAD or PDN using one of several statistical inference models.

[0142] • Implementation Method A: ARMA (Autoregressive Moving Average) or ARIMA (Autoregressive Integral Moving Average) model. A statistical regression model of the trend from a patient's temperature data can be used to predict future values ​​of this temperature data, which can be used to determine the future presence of PAD or PDN. Such a model can be used to predict future temperature data using only previous data from the same patient. One advantage of this approach is its ability to handle the prediction of non-steady-state processes, such as evidence of non-steady-state progressive foot disease. Another advantage is its ability to handle seasonal and other periodic fluctuations.

[0143] • Implementation Method B: Kalman Filtering. Similarly, Kalman filtering can fit standardized data from patients and / or the temperature values ​​themselves from each group and can be used to predict future values ​​of these temperature data. Using this method, Kalman filtering can be used to determine whether a PAD or PDN will exist in the future. This implementation relates to Implementation Method A of Exemplary Method 10 below, assuming all variables follow a Gaussian distribution. Using Kalman filtering to predict and estimate PAD or PDN has advantages over Implementation Method A of Exemplary Method 10, including improved computational efficiency and stability.

[0144] Exemplary Method 10: Anomaly Detection

[0145] Foot temperature data from patients processed using one or more of the methods detailed in Exemplary Standardized Reference Methods 1-8 can be used to monitor the progression of PAD or PDN using one of several anomaly detection methods.

[0146] • Implementation Method A: Unsupervised Detection. One or more unsupervised anomaly detection methods can be applied to foot temperature data from patients. For example, density-based techniques, such as clustering and cluster membership tests, can be applied to foot temperature data to determine whether a trend in foot temperature indicates progression of PAD or PDN. More sophisticated unsupervised detection methods, such as Hidden Markov Models, can also be used to determine whether a foot temperature trend indicates progression of PAD or PDN. Those skilled in the art will recognize that, in this implementation, training data is not required to build a model to determine whether PAD or PDN is progressing.

[0147] • Implementation Method B: Supervised Detection. If training data is available, supervised anomaly detection techniques can be utilized. In this case, foot temperature data from patients with and without diabetic complications PAD or PDN are used to build a classification model. This model is subsequently evaluated as additional foot temperature data becomes available to determine if PAD or PDN is progressing. Alternatively, foot temperature data from patients exhibiting progression of DM complications can be used to build a classification model. A simple model can compare the foot temperature data over time from new patients with foot temperature data over time from patients with known progression of PAD or PDN or without progression. Nearest neighbor classifiers, dynamic time-warped classifiers, or another time-series classifier can be used when progression data is available.

[0148] • Implementation Method C: Semi-supervised anomaly detection. This method relies on data from patients with or without progression of diabetic complications to build a model without progression. This model can be statistically significant, such as the distribution of foot temperature values ​​from patients with or without progression of complications. Statistical tests can then be applied to determine whether foot temperature data from new patients belong to this baseline model, or whether there is a difference, the latter indicating progression of diabetic complications such as peripheral arterial disease (PAD) or peripheral neuropathy (PDN).

[0149] Various embodiments of the present invention can be implemented, at least in part, using any conventional computer programming language. For example, some embodiments can be implemented using a procedural programming language (such as "C") or an object-oriented programming language (such as "C++"). Other embodiments of the present invention can be implemented as pre-programmed hardware elements (e.g., application-specific integrated circuits, FPGAs, and digital signal processors) or other related components.

[0150] In alternative implementations, the disclosed apparatus and methods (e.g., see the various flowcharts above) may be implemented as a computer program product (or in a computer program) for use with a computer system. Such implementations may include a set of computer instructions fixed on a tangible medium, such as a computer-readable medium (e.g., a floppy disk, CD-ROM, ROM, or fixed disk) or transferable to a computer system via a modem or other interface device, such as a communication adapter connected to a network via a medium.

[0151] The medium can be a tangible medium (e.g., an optical or analog communication line) or a medium implemented using wireless technology (e.g., Wi-Fi, microwave, infrared, or other transmission technologies). The medium can also be a non-transient medium. This set of computer instructions can embody all or part of the functionality of the system described above. The processes described herein are merely exemplary, and it is understood that various alternatives, mathematical equivalents, or derivations thereof are within the scope of this invention.

[0152] Those skilled in the art will understand that such computer instructions can be written in a variety of programming languages ​​for use in many computer architectures or operating systems. Furthermore, such instructions can be stored in any storage device, such as semiconductor, magnetic, optical, or other storage devices, and can be transmitted using any communication technology, such as optical, infrared, microwave, or other transmission technologies.

[0153] In other ways, such computer program products can be pre-installed on a computer system (e.g., on system ROM or a fixed disk) as removable media with accompanying printed or electronic documents (e.g., compressed packaging software), or distributed from a server or electronic bulletin board via a larger network (e.g., the Internet or the World Wide Web). Of course, some embodiments of the invention can be implemented as a combination of software (e.g., computer program products) and hardware. Other embodiments of the invention are implemented entirely as hardware or entirely as software.

[0154] The embodiments of the present invention described above are merely exemplary; many variations and modifications will be apparent to those skilled in the art. These variations and modifications are all within the scope of the present invention.

Claims

1. A monitor for managing peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in patients with plantar surface involvement, the monitor comprising: The body has a base with a top surface, the top surface of which has a receiving area configured to receive the soles of feet; the base forms an open platform or a closed platform. The main body has a group of temperature sensors connected to the top surface of the receiving area, the group of temperature sensors being spaced apart within the receiving area and configured to activate upon receiving a stimulus applied to one or both of the open or closed platform and the group of temperature sensors. The group of temperature sensors is configured to be in thermal communication with the sole of the foot within the receiving area to determine the current temperature at each of a set of spaced-apart locations on the sole of the foot. The group of temperature sensors is configured to generate a set of temperature values, where each location has a corresponding temperature value. The temperature sensor array is configured to generate a set of current temperature values ​​after the receiving area comes into contact with the patient's sole surface. The input device is configured to receive four or more earlier temperature values ​​of the patient's sole surface generated at an earlier time. A normalizer is configured to 1) set a normalization reference for four or more earlier temperature value groups of the patient's sole surface generated at an earlier time and 2) the current temperature value group, in order to generate standardized data by normalizing the four or more earlier temperature value groups and the current temperature value group relative to the normalization reference. A modeler operatively connected to the normalizer, the modeler being configured to transform the normalized data into model information representing the progression of PDN or PAD, the modeler being configured to use the model information to determine the trajectory of the patient's PDN or PAD.

2. The monitor according to claim 1, wherein, Each of the earlier temperature value groups is separated from the other earlier temperature data value groups by at least one day in time.

3. The monitor of claim 1, wherein the normalizer is configured to apply a normalization function to the earlier temperature value group and the current temperature value group using the normalization reference to generate the normalized data. The standardized reference includes one or more opposite temperature values.

4. The monitor of claim 1, wherein the normalizer is configured to apply a normalization function to the earlier temperature value group and the current temperature value group using the normalization reference to generate the normalized data. The standardized reference includes one or more temperature values ​​on the same side.

5. The monitor of claim 1, wherein the normalizer is configured to apply a normalization function to the earlier temperature value group and the current temperature value group using the normalization reference to generate the normalized data. The standardized reference includes one or more ambient temperature values.

6. The monitor according to claim 1, wherein, The input device is configured to receive 4 earlier groups to 10,000 earlier groups.

7. The monitor according to claim 1, wherein, The modeler is configured to select a model that represents the standardized data as a simpler system; and The model is applied to the standardized data to generate the model information.

8. The monitor of claim 1, further comprising a predictor configured to use the model information to predict the future state of a patient's PDN or PAD relative to the current temperature value group.

9. The monitor according to claim 1, wherein, The trajectory includes one or both of the rate of change of the model information and the magnitude of the model information.

10. The monitor according to claim 1, wherein, The modeler is configured to transform multiple earlier temperature value sets and the current temperature value set into model information representing the progress of PDN, relative to the normalized reference. The modeler is then configured to use this information to determine the trajectory of the patient's PDN.

11. The monitor according to claim 1, wherein, The modeler is configured to transform multiple earlier temperature value sets and the current temperature value set into model information representing the progress of PAD, relative to the normalized reference. The modeler is then configured to use this information to determine the trajectory of the patient's PAD.

12. A method for monitoring peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in patients with plantar surface involvement, the method comprising: A main body is provided with a base having a top surface, the top surface of which has a receiving area configured to receive the soles of feet; the base forms an open platform or a closed platform. The main body has a set of temperature sensors connected to the top surface of the receiving area, within the receiving area, and configured to activate upon receiving a stimulus applied to one or both of the open or closed platform and the temperature sensor set. The temperature sensor set is configured to be in thermal communication with the sole of the foot within the receiving area to determine the current temperature at each of a set of different locations on the sole of the foot. The temperature sensor set is configured to generate a set of temperature values, where each location has a corresponding temperature value. The patient's sole surface is brought into contact with the receiving area so that the temperature sensor array generates the current temperature value set. Access four or more earlier temperature sets of the patient's sole surface generated at an earlier time. Standardized references are established for 1) four or more earlier temperature sets of the patient's foot surface generated at an earlier time and 2) the current temperature set, to generate standardized data by standardizing the four or more earlier temperature sets and the current temperature set relative to the standardized references. The standardized data is transformed into model information representing the progress of PDN or PAD, and The model information is used to determine the trajectory of the patient's PDN or PAD.

13. The method according to claim 12, wherein, Each of the earlier temperature value groups is separated from the other earlier temperature data value groups by at least one day in time.

14. The method according to claim 12, wherein, The setting of the standardization reference includes: applying a standardization function to the earlier temperature value group and the current temperature value group using the standardization reference. The standardized reference includes one or more opposite temperature values.

15. The method according to claim 12, wherein, The setting of the standardization reference includes: applying a standardization function to the earlier temperature value group and the current temperature value group using the standardization reference. The standardized reference includes one or more temperature values ​​on the same side.

16. The method according to claim 12, wherein, The setting of the standardization reference includes: applying a standardization function to the earlier temperature value group and the current temperature value group using the standardization reference. The standardized reference includes one or more ambient temperature values.

17. The method according to claim 12, wherein, The phrase "accessing four or more earlier groups" includes accessing from four earlier groups to 10,000 earlier groups.

18. The method according to claim 12, wherein, The transformation includes: Choose a model that represents the standardized data as a simpler system; and The model is applied to the standardized data to generate the model information.

19. The method of claim 12, further comprising using the model information to predict the future state of a patient's PDN or PAD relative to the current temperature value group.

20. The method according to claim 12, wherein, The trajectory includes one or both of the rate of change of the model information and the magnitude of the model information.

21. The method according to claim 12, wherein, The transformation includes: relative to the standardized reference, converting multiple earlier temperature value sets and the current temperature value set into model information representing the progress of PDN. Furthermore, the determination mentioned therein includes using the information to determine the trajectory of the patient's PDN.

22. The method according to claim 12, wherein, The transformation includes converting multiple earlier temperature value sets and the current temperature value set into model information representing the progress of PAD, relative to the standardized reference. Furthermore, the determination mentioned therein includes using the information to determine the trajectory of the patient's PAD.

23. The method of claim 12, further comprising using the temperature sensor set to generate the earlier temperature value set.

24. A system for monitoring peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in patients with a plantar surface, the system comprising: The body has a base with a top surface, the top surface of which has a receiving area configured to receive the soles of feet; the base forms an open platform or a closed platform. The main body has a group of temperature sensors connected to the top surface of the receiving area, the group of temperature sensors being spaced apart within the receiving area and configured to activate upon receiving a stimulus applied to one or both of the open or closed platform and the group of temperature sensors. The group of temperature sensors is configured to be in thermal communication with the sole of the foot within the receiving area to determine the current temperature at each of a set of spaced-apart locations on the sole of the foot. The group of temperature sensors is configured to generate a set of temperature values, where each location has a corresponding temperature value. The temperature sensor array is configured to generate a set of current temperature values ​​after the receiving area comes into contact with the patient's sole surface; and Computer program product for a computer system, the computer program product including a tangible, non-transient, computer-usable medium having computer-readable program code thereon, the computer-readable program code including: Program code for receiving four or more earlier temperature sets of the patient's sole surface generated at an earlier time. Program code for setting a standardized reference for 1) four or more earlier temperature value groups of the patient's sole surface generated at an earlier time and 2) the current temperature value group, to generate standardized data by standardizing the four or more earlier temperature value groups and the current temperature value group relative to the standardized reference. Program code used to convert the standardized data into model information representing the progress of PDN or PAD; and Program code for using the model information to determine the trajectory of a patient's PDN or PAD.

25. The system according to claim 24, wherein, Each of the earlier temperature value groups is separated from the other earlier temperature data value groups by at least one day in time.

26. The system of claim 24, wherein the program code for setting the normalization reference includes program code for applying a normalization function to the earlier temperature value group and the current temperature value group using the normalization reference. The standardized reference includes one or more opposite temperature values.

27. The system of claim 24, wherein the program code for setting the normalization reference includes program code that applies a normalization function to the earlier temperature value group and the current temperature value group using the normalization reference. The standardized reference includes one or more temperature values ​​on the same side.

28. The system of claim 24, wherein the program code for setting the normalization reference includes program code for applying a normalization function to the earlier temperature value group and the current temperature value group using the normalization reference. The standardized reference includes one or more ambient temperature values.

29. The system according to claim 24, wherein, Program code for accessing more than four earlier groups includes accessing 4 to 10,000 earlier groups.

30. The system according to claim 24, wherein, The program code used for the conversion includes: Program code for selecting a model that represents the standardized data as a simpler system; and Program code used to apply the model to the standardized data to generate the model information.

31. The system of claim 24, further comprising program code for using the model information to predict the future state of a patient's PDN or PAD relative to the current temperature value group.

32. The system according to claim 24, wherein, The trajectory includes one or both of the rate of change of the model information and the magnitude of the model information.

33. The system according to claim 24, wherein, The program code used for the transformation includes: program code that, relative to the standardized reference, transforms multiple earlier temperature value sets and the current temperature value set into model information representing PDN progress. Furthermore, the program code used for determination includes program code for using the information to determine the trajectory of the patient's PDN.

34. The system according to claim 24, wherein, The program code used for the conversion includes: program code that, relative to the standardized reference, converts multiple earlier temperature value sets and the current temperature value set into model information representing the progress of PAD. Furthermore, the program code used for determination includes program code for using the information to determine the trajectory of the patient's PAD.

35. A non-contact method for monitoring peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in patients with plantar surface involvement, the method comprising: Provide thermal imagers with infrared radiation sensors; The infrared radiation sensor is directed toward the patient's sole surface to generate a current set of thermal images representing the current temperature across the entire sole surface. Access more than four earlier thermal image sets of the patient's sole surface generated at an earlier time; A standardized reference is set for 1) four or more earlier thermal image datasets of the patient's foot surface generated at an earlier time and 2) the current thermal image dataset, to generate standardized data by standardizing the four or more earlier thermal image datasets and the current thermal image dataset relative to the standardized reference. The standardized data is transformed into model information representing the progress of PDN or PAD; and The model information is used to determine the trajectory of the patient's PDN or PAD.

36. The method according to claim 35, wherein, Each of the earlier thermal image data sets is separated from the other earlier thermal image data sets by at least one day in time.

37. The method of claim 35, wherein the setting of the standardized reference includes: A normalization function is applied, which includes earlier and current thermal image data sets as inputs to the normalization function. The setup also includes applying the normalization function to one or more contralateral temperature values.

38. The method of claim 35, wherein the setting of the standardized reference includes: The normalization function is applied to the earlier thermal image data set and the current thermal image data set, which are used as inputs to the normalization function. The settings also include applying the normalization function to one or more temperature values ​​on the same side.

39. The method of claim 35, wherein the setting of the standardized reference includes: The normalization function is applied to the earlier thermal image data set and the current thermal image data set, which are used as inputs to the normalization function. The settings also include applying a normalization function to one or more ambient temperature values.

40. The method of claim 35, wherein, The transformation includes: The model chosen represents both the earlier and current thermal imaging datasets as simpler systems; and The model is applied to multiple earlier thermal imaging datasets and the current thermal imaging dataset to generate the model information.

41. The method according to claim 35, wherein, The trajectory includes one or both of the rate of change of the model information and the magnitude of the model information.

42. A monitor for monitoring peripheral diabetic neuropathy ("PDN") and / or peripheral artery disease ("PAD") in patients with a plantar surface, the monitor comprising: A thermal imager with an infrared radiation sensor configured to selectively generate a current set of thermal images representing the current temperature over the entire surface of the sole of the foot when the sensor is directed toward but not in contact with the patient’s sole surface. The input device is configured to receive four or more earlier sets of thermal images of the patient's sole surface, generated at an earlier time. A normalizer is configured to: 1) set a normalization reference for four or more earlier thermal image datasets of the patient's plantar surface generated at an earlier time, and 2) the current thermal image dataset, to generate normalized data by normalizing the four or more earlier thermal image datasets and the current thermal image dataset relative to the normalization reference. A modeler operatively connected to the normalizer, the modeler being configured to transform the normalized data into model information representing the progression of PDN or PAD, and the modeler being configured to determine the trajectory of the patient's PDN or PAD as a function of the model information.

43. The monitor according to claim 42, wherein, Each of the earlier thermal imaging data sets is separated from the other earlier thermal imaging data sets by at least one day in time.

44. The monitor according to claim 42, wherein, The normalizer is configured to apply a normalization function, which includes earlier and current sets of thermal imaging data as inputs to the normalization function. The normalizer is also configured to apply the normalization function to one or more contralateral temperature values.

45. The monitor according to claim 42, wherein, The normalizer is configured to apply a normalization function, which includes earlier and current sets of thermal imaging data as inputs to the normalization function. The normalizer is also configured to apply the normalization function to one or more temperature values ​​on the same side.

46. ​​The monitor according to claim 42, wherein, The normalizer is configured to apply a normalization function, which includes earlier and current sets of thermal imaging data as inputs to the normalization function. The normalizer is also configured to apply the normalization function to one or more ambient temperature values.

47. The monitor according to claim 42, wherein, The normalizer is configured as follows: Choose a model that characterizes both the earlier and current thermal imaging datasets as simpler systems; and The model is applied to multiple earlier thermal imaging datasets and current thermal imaging datasets to generate the model information.

48. The monitor according to claim 42, wherein, The trajectory includes one or both of the rate of change of the model information and the magnitude of the model information.

Images