Patent Application
20160150976
The present disclosure relates generally to a thermal imaging system, and more specifically, to exemplary embodiments of an exemplary thermal imaging system, device, apparatus, method and computer-accessible medium for diagnosing a disease or condition.
The slow progress towards the United Nations Millennium Development Goal #4—to reduce childhood mortality in developing countries by two thirds by 2015—has been partly related to the lack of progress in early recognition, diagnosis and treatment of childhood pneumonia and neonatal pneumonia. Severe pneumonia can be one of the most common reasons that children under age 5 die (see e.g., Reference 11), but can often be elusive and difficult to diagnose in young infants and children based on history and physical examination alone. (See e.g., References 2, 5, 7 and 9). In well-equipped health care facilities, chest x-rays can be used to aid in the diagnosis of severe pneumonia, and to guide critical treatment decisions (e.g., whether a patient needs parenteral antibiotics or not, hospitalization or not, use of oxygen or not). However, in low and middle-income countries (“LMIC”s), children are typically taken to first-level health clinics where chest x-rays may not be available, and the diagnosis of severe pneumonia can often be missed.
Distinguishing pneumonia from bronchiolitis and other viral respiratory infections can be especially difficult in infants and young children. Treatment guidelines from the American Academy of Pediatrics (“AAP”), the American College of Chest Physicians (“ACCP”), and the Centers for Disease Control and Prevention (“CDC”) all recommend not treating viral illnesses with antibiotics. However, for any patient with cough and fever, it can be important to be sure that the patient does not have pneumonia, because pneumonia can be more likely due to bacteria and does need antibiotic treatment. By detecting temperature patterns, case reports in technical and scientific articles have demonstrated thermal imaging has a strong diagnostic potential for detecting lobar consolidation or asymmetric infiltrates consistent with the diagnosis of pneumonia. (See e.g., References 6 and 10).
Thermal imaging capability can provide frontline healthcare workers with a portable and durable technology to make quick, appropriate decisions to distinguish bacterial pneumonia (e.g., a lower respiratory tract disease) from other respiratory infections, such as acute bronchiolitis (e.g., a lower respiratory tract disease) or acute bronchitis (e.g., an upper respiratory tract disease), which are usually caused by viruses over 90% of the time. With a point-of-care thermal imaging device, a healthcare worker can more appropriately guide the timely use of antibiotics for those who have pneumonia. This can be important for resource-poor settings, where medicines and resources can be extremely limited. The scant doses of antibiotics, single oxygen canisters or resources to transfer a patient to the district hospital can be given only to those with severe pneumonia. In addition, the information provided by a thermal scanner can assay in the evaluation of patients of any age, providing critical information to guide healthcare workers to better triage the patient pool to improve health outcomes.
Body temperature can be one of the vital signs routinely measured by healthcare providers and veterinarians to detect or monitor medical problems. Body temperature can vary by gender, time of day, recent activity, and other factors, but normally can range from about 36.5° C. to 37.2° C. Temperatures higher and/or lower than this range can indicate the presence of infection, inflammation, fever and other processes.
Thermometry has been extensively used throughout the history of medicine. Traditional thermometers need direct contact and typically have a slow response. Infrared thermometry can be a technique measuring the thermal infrared radiation from a heated body, and thus does not need direct contact. Since its discovery in the early 1900's, infrared thermometry has been increasingly used for non-contact temperature assessment in medicine, civil engineering and science, from point measurements to temperature maps in the form of a thermal image. A thermal camera can be attractive because it can scan a large area in real-time. It has been widely used in airports to rapidly identify fever and infectious diseases in a high-traffic area. Because thermal infrared radiation, with wavelength ranging from about 9,000 nm to about 14,000 nm, can be invisible to the naked eyes of a human, a thermal camera can also be used for the purpose of night-vision.
Optical sensors, such as CCD and CMOS, have a limited sensitivity in the near-infrared (“NIR”) range and thus can be adapted to obtain thermal images. However, a suitable temperature range of these sensors can be relatively high, typically above 280° C. Specialized thermal imaging sensors, such as Focal Plane Arrays (“FPA”), can directly measure the thermal infrared radiations radiated from a wide range of temperatures. FPA-based thermal cameras are generally more affordable compared to the CCD/CMOS based thermal cameras; however, its resolution can be quite limited, only in the range of a few hundred pixels in each axis of the image. Because an FPA sensor can be more complicated to fabricate, a high-resolution FPA sensor can be very high in cost.
With over 6 billion mobile phones being actively used worldwide, there is one for every person over the age 10. Technological advances have contributed to the development of the smart phone, which can be a cellular phone that has built-in applications and internet access. Despite an uneven smartphone adoption rate, it is estimated that by 2015, over 2 billion smart phones will be in use globally. Africa's one billion people have 750 million phones, and by the year 2016 nearly 30% will be smart phones. (See e.g., Reference 3). A smart phone can combine the capabilities of connectivity, imaging and computing into a single, portable and low-cost platform, and has enormous potential for improving healthcare to patients, especially in global health. Given the nearly complete lack of diagnostic imaging and superior availability of camera phones in resource poor settings, the use of mobile phones in such settings can substantially improve the accessibility of diagnostic tools, which can be life-saving for the young infants and children most at risk.
Thus, it may be beneficial to provide an exemplary device, apparatus, method and computer-accessible medium for providing and/or facilitating thermal imaging that can be simple to use, and which can overcome at least some of the deficiencies described herein above.
Exemplary of the present disclosure can include an apparatus for generating thermal imaging information of a biological structure(s), including detector arrangement(s) which can be configured to receive and detect light radiation(s), a determination arrangement(s) which can be configured to determine a position or an orientation of the detector arrangement(s) with respect to the biological structure(s) to generate data, and a computer arrangement configured to generate the thermal imaging information based on the radiation and the data when the detector arrangement(s) is moved. The thermal imaging information can include a 2-dimension temperature map(s) of a portion(s) of the biological structure(s). The thermal imaging information can include a 3-dimension temperature map(s) of a portion(s) of the biological structure(s).
In some exemplary embodiments of the present disclosure, an indication of whether there can be an infection or an inflammation in a portion(s) of the biological structure(s) can be provided based on the thermal imaging information. The determination arrangement(s) can be configured to determine at a position(s) or the orientation using a sensor arrangement(s), which can include a gyroscope, an accelerometer or a camera. A plurality of images can be generated based on the radiation prior to the generation of the thermal imaging information. A reference image(s) can be generated based on the plurality of images. The thermal imaging information can be generated based on the reference image(s). It can be determined whether the reference image(s) contains all predetermined elements of the biological structures prior to the generation of the thermal imaging information.
In certain exemplary embodiments of the present disclosure, a diagnosis of the biological structure(s) can be determined based on the thermal imaging information. The diagnosis being determined can include a probability or a likelihood of an infection or an inflammation of the biological structure(s). The diagnosis being determined can include pneumonia. The diagnosis can be based on a temperature difference between a first portion of the biological structure(s) and a different second portion of the biological structure(s).
In another exemplary embodiment can be an exemplary system, method and computer-accessible medium for determining a diagnosis for a biological structure(s), which can include receiving first information related to a plurality of images of the biological structure(s), combining the plurality of images into a reference image(s), where the reference image(s) can include thermal imaging information associated with the biological structure(s), determining a sufficiency of the reference image(s), and determining the diagnosis for the biological structure(s) based on the reference image(s), and as a function of the sufficiency.
In some exemplary embodiments of the present disclosure, the plurality of images can include a thermal image(s) and a visual image(s). The first information can further include of a position or an orientation of the images. The images can be combined into the reference image(s) using a stitching procedure. The sufficiency can include a determination of whether the biological structure(s) can be rotated in the reference image(s) and/or whether the reference image(s) can include all critical elements of the biological structure(s). The diagnosis can include a probability or a likelihood of an infection or an inflammation of the biological structure(s). The diagnosis can include pneumonia.
In a further exemplary embodiment of the present disclosure, an apparatus can be provided for obtaining thermal imaging information of biological structure(s), including a detector arrangement(s) which can be configured to receive and detect a light radiation(s), a controller arrangement(s) which can be configured to control multiple angles or multiple positions at which the light radiation(s) can be received from the biological structure(s), and detected by the detector arrangement(s) to generate data, and a computer arrangement configured to generate the thermal imaging information based on the radiation and the data when the positions or the angles of the receipt of the radiation are changed.
According to still a further exemplary embodiment of the present disclosure, an apparatus can be provided for determining a probability or a likelihood of an infection or an inflammation of biological structure(s), which can include a detector arrangement(s) which can be configured to receive and detect a radiation(s) from the biological structure(s). A computer arrangement can be configured to generate thermal imaging information for a portion(s) of the biological structure(s) based on the radiation, and determine the probability or the likelihood of the infection or the inflammation based on the thermal imaging information. The thermal imaging information can be compared to a further imaging information, and the probability or the likelihood of the infection or the inflammation can be determined based on the comparison. The further imaging information includes a previously-obtained thermal imaging information for the portion(s). The probability or the likelihood of the infection or the inflammation is for pneumonia. The biological sample(s) can be a chest cavity. A recommendation for whether a treatment can be prescribed can be generated based on the probability or the likelihood of the infection or the inflammation.
These and other objects, features and advantages of the exemplary embodiments of the present disclosure will become apparent upon reading the following detailed description of the exemplary embodiments of the present disclosure, when taken in conjunction with the appended claims.
Further objects, features and advantages of the present disclosure will become apparent from the following detailed description taken in conjunction with the accompanying Figures showing illustrative embodiments of the present disclosure, in which:
Throughout the drawings, the same reference numerals and characters, unless otherwise stated, are used to denote like features, elements, components or portions of the illustrated embodiments. Moreover, while the present disclosure will now be described in detail with reference to the figures, it is done so in connection with the illustrative embodiments and is not limited by the particular embodiments illustrated in the figures and appended claims.
The exemplary system/device/apparatus/method/computer-accessible medium according to exemplary embodiments of the present disclosure can be used to achieve high resolution two-dimensional (“2D”) or three-dimensional (“3D”) temperature maps by utilizing a series of low-resolution thermal images, and can track information obtained from a positional and/or imaging sensor. Such exemplary techniques can facilitate the use of a low-cost thermal sensor or device to create temperature maps that can be greater than the native resolution of the sensor itself, and thus, can be a super-resolution technique. Such exemplary procedures can be important when making a thermal-imaging based medical diagnosis in the resource-poor regions, as this exemplary approach can facilitate the production of clinically useful thermal images without needing and maintaining an expensive high-end thermal camera. Because many low-resolution thermal sensors can be compact and light-weight, this exemplary approach can have the advantage of making portable or embedded thermal measurement systems. This exemplary technique may not be restricted to low-resolution sensors alone. When combined with a high-resolution thermal camera, the exemplary procedures can also produce improved image qualities including of better resolution and signal to noise ratio.
The thermal images can be displayed on devices such as mobile phones, tablets, laptops and other computers. Since health care professionals, and those who care for animals, do not routinely evaluate thermal images, the display device can be coupled with a software application to assist with image interpretation based on an image analysis procedure, and machine based learning of normal and abnormal heat patterns in humans and animals.
Described below is an exemplary system/device/apparatus/method/computer-accessible medium according to exemplary embodiments of the present disclosure that can facilitate a rapid high-resolution thermal imaging using low cost thermal sensors and scanning mechanisms. In an exemplary embodiment, a metallic mirror can reflect not only the visible light, but also an invisible thermal infrared light. One or multiple motorized rotating mirrors (e.g., a Galvo) can be positioned in front of a stationary thermal sensor. By rotating the mirror, or mirrors, e.g., in x/y axes, a spatial thermal image can be acquired. Because a Galvo mirror device can facilitate rapid scanning and precise control of the dwelling time, such method can obtain high resolution and large field-of-view scan within a short time.
According to another exemplary embodiment of the present disclosure, it is possible to use an exemplary multi-pixel, or array, thermal sensor to achieve high resolution scans. An array sensor can be rotated in the x, x/y or x/y/z, /y direction, or any combination thereof. Translation of the sensor can also be performed. The exemplary thermal sensor can be combined with the Galvo mirror(s), as described above. Because an array sensor can have a wide field-of-view, and coarser spatial sampling, a deconvolution method/procedure can be used to interpolate a refined image of high spatial resolution.
The Galvo mirror(s) has/have been conventionally used to scan lasers of visible or infrared range for the purpose of illumination. In an exemplary embodiment, the Galvo mirror(s), with the invisible thermal infrared light at the detection side can be used.
Exemplary embodiments of the exemplary system/device/apparatus can utilize low-cost hardware, thus making it possible to make hand-held high resolution thermal imaging systems coupled with a mobile phone or tablet. Exemplary embodiments can include a low cost mobile phone thermal imager attachment for home insulation and a medical thermal imaging system for fever or inflammatory disease detection.
An exemplary mobile-phone based thermal imaging device according to an exemplary embodiment of the present disclosure can be used to differentiate viral versus bacterial infections in clinical medicine. Clinicians can use this safe, noninvasive, and reliable thermal imaging device as a point of care screening tool to identify patients with likely bacterial infections, especially respiratory infections. This exemplary thermal imaging technique can analyze temperature patterns to predict the probability of bacterial infection. With a simple-to-use thermal camera, the clinician can take a thermal picture of the patient, and receive the interpretation (e.g., computer interpretation) of the likeliness of bacterial infection, facilitating the clinician to make a decision about the need for antibiotic treatment in a single office visit.
The exemplary system/device/apparatus/method/computer-accessible medium according to an exemplary embodiment of the present disclosure can reduce health care costs, improve rational use of antibiotics and minimize radiation exposure to the patient.
In another exemplary embodiment of the present disclosure, an exemplary thermal imaging device can take a thermal picture of the patient, and provide an answer of yes/no for likelihood of bacterial infection. For example, an interpretation of the probability of bacterial infection can be based on correlation with the gold standard of chest x-rays to diagnose pneumonia. According to further exemplary embodiments of the present disclosure, it is possible to provide certain technology to produce a low cost thermal camera facilitating wide spread distribution.
Thermal images can be commonly acquired by thermal cameras. Scientific-grade thermal cameras can be very expensive, currently costing between $5,000 and $50,000. Even a low-grade thermal camera, primarily designed for house insulation inspections (see e.g., Reference 1), can easily cost over $1,000 and may not be feasible for a wide deployment in resource-poor regions. Super low-cost thermal cameras (e.g., less than $100) became possible in recent years due to the emergence of miniature thermal sensors, and low-cost microcontrollers. (See, e.g., Reference 8). As described herein, the exemplary system/device/apparatus/method/computer-accessible medium according to an exemplary embodiment of the present disclosure can be or include a snap-on thermal imaging attachment for mobile phones. This exemplary attachment can contain or include a miniature thermal sensor, a compact and high speed scanning mechanism and a microcontroller within a compact snap-on enclosure that can work with all or most smart phones. A simple and easy to use/operate application (e.g., mobile phone application or app) can be used to acquire thermal images from the attachment and display the results as overlays to the phone's camera view. Advanced 3D thermal map displays can be possible by incorporating exemplary stereo-based surface reconstruction techniques. (See, e.g., Reference 5).
The exemplary system/device/apparatus/method/computer-accessible medium according to an exemplary embodiment of the present disclosure can have the following exemplary characteristics: (i) Instantaneous Use: the device can complete a clinically-usable thermal scan; (ii) Universal Applications: the attachment can be compatible with most smart phones (e.g., Android smartphones, Apple smartphones, Windows smartphones etc.); (iii) Informative Results: the app can facilitate overlaying the thermal images with camera images; and (iv) Easy-to-operate: a healthcare worker with minimal training can perform a quality scan.
A compact thermal scanner operated with an Android smartphone was developed and has been successfully tested to acquire thermal images from a healthy volunteer. An exemplary illustration of the exemplary thermal scanner attachment is shown in
Using the above shown system/device/apparatus/method/computer-accessible medium according to an exemplary embodiment of the present disclosure, thermal images were acquired from a healthy volunteer.
Scanning speeds to acquire images can be important. For example, each thermal image in
The use of the Galvo can obtain more precise scan control and high resolution thermal maps within seconds. Furthermore, there can be low-cost array thermal sensors (e.g., 16×4 and 8×8 pixel arrays) currently available. For the same image size, a 64-pixel array sensor can reduce the acquisition time below 1.5 seconds. The cost and form-factor of an array sensor can be close to a point sensor used in the prototype device; however, the angular resolution of a pixel in an array sensor can be inferior to the use of a point sensor. In order to use an array sensor with the Galvo, exemplary real-time data interpolation and de-convolution scheme(s) can be utilized to achieve both high speed and high resolution. Based on these exemplary techniques, a battery-powered thermal camera attachment shown in
The exemplary device/system/apparatus 335 can have dimensions of about 1×2×1 cm, and can capture high speed images (e.g., 7-10 frames per second low-resolution images). The exemplary device/system/apparatus 335 can (i) be Bluetooth enabled, (ii) have low power consumption (e.g., 50 hours of continuous imaging with 3 AA batteries), and (iii) be fully integrated with a smartphone (e.g., an Android, iPhone or Windows Phone smartphone), and can have integrated software with co-registered thermal and visual images.
The exemplary software can program the exemplary system/device/apparatus to passively receive the thermal infrared light radiated from a heated body (e.g., a human, living animal, stove or a hot engine). The electromagnetic (“EM”) radiation from this attachment can be limited to the Bluetooth signal used to communicate between the attachment and a smartphone. Because Bluetooth uses very low power EM radiation, it can be very safe for use in human, similar to numerous existing Bluetooth-enabled smartphones, headsets and laptops.
In order to obtain a wide field-of-view high-resolution thermal image in a portable or non-portable device, the following exemplary procedures can be used:
A reference image of the chest 405 can be used. (e.g., see
Thermal images of a given view or region-of-interest can be obtained from at least two positions (e.g., sensor positions) or orientations; this can be done by moving the thermal sensor through either a motorized motion, or an uncontrolled motion, such as a slow hand swipe or shaking. Examples of the uncontrolled motion can include: (i) the shaking produced at a hand-held position, (ii) a slow circular hand-rotation (e.g., a few seconds per cycle), (iii) slow reciprocal hand-twisting (e.g., a few seconds per cycle), (iv) an approximated linear translation and (v) a slow vibrational motion enabled through a spring-loaded sensor enclosure. The acquired infrared image is shown in
Both thermal images and the positional readings can be obtained from the thermal sensor and the positional sensor, respectively, and the positions/directions of the sensor can be changed (e.g., continuously changed). This can produce a multi-view thermal measurement around the target.
Using the positional information associated with each thermal reading, multi-view low-resolution images can be combined into a high-resolution thermal reading. This can be done by a multi-frame blind deconvolution or an optimization process. For example, a set of coordinates over the target space can be defined. For each position, a temperature reading that best fit, in the least square sense, the thermal measurements of all views can be solved for based on the light-of-sight based on the directional vectors obtained from the positional information.
An exemplary sensor can include a digital camera, and the position and orientation of the camera sensor can be automatically calibrated by an exemplary multi-view stereo (“MSV”) technique. Three dimensional thermal imaging can be facilitated when combining the multi-frame thermal images with the corresponding visual camera images. Applying the exemplary MSV procedure to the visual camera images, a 3D surface of the target and the surrounding structures can be calculated, and a temperature map distribution on this 3D surface can be solved for based on a least-square fitting of the thermal measurements, and the directional information can be restored from the exemplary MSV calculations.
The obtained 2D or 3D thermal images can be displayed on a monitor, facilitating the inspection and manipulation of the images. The scan directions, distance-to-target and the magnitude of the controlled/uncontrolled motion can be adjusted to obtain a refined thermal image that can be adapted based on the interpretation or findings from the initial scan.
In order to conduct the image analysis, the following exemplary procedures can be used. While the procedures below use the patient's chest as an example, it should be noted that the exemplary procedures can be used on any suitable body part or portion of the body.
Image registration and morphing can include the processes of mapping one image onto a reference image (e.g., either the exemplary thermal image or the exemplary visual image), while preserving the spatial information present in the original image. The first exemplary procedure for image registration and morphing can be the identification of the outline of the patient's body based on temperature differences between the patient and the external environment (see, e.g., procedure 1005 of
If the outline of the patient's body can be determined from the image, then at procedure 1010, using the exemplary system/apparatus/device/method/computer-accessible medium according to an exemplary embodiment of the present disclosure, it can be determined if certain important or critical elements needed for an eventual diagnosis are present. For example, for evaluation of the chest, the shoulders, the axilla and the likely location of the diaphragm should be present. The width of the chest can be determined as the distance between the shoulders and the length of the chest can be at least about 1.5 times the distance between the shoulders extending inferiorly from the level of the shoulders in the exemplary case of a chest. The determination of the sufficiency of the image can be made by locating readily identifiable landmarks in the image, such as the neck and shoulders, and then calculating, based on the size and position within the image of the neck and shoulders, if other key elements of the patient are present. If this information cannot be obtained from the image, the operator can be informed of the need to acquire additional images.
If the image includes the necessary elements of the patient, then, at procedure 1015, the image can be mapped onto a reference image. The mapping procedure can be performed by the application of a series of geometric transformations to the acquired image, which can align the acquired image with the reference image. The transformations can include translation, scaling and rotation around an axis. Exemplary techniques to determine the transformations can include determining the patient's center of mass in the image for translation, determining the patient's size in the image for scaling and determining the patient's skew in the image for rotation. Following these initial transformations, an iterative, hill-climbing technique can be used to find a local optimal alignment (e.g., procedure 1020). The exemplary process can produce a numerical correlation coefficient which can be used to accept or reject the image based on quality of alignment. If the correlation coefficient is not sufficiently high, the operator can be informed of the need to acquire additional images.
Following the exemplary image registration and morphing procedure above, the acquired infrared image can be optimally aligned to the limits of the geometric transforms employed (e.g. with respect to minimizing sum of squared error) to the reference image. However, the image may not be perfectly aligned. Additional heuristic processing can serve to improve the image alignment even further. Exemplary heuristics can include rules which rely on knowledge of human anatomy, and upon clues that can be derived from the thermal image. Exemplary clues can be typical hot-spots found around the neck and over the xiphoid process. Following the application of these exemplary heuristics, the image can be aligned further as shown in
When it can be determined that each patient has a sufficient number of images of acceptable quality for further analysis, the exemplary procedure can continue. Alternatively, instead of the all of the images acquired above, one anterior and posterior view can be used. If only one of the anterior or posterior view can be available, this information can be incorporated in the image analysis result. If images are still insufficient, the operator can be informed of the need to acquire additional images.
In order to spatially map the body, with the exemplary device/system/apparatus/method/computer-accessible medium according to an exemplary embodiment of the present disclosure, an exemplary image analysis area can be determined based on a location of the left chest and the right chest and a top and a bottom of each side of the left chest and the right chest (e.g., procedure 1025). Each pixel can have a precise location within each half of the chest. Anterior and posterior images can be aligned for analysis (e.g., image registration). The image can be partitioned (see, e.g., procedure 1030) into areas to be analyzed and areas to be excluded.
Following the exemplary alignment to the reference image, most or all pixels not including lung tissue can be removed from the image (see, e.g., procedure 1035). Known hot areas can be excluded based on the exemplary image library (e.g., axillae, neck and midline spine). Exemplary locations of these hot spots can be based on the initial outline of the image accepted for analysis.
The lung images from
The exemplary processing procedure shown in
Using only the hot-spot information from the previous procedure, the likelihood that “hot” areas can represent pathology can be determined. Areas that can be symmetrically “hot”, and that can be unlikely to be consistent with pathology, can be excluded. The exemplary image library can be used to determine symmetric structures and shapes inconsistent with pathology. Also, hot areas too small in size can be excluded as these can be likely to be due to random variation in temperature. After the hotspots are removed, the exemplary image can be generated (see, e.g., procedure 1050), and the exemplary image library can be used to determine minimum size limits. A probability of the image being consistent with bacterial infection based on asymmetry and/or temperature patterns in right and left chest can be determined. The procedure used to make a diagnosis can also be trainable based on the ever expanding image library. The exemplary conclusions from the exemplary system/device/apparatus/method/computer-accessible medium, according to an exemplary embodiment of the present disclosure, can be consistent with a positive or negative diagnosis of pneumonia. If inconclusive, different rules can be weighted to derive an overall diagnosis along with a confidence measure of the diagnosis. These exemplary weightings can all be determined from training data including, but not limited to, machine learning or exemplary neural network techniques.
In a healthy individual, it can be expected that the skin of the torso over the area of the lungs can be of a uniform temperature. A digital infrared image of a lung uniform in temperature, however, can contain noise and artifacts due to the imperfections of the measurement of infrared energy radiated from the healthy patient. A source of error can be quantization noise due to the conversion of an infinitely variable analog measurement into a discrete digital value. Quantization noise can be characterized as being of high spectral frequency as it can be relatively uncorrelated from one pixel to the next.
The exemplary diagnosis technique/procedure according to an exemplary embodiment of the present disclosure can utilize the assumption that an elevation of skin temperature can be detectable in the skin area adjacent to the area of infection in the lungs. This exemplary elevation in skin temperature can be due to the patient's inflammatory response to the infection. An assumption can also be that the area of inflammation can be large in magnitude relative to the quantization error, and is can be in size relative to the resolution of the detector. This can result in the area of inflammation being of low spectral frequency relative to the frequency of the noise.
The diagnosis of pneumonia from an x-ray, or from an infrared image, may likely only be made by a medical expert. The automated detection system described above applies this knowledge in an exemplary computer program/procedure. This can be achieved through as a rule-based system. Rules within the system can be fed from hot-spot information obtained from the image. Hot spots can be characterized by their size, intensity, shape and location within the lung. Rules can have the form of, for example, “if size>A and intensity>B and shape=elliptical and location=lower-left-lobe, then this can indicate a positive result for pneumonia with a likelihood of D %”. In this example, the values A, B, C, and D can be constants. In the event multiple rules apply to a set of observations, other rules can exist to form an overall diagnosis from the other rules that “fired”.
As described herein, the values of A, B, C and D can be determined from training images. An expert can diagnoses a number of training images as being positive or negative for pneumonia. Well-known optimization techniques can then be used to automatically adjust all constants of the exemplary system/method/computer-accessible medium such that the diagnosis generated from the exemplary system/method/computer-accessible medium for the training images can closely match the diagnosis made by the expert. The assumption can be that if the training set can be large enough, then the expert and the exemplary system/method/computer-accessible medium can derive the same diagnosis for new images not part of the original set of training images. The rules and parameters can be adjusted as new training images become available.
In addition to the above, multiple thermal sensors can be used simultaneously to achieve faster image acquisition. There can be a positional sensor associated with each thermal sensor or a single positional sensor for the entire thermal sensor group as long as the relative positions between the thermal sensors can be known. The 2D and 3D thermal maps can be acquired continuously so that the temperature variations over time at any given position can be restored. This can make it possible for real-time monitoring of temperature of the target. If the thermal sensor has a higher temporal resolution than the positional sensor readings, an interpolation between the adjacent positional readings can be assumed to correspond to the intermediate thermal image frames. A fiducial marker can also be used, which can be visible in both the thermal images and the digital camera images to make an automatic correspondence between the two sets of measurements.
As shown in
Further, the exemplary processing arrangement 1202 can be provided with or include an input/output arrangement 1214, which can include, for example a wired network, a wireless network, the internet, an intranet, a data collection probe, a sensor, etc. As shown in
The foregoing merely illustrates the principles of the disclosure. Various modifications and alterations to the described embodiments will be apparent to those skilled in the art in view of the teachings herein. It will thus be appreciated that those skilled in the art will be able to devise numerous systems, arrangements, and procedures which, although not explicitly shown or described herein, embody the principles of the disclosure and can be thus within the spirit and scope of the disclosure. Various different exemplary embodiments can be used together with one another, as well as interchangeably therewith, as should be understood by those having ordinary skill in the art. In addition, certain terms used in the present disclosure, including the specification, drawings and claims thereof, can be used synonymously in certain instances, including, but not limited to, for example, data and information. It should be understood that, while these words, and/or other words that can be synonymous to one another, can be used synonymously herein, that there can be instances when such words can be intended to not be used synonymously. Further, to the extent that the prior art knowledge has not been explicitly incorporated by reference herein above, it is explicitly incorporated herein in its entirety. All publications referenced are incorporated herein by reference in their entireties.
The following references are hereby incorporated by reference in their entirety.
This application relates to and claims priority from U.S. Patent Application No. 61/839,137, filed on Jun. 25, 2013, the entire disclosure of which is incorporated herein by reference.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2014/044125 | 6/25/2014 | WO | 00 |
| Number | Date | Country | |
|---|---|---|---|
| 61839137 | Jun 2013 | US |
