Systems, Apparatus, and Methods for Image Interpretation - Patent application
A portable imaging device converts fluorescent wound images to HSV color space and applies threshold conditions to highlight bacterial regions, addressing the lack of sophisticated imaging in wound care and enhancing bacterial detection accuracy.
Patent Information
- Application Number
- JP2024531066
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-11-24
- Filing Date
- 2022-11-23
- Publication Date
- 2024-12-03
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Current wound care practices lack sophisticated imaging modalities for accurately identifying bacterial presence and distribution in wounds, leading to suboptimal treatment and potential misdiagnosis due to the invisibility of bacteria to the naked eye and interference from non-bacterial sources.
A portable handheld device with an imaging system that converts fluorescent images to HSV color space, applies threshold conditions based on bacterial fluorescence signatures, and generates an overlay image to highlight bacterial regions, utilizing color and texture analysis for precise bacterial identification.
Enhances the ability to accurately detect bacterial presence and distribution in wounds, reducing treatment errors and improving wound care outcomes by providing real-time, point-of-care bacterial identification.
Smart Images

Figure 2024544602000001_ABST
Abstract
Description
[Technical field]
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Application No. 63 / 282,894, filed with the U.S. Patent and Trademark Office on November 24, 2021, the entire contents of which are incorporated herein by reference.
[0002] Technical Field Systems, devices, and methods for image interpretation are disclosed. The images may be various types of images, including fluorescent images. In particular, the systems, devices, and methods may utilize, interpret, and / or modify images acquired by an imaging system to identify or aid in identifying characteristics of interest associated with a target in the image. In various applications, for example, the target may be a wound, and the systems, devices, and methods herein may be used to identify, quantify, and / or differentiate bacteria present in the wound. [Background technology]
[0003] background Wound care is a major clinical challenge. Healing and chronic non-healing wounds are associated with many biological tissue changes including inflammation, necrosis, exudate production, bleeding, proliferation, connective tissue remodeling, and the presence, proliferation, and infection of bacteria, which are common major concerns. Some wound infections are not clinically apparent, contributing to the increased personal, emotional, and economic burden associated with wound care, especially in the elderly population. For example, Pseudomonas aeruginosa and Staphylococcus aureus are genera of bacteria that are common in the hospital environment and are common causes of bacterial infections. Currently, the clinical gold standard for wound assessment involves direct visual inspection of the wound under white light illumination for typical signs and symptoms of infection. This is often combined with swab cultures or tissue biopsy samples for laboratory testing.
[0004] Certain medical specialties (e.g., cardiology, oncology, neurology, orthopedics, etc.) rely on specific imaging modalities (e.g., x-ray, ultrasound, magnetic resonance imaging (MRI), computed tomography (CT) scans, etc.) to aid in diagnosis and evaluation. Clinicians in such specialties may use advanced and established methods of interpreting images. In contrast, in wound care specialties, the standard of care has historically not relied on such imaging modalities, and no such advanced or established methods of interpreting images exist. Some clinicians may use cameras to capture images of wounds in standard photographic formats, but these formats do not identify or present any bacterial information within the wound.
[0005] Qualitative and subjective visual assessments only provide a global picture of the wound site and do not provide information about the underlying biological, biochemical, and molecular changes occurring at the tissue and cellular levels. Furthermore, bacteria are not visible to the naked eye, resulting in suboptimal wound sampling and an inability to properly track changes in bacterial growth in the wound. This can hinder healing and timely selection of optimal antimicrobial treatments. Furthermore, it can be difficult to distinguish certain markers of bacterial presence from similar markers caused by non-bacterial sources. For example, fluorescent images can contain reflections from non-bacterial sources (e.g., tattoos, fingernails, toenails, jewelry, background environment, etc.) that appear similar in color to the fluorescence expected from a particular bacterial strain. These situations can result in the misidentification of a wound as containing bacteria when it does not (possibly resulting in medically unnecessary treatment) and / or the misidentification of a wound as not containing bacteria when it is actually infected (possibly resulting in severe infection spread, reduced quality of life, loss of digits, etc.).
[0006] Thus, there is a need for systems, devices and methods for the analysis, interpretation and / or correction of medical images (such as fluorescent images of wounds or skin) that can reliably indicate the presence and location of areas potentially having bacterial levels above certain clinically significant thresholds, thereby enabling better treatment and reducing both morbidity and mortality resulting from elevated bacterial burdens, particularly those associated with chronic wounds. Summary of the Invention [Means for solving the problem]
[0007] overview In consideration of these and other circumstances, the present disclosure provides methods, systems, and devices that provide point-of-care interpretation of images to aid in identifying regions of interest within the images. In one example, the disclosed methods, systems, and devices aid in point-of-care interpretation of fluorescent images of tissue, such as wounds, that aid in identifying regions of bacteria within the images by highlighting the regions of bacteria on the images to bring them to the attention of a user / clinician.
[0008] In one aspect of the disclosure, a portable handheld device is provided, the portable handheld device comprising an imaging device configured to receive a fluorescent image, a memory, and a processor configured to: convert the fluorescent image into an HSV color space, thereby generating a first converted image; and compare a saturation parameter of the first converted image to a first saturation threshold condition for a first range of hue parameters of the first converted image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescent signature; and a processor configured to perform a color analysis of the fluorescent image including comparing a value parameter of the first color to a first value threshold condition, the first value threshold condition being based on a correlation between the first color and a first bacterial fluorescence signature; identifying regions of the fluorescent image where, for a first range of the hue parameter, the saturation parameter satisfies the first saturation threshold condition and the first value parameter satisfies the value threshold parameter; discarding identified regions that are smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions on the fluorescent image.
[0009] In another aspect of the disclosure, a system is provided, the system including a display device, an imaging device, the imaging device including an excitation light source configured to irradiate a target area with excitation light and an image sensor configured to capture a fluorescence image responsive to a response of the target area to the excitation light, a housing, and a processing circuit disposed within the housing, the processing circuit including a processor, the processor converting the fluorescence image into an HSV color space, thereby generating a first converted image, and comparing a saturation parameter of the first converted image to a first saturation threshold condition for a first range of hue parameters of the first converted image, the first saturation threshold condition being a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature. and comparing a value parameter of the first transformed image to a first value threshold condition for a first range of hue parameters, the first value threshold condition being based on a correlation between the first color and the first bacterial fluorescence signature; and identifying regions of the fluorescent image where the saturation parameter satisfies the first saturation threshold condition and the first value parameter satisfies the value threshold parameter for the first range of hue parameters; discarding identified regions smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions on the fluorescent image.
[0010] In yet another aspect of the present disclosure, a method of fluorescence image interpretation is provided, the method including: receiving a fluorescence image; converting the fluorescence image to an HSV color space, thereby generating a first transformed image; comparing a saturation parameter of the first transformed image to a first saturation threshold condition, for a first range of hue parameters of the first transformed image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature; comparing a value parameter of the first transformed image to a first value threshold condition, for the first range of hue parameters, the first value threshold condition being based on a correlation between the first color and the first bacterial fluorescence signature; identifying regions of the fluorescence image where the saturation parameter satisfies the first saturation threshold condition and where the first value parameter satisfies the value threshold parameter, for the first range of hue parameters; discarding the identified regions smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions on the fluorescence image.
[0011] In yet another aspect of the present disclosure, a portable handheld device is provided, the portable handheld device comprising: an imaging device configured to receive a fluorescence image; a memory; and a processor configured to perform texture analysis of the fluorescence image, including generating a first transformed image by converting the fluorescence image to grayscale; comparing, for each pixel, an intensity parameter of the first transformed image to an intensity threshold; temporarily identifying regions within a predetermined pixel distance of pixels whose intensity parameter exceeds the intensity threshold; generating a second transformed image by converting the fluorescence image to a predetermined color space; determining, in each channel of the second transformed image, respective gradients of the temporarily identified regions; and permanently identifying one of the temporarily identified regions where the respective gradient of each channel of the second transformed image exceeds the gradient threshold.
[0012] In yet another aspect of the present disclosure, a method is provided for interpreting clinical images, the method including receiving, on a handheld imaging device, an input image; performing color and / or texture analysis on the input image to identify areas of clinical interest in the input image; generating an overlay, the overlay including one or more colors, highlighting, arrows, or other highlighting methods that, when placed on the input image, provide an indication of one or more clinical features of interest; and displaying, in real time, an output image including the input image and the overlay, where the clinical features of interest on the input image are highlighted by the overlay to a user of the imaging device.
[0013] In yet another aspect of the present disclosure, a portable handheld device is provided, the portable handheld device comprising: an imaging device configured to receive an input image; a memory; and a processor configured to convert the input image into a predetermined color space including a first channel corresponding to a first parameter, a second channel corresponding to a second parameter, and a third channel corresponding to a third parameter, thereby generating a converted image; comparing a second parameter of the first converted image to a threshold condition for a first range of the first parameter of the converted image, the first threshold condition being based on a correlation between the first range of the first parameter and a bacterial signature; and a processor configured to perform a color analysis of the input image including comparing, for a first range of the one parameter, a third parameter of the transformed image to a second threshold condition, the second threshold condition being based on a correlation between the first range of the first parameter and a bacterial signature; identifying, for the first range of the first parameter, regions of the input image where the second parameter satisfies the first threshold condition and the third parameter satisfies the second threshold condition; discarding the identified regions that are smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions of the input image.
[0014] Aspects of the present disclosure thus provide improvements in at least the art of fluorescence imaging, as well as related arts such as image processing, medical instruments, biophotonics, wound treatment, etc. Additional aspects of the present disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the present disclosure. The aspects of the present disclosure and advantages arising therefrom may be realized and attained by means of the elements and combinations particularly pointed out in the appended claims and their equivalents.
[0015] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the scope of the disclosure and claims.
[0016] BRIEF DESCRIPTION OF THE DRAWINGS These and other aspects of the present disclosure will be described with reference to the accompanying drawings. [Brief description of the drawings]
[0017] [Figure 1A] 1 illustrates an exemplary apparatus according to various aspects of the present disclosure. [Figure 1B] 1 illustrates an exemplary apparatus according to various aspects of the present disclosure. [Diagram 2] 1 illustrates another exemplary apparatus according to various aspects of the present disclosure. [Diagram 3] 1 shows an exemplary schematic diagram of an apparatus according to various aspects of the present disclosure. [Figure 4A] 1 illustrates an exemplary process flow according to various aspects of the present disclosure. [Figure 4B] 1 illustrates an exemplary process flow according to various aspects of the present disclosure. [Figure 5A] 1 illustrates an exemplary image according to various aspects of the present disclosure. [Figure 5B] 1 illustrates an exemplary image according to various aspects of the present disclosure. [Figure 6A] 1 illustrates an exemplary image according to various aspects of the present disclosure. [Figure 6B] 1 illustrates an exemplary image according to various aspects of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0018] Detailed Description Reference will now be made in detail to various exemplary embodiments, examples of which are illustrated in the accompanying drawings. The various exemplary embodiments are not intended to limit the present disclosure. On the contrary, the present disclosure is intended to cover alternatives, modifications, and equivalents of the exemplary embodiments. In the drawings and the description, like elements are numbered like. It should be noted that features individually described herein can be combined with each other in any technically appropriate manner to disclose additional embodiments of the present disclosure.
[0019] The present disclosure provides devices, systems, and computer-implemented methods that provide automated image interpretation to provide guidance for clinicians reviewing photographs. For example, when a fluorescent image is captured and analyzed, areas of bacterial fluorescence are identified and shown on the image, for example by outlining areas of suspected infection. In some embodiments, this is accomplished by detecting color shades (e.g., red, yellow, and cyan) within the image that are shown to be indicative of bacteria. Image interpretation elements and operations analyze the image for one or more bacterial fluorescent signals (e.g., for both red and cyan bacterial fluorescent signals).
[0020] Although the present disclosure has been described primarily in terms of identifying wound components and features captured in fluorescent images, the devices, systems, and methods disclosed herein can also be used to identify features of any target surface or object having a distinguishable color or texture profile that can be correlated to the condition and / or presence, location, and / or abundance of microorganisms such as bacteria. For example, the present disclosure can be used to identify features of resected tissue, such as cancer tissue (e.g., lumpectomy for best cancer surgery) captured in fluorescent, white light, or other images. When used with resected tissue, the devices and methods can be used to identify features such as, for example, tissue components, tumor size, tumor edges, tumor borders, and tissue vascularization shown in the fluorescent images. Furthermore, while the present disclosure has been described primarily in terms of specific wavelengths of fluorescence and corresponding or associated bacterial species, etc., the disclosure is not so limited. In actual implementations, the devices, systems, and methods herein can be recalibrated for any bacterial species based on the unique fluorescent emission signature of a given wavelength or wavelength range for the bacterial species, and can be used with other tissue components (normal or abnormal) for which unique fluorescent signatures are known, as well. The present disclosure may further be implemented with respect to non-bacterial signatures, including, but not limited to, viruses, fungi, yeast, and / or other microorganisms. The devices, systems, and methods disclosed herein are not limited to use with fluorescent images, but may instead be used to analyze other types of images, such as, for example, white light images. The above recalibration or additional implementations may be accomplished with little or no modification to the remainder of the methods, systems, and devices described herein, by selecting specific thresholds depending on the particular implementation.
[0021] Exemplary wound monitoring devices described herein include handheld / portable optical digital imaging devices having specific excitation light sources and optical bandpass filters attached thereto. These devices include, but are not limited to, those described in International Patent Application Publication No. WO 2009 / 140757, International Patent Application Publication No. WO 2019 / 148268, International Patent Application Publication No. WO 2020 / 0148725, and / or International Patent Application Publication No. WO 2020 / 0148726, the entire contents of each of which are incorporated herein by reference in their entirety. The imaging devices and systems further described herein can be used to image and analyze the fluorescence of components in a wound due to exposure to excitation light. For example, in wounds caused by or having the presence of bacteria including Pseudomonas aeruginosa, Pseudomonas aeruginosa fluoresces at one or more bands of wavelengths with specific spectral signatures, i.e., known peaks, when subjected to excitation light. The excitation light may include any light of a known wavelength, such as a peak at 405 nm, or a range of wavelengths having a known peak. Capturing and analyzing this data allows for identification of the presence of bacteria generally, as well as the presence, location, and / or abundance of specific types of bacteria. Devices and systems are trained to identify, type, and quantify the presence of bacteria as well as additional characteristics of the wound.
[0022] One example of a wound monitoring device is a portable handheld imaging system that includes an imaging device having two or more cameras (i.e., camera sensors) and a processor coupled to the imaging device for analyzing images captured from the camera sensors to perform algorithmic or other operations, as described in more detail below. The imaging device includes, for example, a first primary camera sensor and a second secondary camera sensor. The first primary camera sensor and the second secondary camera sensor may be configured to capture standard white light (WL), fluorescent (FL), near infrared (NIR), or infrared (IR) images. The sensor may be so configured by use with a dedicated filter selectable from a plurality of filters (e.g., filter wheel, tunable filter, etc.) associated with the imaging device. Thus, the methods disclosed herein may be used to measure features captured in WL, FL, NIR, or IR images. In some implementations, the first camera sensor is separated from the second camera sensor by a predetermined fixed separation distance to enable determination of disparity values of the primary and secondary images (taken by the primary and secondary camera sensors, respectively).
[0023] 1A-1B illustrate an exemplary wound monitoring device 100 according to the present disclosure, with FIG. 1A being a front perspective view and FIG. 1B being a rear perspective view of the wound monitoring device 100. The device 100 may be, for example, a MolecuLight DX® device developed by MolecuLight®. The device 100 is non-contact and no imaging contrast agent is required for white light and / or fluorescent imaging. As shown, the wound monitoring device 100 includes a display device 110, an imaging device 120, and a housing 130.
[0024] The imaging device 120 has at least one image sensor, such as, for example, a first image sensor 121, a second image sensor 122, and a third image sensor 123. Each of the first image sensor 121, the second image sensor 122, and the third image sensor 123 may be individually implemented as an image sensor that can be used for one or more of WL, FL, IR, and thermal imaging. In one example, the first image sensor 121 and the third image sensor 123 are configured together for stereoscopic white light imaging, and the second image sensor 122 is configured for fluorescent imaging. In another example, the first image sensor 121 and the third image sensor 123 are configured together for stereoscopic fluorescent imaging, and the second image sensor 122 is configured for white light imaging. In yet another example, the first image sensor 121 is configured for white light imaging, the second image sensor 122 is configured for fluorescent imaging at a first wavelength or wavelength range, and the third image sensor 123 is configured for fluorescent imaging at a second wavelength or wavelength range. The physical arrangement (i.e., ordering) of the first image sensor 121, the second image sensor 122, and the third image sensor 123 may also differ from that shown in FIG. 1B. For ease of illustration, the wound monitoring device 100 shows three image sensors, but the present disclosure includes an imaging device 120 including any number of image sensors, so long as there is at least one image sensor capable of receiving a fluorescent signal. In the illustration of FIG. 2, the first image sensor 121, the second image sensor 122, and the third image sensor 123 are disposed within an optical battery 124. The optical battery 124 may further include one or more excitation light sources, such as an excitation light source 125 implemented as a light-emitting diode (LED). The optical battery 124 may further include at least one white light torch, a range finder, a temperature sensor, and the like.
[0025] The housing 130 may include a physical user interface, such as a power button, one or more input buttons. The housing 130 may also include various input and output ports, such as a wired charging port, an inductive charging port, a universal serial bus (USB) port, and / or other peripheral ports. As shown in FIG. 1B, the housing 130 includes an optical mount 131 to which the imaging device 120 is attached. The optical mount 131 may provide a permanent physical mount or may provide a removable mount in implementations in which the imaging device 120 is modular (i.e., may be replaced with other imaging devices, such as an endoscopic head). The housing 130 may be of a unitary construction or may include multiple housing components attached to each other. Although not specifically shown in FIGS. 1A-1B, the housing 130 may include an interior space in which various components, such as a processor (e.g., a central processing unit (CPU)), memory, program storage, etc., are located.
[0026] FIG. 2 shows four views (front, right side, back, and left side elevation views) of another exemplary wound monitoring device 200 according to the present disclosure. The device 200 can be, for example, a MolecuLight i:X® device developed by MolecuLight®. The device 200 allows clinicians to quickly, safely, and easily visualize the presence and distribution of bacteria in skin and wounds in real time, including but not limited to at the point of care. The device 200 is non-contact and no imaging contrast agents are required for white light and / or fluorescent imaging. The device 200 is shown as a handheld portable medical device comprised of a high-resolution color LCD display and a touch-sensitive screen 208 with integrated optics and microelectronics and an internal battery power source. The device 200 further includes a power button 201 for turning the device on and off, a display screen power button 202 for turning the display screen 208 on and off, a system status LED 203 for indicating overall device performance, a battery status LED 204 for indicating the device's battery charge, a range finder LED system 205 for indicating the optimal distance from the wound being targeted or imaged, an ambient light status LED 206 for indicating the optimal lighting environment for fluorescence mode imaging, a heat sink 207 for dissipating heat when the device 200 may become hot after extended use, a home button 209 for providing access to the image and video capture functions of the device 200, and a port 210 for charging and data transfer. The port 210 can be used with any generic or dedicated cable, such as a USB or MolecuLight i:X® connection cable.
[0027] The device 200 further includes a rocker switch 211 that allows switching between a standard imaging mode and a fluorescent imaging mode. For example, the device 200 captures real-time images (e.g., in JPG format) and videos (e.g., in MOV format) using both the standard imaging mode and the fluorescent imaging mode. The standard imaging mode is generally used for standard photography, i.e., capturing RGB images and videos of a target illuminated with standard white light. The fluorescent imaging mode is used for capturing RGB images and videos of a target illuminated with light having a known peak wavelength and intended to generate fluorescence from the specific target being excited by the light. Thus, the device 200 further includes an LED 212 with a specific wavelength or range of wavelengths for illuminating the target when in the fluorescent imaging mode, a camera lens 213 that allows image and video capture, a range finder sensor 214 for detecting the optimal distance from the wound or surrounding skin, and an ambient light sensor 215 for detecting the optimal lighting conditions for the fluorescent imaging mode. Additionally, device 200 includes a holding contour 217 to allow a user to firmly grip the device, and a charging port 218 to allow charging of the device using a standard or proprietary power adapter.
[0028] Figure 3 shows an exemplary schematic diagram of a wound monitoring device 300 according to the present disclosure, which may be the same as or similar to the devices 100 or 200 shown in Figures 1-2, or may be a different device. As shown in Figure 3, the wound monitoring device 300 includes a fluorescent image sensor 310, an excitation light source 320, a controller 330, a display device 340, a memory 350, a user interface (UI) 360, a power source 370, communication processing circuitry 380, and input / output (I / O) processing circuitry 390.
[0029] The fluorescence image sensor 310 may be implemented as or include a photoelectric conversion device that receives incident electromagnetic radiation (e.g., light) and converts the radiation into signal charges that can be used to generate an image of the image sensor's field of view. Such an image sensor may have, for example, a complementary metal-oxide semiconductor (CMOS) architecture, a charge-coupled device (CCD) architecture, or the like. The photoelectric conversion device may include an array of a plurality of individual pixel circuits, each of which includes a light-receiving element such as a photodiode. The image sensor may include additional processing circuitry, such as drive circuitry, timing circuitry, memory circuitry, control circuitry, output circuitry, power supply circuitry, and buses.
[0030] The fluorescence image sensor 310 may include a filter or multiple filters to block certain wavelengths of light and pass other wavelengths of light to the fluorescence image sensor 310. In implementations where multiple filters are present, the filters may be arranged in a device that allows individual filters to be individually selected and moved in front of the fluorescence image sensor 310 (e.g., a filter wheel) to selectively detect specific light signals based on the wavelength of light. The fluorescence image sensor 310 may be part of a digital camera, for example, with at least ISO 800 sensitivity, but more preferably with ISO 3200 sensitivity, and may be combined with one or more emission filters or other equally effective (e.g., miniaturized) mechanized spectral filtering mechanisms (e.g., acousto-optical or liquid crystal tunable filters). Each optical filter may have a different discrete spectral bandwidth and may be a bandpass filter, a multiband filter, a shortpass filter, or a longpass filter. In one example, the fluorescence image sensor 310 includes a first filter having a passband of 600-660 nm and a second filter having a passband of 470-520 nm or 500-550 nm. The fluorescence image sensor 310 can include a lens that is aligned to point at a target area (e.g., the area being imaged) and can detect optical signals emanating from the target area or surface.
[0031] The excitation light source 320 may be or include one or more light emitting elements that generate excitation light or illumination, e.g., monochromatic or white light with a wavelength peak of 400-450 nm, or any other combination of single or multiple wavelengths (e.g., wavelengths in the ultraviolet / visible / near infrared / infrared range), to illuminate a target object (e.g., a wound or other area of interest) to induce a light signal (e.g., fluorescence). For example, the excitation light source 320 may be a blue or violet LED array emitting light at about 405 nm (e.g., ±5 nm) and may be coupled with an additional bandpass filter centered at about 405 nm to remove / minimize side spectral bands of light from the LED array output so as not to cause light leakage to an imaging detector that has its own optical filter. The excitation light source 320 may additionally or alternatively include laser diodes and / or filtered light arranged in various geometries. The apparatus 300 may include methods or devices (e.g., heat sinks or cooling fans) for dissipating heat and cooling the excitation light source 320. The apparatus 300 may include a system or device (e.g., an optical bandpass filter) for removing any undesirable wavelengths of light from the excitation light source 320 used to illuminate the object being imaged.
[0032] The controller 330 may be one or more computer processors (processing units), electronic processors such as microprocessors, digital signal processors, controllers and microcontrollers. Logic may be formed from signals stored in a computer-readable medium such as memory 350, which in an exemplary embodiment may be random access memory (RAM), read-only memory (ROM), erasable / electrically erasable programmable read-only memories (EPROMS / EEPROMS), flash memory, etc. Logic may also include digital and / or analog hardware circuits, e.g., hardware circuits including logical AND, OR, XOR, NAND, NOR, and other logical operations. Logic may be formed from a combination of software and hardware. On a network, logic may be programmed on a server or a complex of servers. A particular logical unit is not limited to a single logical location on the network. Furthermore, the modules need not be executed in a particular order. Each module may call another module when it needs to be executed.
[0033] The display device 340 may be any type of display, including but not limited to a liquid crystal display (LCD), a quantum dot display, an organic light emitting display (OLED), a thin-film transistor (TFT) display, etc. The display device 340 may be configured to provide a real-time display of the field of view of the fluorescence image sensor 310, display images stored in the memory 350 or stored remotely, provide graphic overlays, etc. The display device 340 may include a touch panel as a user interface to allow input from a clinician. The display device 340 may be configured to present one or more graphical user interfaces (GUIs).
[0034] The UI 360 may include physical and virtual mechanisms for a user to interact with the device 300, including physical buttons, dials, switches, etc., interactive icons displayed on the display device 340, and / or one or more GUIs. The power source 370 may be an AC / DC power source, a small battery bank, or a rechargeable battery pack. Additionally or alternatively, the device 300 may be adapted to connect to an external power source. The communication processing circuitry 380 may include wired and / or wireless communication processing circuitry for enabling communication between the device 300 and an external device. The communication processing circuitry 380 may be configured to communicate using a wired communication medium, such as a coaxial cable, a USB cable, an Ethernet cable, and / or using a wireless communication protocol, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 4G cellular communication, 5G wireless communication, etc.
[0035] The I / O processing circuitry 390 may be or include one or more of the following: an interface for a head mounted display, an interface for an external printer, an interface for a tablet computer, laptop computer, desktop computer or other computing device, an interface for a device that allows for the use of additional memory, an interface for a microphone. The device 300 may have a housing that houses all the components in one entity. The housing may include a means for securing any digital imaging device therein. The housing may be designed to be handheld, compact, and / or portable. The housing may be one or more enclosures or a separate housing.
[0036] The device 300 can be configured with application programs (e.g., stored in the memory 350 and loaded for execution by the controller 330) for executing various algorithms, including those that provide image interpretation of the fluorescent image. For example, the processor may be configured to "interpret" the fluorescent image by, for example, analyzing the hue, saturation, and value of the colors present in the fluorescent image and, based on the analysis, identifying areas of the image that correspond to one or more areas of bacterial presence, where the bacterial presence and / or abundance may be based on a particular quantitative threshold, or a spectral signature exceeding a particular type of bacteria. The analysis may include, for example, comparing the HSV parameters of the image to threshold parameters, where the threshold is set to remove pixels that are suspicious or non-suspicious or contain a threshold amount of bacteria. The processor may then create an overlay placed on the original image to highlight these areas of the original image and indicate that the clinician should consider these areas suspicious (e.g., as containing a particular level of bacteria and / or a particular type of bacteria) or should consider physical follow-up, such as debridement or swabbing.
[0037] If color analysis alone is not sufficient to identify a particular region of interest within the fluorescent image, a texture analysis of the image can also be performed, as described below. This analysis may be performed in parallel or sequentially with the color analysis. For example, texture analysis may help identify regions of the image where the bacterial signature of interest includes a non-primary color, such as cyan or blush (pink). This analysis, in combination with the color analysis, allows the processor to identify specific regions in the fluorescent image that exhibit a particular bacterial fluorescent signature.
[0038] In this manner, the disclosed apparatus and methods aid the clinician in interpreting fluorescence images of a wound.
[0039] According to one exemplary embodiment, the processor 330 may be configured to perform a color analysis of the fluorescent image, including converting the fluorescent image to an HSV color space to generate a first converted image, comparing a saturation parameter of the first converted image to a first saturation threshold condition, the first saturation threshold condition being based on a correlation between the first color and the first bacterial fluorescence signature, comparing a value parameter of the first converted image to a first value threshold condition, the first value threshold condition being based on a correlation between the first color and the first bacterial fluorescence signature, identifying regions of the fluorescent image in which the saturation parameter satisfies the first saturation threshold condition (e.g., by flagging corresponding data), discarding identified regions that are smaller than a size threshold (e.g., by unflaggling data corresponding to the flagged regions), and generating an overlay image by outlining the remaining identified regions on the fluorescent image. The color analysis may be repeated for multiple colors, each with its own saturation and / or value threshold conditions, based on the correlation between the individual colors and various bacterial fluorescence signatures. In other examples of the present disclosure, color spaces other than HSV may be used for the color analysis, including, but not limited to, HSL, CIEXYZ, CIELUV, L*a*b*, RGB, YCbCr, YUV, LCh, CMY, CMYK, and custom color spaces. When using color spaces other than HSV, the threshold conditions (e.g., hue threshold, saturation threshold, and / or value threshold) may be directly converted to the other color space. Note that if the color space used in the color analysis is the same as the color space in which the input image was captured (e.g., both are RGB), the conversion operation may be omitted.
[0040] The processor 330 may be configured to perform texture analysis of the fluorescence image, including converting the fluorescence image to grayscale to generate a second converted image, comparing, for each pixel, an intensity parameter of the second converted image to an intensity threshold, temporarily identifying regions within a predetermined pixel distance of the pixels whose intensity parameter exceeds the intensity threshold, generating a third converted image by converting the fluorescence image to an L*a*b* color space, determining, in each channel of the third converted image, the gradient of each of the temporarily identified regions, and permanently identifying one of the temporarily identified regions whose gradient of each channel of the third converted image exceeds the gradient threshold. In other examples of the present disclosure, color spaces other than L*a*b* may be used for the texture analysis, including, but not limited to, HSV, CIEXYZ, CIELUV, RGB, YCbCr, YUV, LCh, CMY, CMYK, and custom color spaces. When a color space other than L*a*b* is used, the thresholds (e.g., intensity threshold and / or gradient threshold) may be newly selected and / or generated based on characteristics of the other color space. If a color space other than HSV is used, the threshold conditions (e.g., hue threshold, saturation threshold, and / or value threshold) may be directly converted to the other color space. Note that if the color space used in the texture analysis is the same as the color space in which the input image was captured (e.g., both are RGB), the conversion operations may be omitted. If the color space used in the texture analysis is the same as the color space used in the color analysis, one of the conversion operations may be omitted and / or the two conversion operations may be combined (depending on the order in which the color analysis and texture analysis are performed and / or whether the analyses are performed sequentially or in parallel).
[0041] Texture analysis can be used to distinguish between bacterial and non-bacterial sources of apparent fluorescence (e.g., bone, tendon, tattoo ink, etc.). 4 It can be used to identify individual bacterial signatures that correspond to bacterial concentrations of colony forming units / gram (cfu / g) or greater, and may indicate the presence of any one of the following:
[0042] [Table 1]
[0043] The above list is non-limiting and exemplary only, and is not inclusive of all species to which the disclosed apparatus and methods may be applied. 4 In addition to identifying bacteria present at concentrations of cfu / g or greater, the image analysis of the present disclosure may be used to identify other characteristics that a clinician may consider useful or relevant to diagnosis, so long as there is some perceptible difference in particular color, intensity, or texture features within the image that may form the basis for selecting one or more thresholds.
[0044] FIG. 4A illustrates an example of operations according to the present disclosure. The operations of FIG. 4A include color and texture threshold analysis and may be performed for a single color. The operations include receiving an input image at 410A. The input image may be a still image retrieved from memory or captured by an image sensor, a video image retrieved from memory and / or captured by an image sensor, and / or a real-time video stream from an image sensor. The input image may be a fluorescent image capture capturing a response of a target area (e.g., a wound) exposed to an excitation light. FIG. 4A illustrates two thresholding operations performed in parallel on a target color: operation 420A performing a fluorescent color thresholding and operation 430A performing a fluorescent texture thresholding. Operations 420A and 430A are not limited to being performed in parallel and may be performed sequentially in any order. For some implementations (e.g., for a particular color), only operation 420A or only operation 430A may be performed. The color and / or texture threshold analysis of FIG. 4A may be performed on multiple colors, sequentially or in parallel. For example, operation 420A may be performed only for a first color, then both operations 420A and 420A may be performed for a second color, and so on.
[0045] Act 420A includes act 421A, which converts an input image (which may be captured, for example, in red-green-blue (RGB) color space) to HSV color space, thereby generating a converted image. The HSV color space includes three channels: hue (H), saturation (S), and value (V). The H parameter is expressed in degrees from 0° to 360°, and the S and V parameters are expressed as percentages from 0% to 100%. To convert from the RGB color space to the HSV color space, for example, a conversion algorithm may be applied to the input image on a pixel-by-pixel basis. The conversion algorithm may include normalizing each of the R, G, and B pixel values to one, calculating the hue using a formula that depends on whether the normalized R, G, or B parameter is maximum, calculating the saturation based on the difference between the maximum normalized R, G, and B parameter and the minimum and based on the value of the maximum normalized R, G, and B parameter, and calculating a value based on the maximum normalized R, G, and B parameter.
[0046] Operation 420A includes operation 422A, which compares each of the H, S, and V channels of the transformed image to a threshold condition. Pixels that meet the threshold condition for each channel may be identified as candidate pixels. In the following example, pixels that may not be "true" color pixels are eliminated by thresholding. This may be done, for example, to identify pixels that correspond to bacteria having a particular concentration or to limit pixels to those that correspond to a particular bacterial species. The thresholding operation may be done for all pixels in each successive channel (i.e., all pixels for H, all pixels for S, all pixels for V), or for all channels in each successive pixel (i.e., H, S, and V for the top left pixel, followed by H, S, and V for the right adjacent pixel, and so on). In one example of operation 422A, the hue threshold condition may be that pixels with an H parameter within 20° of the target color are considered "color candidate" pixels, the saturation threshold condition may be that pixels with an S parameter greater than 75% are considered "color candidate" pixels, and the value threshold condition may be that pixels with a V parameter greater than 75% are considered "color candidate" pixels. Operation 422A may also include identifying regions of the transformed image (or the original fluorescent image) where each channel satisfies a corresponding threshold condition (e.g., by adding or updating data flags). Thus, pixels that are color candidates in each of the three channels may be considered "target color" pixels.
[0047] Act 430A includes act 431A, which converts the input image to grayscale, thereby generating a converted image. To convert from an RGB color space to grayscale, for example, a conversion algorithm may be applied to the input image on a pixel-by-pixel basis. The conversion algorithm may include determining a simple average of the R, G, and B components, a weighted average of the R, G, and B components, etc. Thus, the converted grayscale image may represent the luminosity or intensity of the input image without regard to color. After act 431A, at 432A, the converted grayscale image may be compared to an intensity threshold on a pixel-by-pixel basis. Pixels corresponding to intensities greater than the intensity threshold may be temporarily identified as "texture candidate" pixels (e.g., identified until act 430A is terminated). In some implementations, all pixels within a predetermined number of temporarily identified pixels may be temporarily identified as well, thereby generating a "texture candidate" pixel region.
[0048] In operation 433A, the input image is transformed into the L*a*b* (also called CIELAB) color space, thereby generating another transformed image. The L*a*b* color space includes three channels: perceptual lightness (L*), relative green-red (a*), and relative blue-yellow (b*). The L* parameter is expressed as a number between 0 and 100, and a* and b* are unbounded with negative numbers indicating more green or more blue, respectively, and positive numbers indicating more red or more yellow, respectively. In some implementations, a* and b* may be bounded, for example, for practical or computational reasons. The parameters may require definition relative to a standard illuminant (i.e., a reference white value), such as the CIE standard illuminants D50 or D65. Operation 433A may include an intermediate operation to transform from RGB to the XYZ (also called CIEXYZ) color space.
[0049] Next, at 434A, the transformed L*a*b* image is analyzed to determine the gradient magnitude across each channel across the image. The gradient refers to the change in each channel value per pixel and may correspond to the rate of increase or decrease in luminosity and the rate of color change across the pixel. Thus, the gradient may provide information about sharp edges or changes in the input image. At operation 435A, the gradient magnitude within the temporarily identified pixel region is compared to thresholds for both the mean and standard deviation (SD) across each channel. Regions (or portions of the temporarily identified regions) where the threshold conditions are met for all three channels (e.g., the mean and standard deviation exceed the respective thresholds) may be considered as "target texture" pixel regions. For target texture pixel regions, the temporary identification may be converted to a permanent identification (e.g., identified until operation 440A or operation 460A is completed). For temporarily identified regions that are not considered target texture pixel regions, the temporary identification may be erased to discard the region. Pixels that are both target color pixels and target texture pixels may be considered target pixels, or pixels indicative of a particular bacterial species.
[0050] After operations 420A and 430A (or after operations 420A or 430A in implementations where only one thresholding is performed), an operation of removing small regions can be performed in operation 440A. Operation 440a can include comparing the size of each identified region to a size threshold (which can be in terms of a two-dimensional area and / or a number of adjacent pixels) and discarding identified regions of the transformed image having a size smaller than the size threshold. Such small regions are below a minimum acceptable area for indicating a bacterial infection.
[0051] In operation 450A, a contour is drawn (e.g., as a separate image or on the input image itself) based on the results of operations 420A and / or 430A to delineate the region of bacteria. An "overlay image" may thereby refer to the overlay itself as an element that is shown, stored, etc., separately from the input image itself, such that no changes are made to the input image itself. Operation 450A may be performed on the entire input image, or on only a portion of the input image. For example, the contour may be drawn only in a central region of the input image (e.g., within a region of a predetermined rectangle or other shape) to reduce color discrepancies when detecting thresholds between the center and edges of the image. Operation 450A may include drawing a boundary of the central region on the input image. In some implementations, the contour drawn in operation 450A may be drawn in a color, such as a color corresponding to the target bacterial species or any other color. In some implementations, the contour drawn in operation 450A may be labeled (e.g., labeled with a color and / or labeled with the name of the bacterial species). In other implementations, the contours may be filled with color or replaced with other forms (e.g., partially transparent color overlays, cross-hatching, etc.) that highlight the relevant areas with arrows or other pointers or pointing icons. In some implementations, the bacterial areas can be identified by overlaying the contours on an image other than the input image (e.g., a white light image corresponding to the input fluorescent image). In operation 460A, the overlay image is output (e.g., displayed to an operator, stored in memory, transmitted to another device, etc.). Compared to image capture, the overlay image output can be performed in real time (e.g., displayed as a live view or preview on the same device that captures the image) or near real time (e.g., within 0.5 seconds).
[0052] Another example of operations according to the present disclosure is shown in FIG. 4B. The operations include receiving an input image at 410B. The input image may be a still image retrieved from memory or captured by an image sensor, a video image retrieved from memory and / or captured by an image sensor, and / or a real-time video stream from an image sensor. The input image may be a fluorescent image capturing a response of a target area (e.g., a wound) exposed to an excitation light. The operations include verifying or otherwise ensuring the accuracy of white balance of the input image at 420B. Operation 420B may include determining whether white balance correction (e.g., self-correction by the image sensor) has been applied to the input image. The determination may be based on metadata included in the input image and / or information related to the image capture device itself. If white balance correction has been applied, operation 420 may include undoing the correction to mitigate image-to-image variation. If white balance correction has not been applied, operation 420B may not include further operations. In this manner, operation 420B sets each input image to the same color temperature and tint based on the temperature and tint values recorded during image capture.
[0053] After the white balance accuracy is verified, one to three (or more) thresholding operations are performed: operation 430B performing red fluorescent color thresholding, operation 440B performing cyan fluorescent texture thresholding, and operation 450B performing cyan fluorescent color thresholding. Operations 430B, 440B, and 450B may be performed in any order, and certain sub-operations (e.g., operations 431B and 451B and / or operations 461B and 462B, described in more detail below) may be combined. Furthermore, in some implementations, only the red operation (i.e., operation 430B) may be performed, and in other implementations, only the cyan operation (i.e., operations 440B and 450B) may be performed. Although FIG. 4B shows operations 440B and 450B being performed in parallel, in practice operations 440B and 450B may be performed sequentially in any order. Additionally or alternatively, these same processes may be performed for other colors in addition to red and cyan. It is also possible to perform texture thresholding on red fluorescence, other fluorescence, and / or white light.
[0054] Operation 430B includes operation 431B, which converts an input image (which may be captured, for example, in an RGB color space) to an HSV color space, thereby generating a converted image. The HSV color space includes three channels: hue (H), saturation (S), and value (V). The H parameter is expressed in degrees from 0° to 360°, and the S and V parameters are expressed in percentages from 0% to 100%. To convert from the RGB color space to the HSV color space, for example, a conversion algorithm may be applied to the input image on a pixel-by-pixel basis. The conversion algorithm may include normalizing each of the R, G, and B pixel values to 1, calculating the hue using a formula that depends on whether the normalized R, G, or B parameter is maximum, calculating the saturation based on the difference between the maximum normalized R, G, and B parameter and the minimum, and based on the value of the maximum normalized R, G, and B parameter, and calculating a value based on the maximum normalized R, G, and B parameter. For example, a pixel with pure cyan color is transformed from an (R,G,B) triplet of (0,255,255) to an (H,S,V) triplet of (180°,100%,100%), and a pixel with pure red color is transformed from an (R,G,B) triplet of (255,0,0) to an (H,S,V) triplet of (0°,100%,100%).
[0055] Operation 430B includes operation 432B, which compares each of the H, S, and V channels of the transformed image to a threshold condition. Pixels that meet the threshold condition for each channel may be identified as candidate pixels. In the example below, pixels that may not be "true" red pixels are thresholded out. This may be done, for example, to identify pixels that correspond to bacteria having a particular concentration, or to limit to pixels that correspond to a particular bacterial species. The thresholding operation may be performed for all pixels in each successive channel (i.e., all pixels for H, all pixels for S, all pixels for V), or for all channels in each successive pixel (i.e., H, S, and V for the top left pixel, followed by H, S, and V for the right adjacent pixel, and so on). In one example of operation 432B, the hue threshold condition may be that pixels having an H parameter less than 20° or greater than 340° are considered to be “red candidate” pixels, the saturation threshold condition may be that pixels having an S parameter greater than 75% are considered to be “red candidate” pixels, and the value threshold condition may be that pixels having a V parameter greater than 75% are considered to be “red candidate” pixels. Operation 432B may also include identifying regions of the transformed image (or original fluorescent image) where each channel satisfies a corresponding threshold condition (e.g., by adding or updating a data flag). Thus, pixels that are red candidate in each of the three channels may be considered to be “red” pixels or pixels having a color with red as a dominant hue.
[0056] After operation 430B, an operation of removing small regions may be performed at 461B. Operation 461B may include comparing the size of each identified region to a size threshold (which may be in terms of a two-dimensional area and / or a number of adjacent pixels) and discarding identified regions of the transformed image having a size smaller than the size threshold. Such small regions are below a minimum acceptable area for indicating a bacterial infection.
[0057] Operation 440B includes operation 441B, which converts the input image to grayscale, thereby generating a converted image. To convert from an RGB color space to grayscale, for example, a conversion algorithm may be applied to the input image on a pixel-by-pixel basis. The conversion algorithm may include determining a simple average of the R, G, and B components, a weighted average of the R, G, and B components, and the like. Thus, the converted grayscale image may represent the luminosity or intensity of the input image regardless of color. After operation 441B, in 442B, the converted grayscale image may be compared to an intensity threshold on a pixel-by-pixel basis, and thus operations 441B and 442B may correspond to determining the brightest or most shining portions of the input image. Pixels corresponding to intensities greater than the intensity threshold may be temporarily identified as "cyan candidate" pixels (e.g., identified until operation 440B is terminated). In some implementations, all pixels within the predetermined number of temporarily identified pixels may be temporarily identified as well, thereby generating a "cyan candidate" pixel region.
[0058] In operation 443B, the input image is transformed into the L*a*b* color space, thereby generating another transformed image. The L*a*b* color space includes three channels: perceptual lightness (L*), relative green-red (a*), and relative blue-yellow (b*). The L* parameter is expressed as a number between 0 and 100, and a* and b* are unbounded, with negative numbers indicating more green or more blue, respectively, and positive numbers indicating more red or more yellow, respectively. In some implementations, a* and b* may be bounded, for example, for practical or computational reasons. The parameters may require definition relative to a standard illuminant (i.e., a reference white value), such as the CIE standard illuminants D50 or D65. Operation 443B may include an intermediate operation to transform from RGB to XYZ color space. For example, a pixel with pure cyan color may be converted directly from an (R,G,B) triplet of (0,255,255) to an (L*,a*,b*) triplet of (91.12,-48.08,-14.14) or intermediately converted to an (X,Y,Z) triplet of (53.81,78.74,106.97).
[0059] Next, at 444B, the transformed L*a*b* image is analyzed to determine the gradient magnitude across each channel across the image. The gradient refers to the change in each channel value per pixel and may correspond to the rate of increase or decrease in luminosity and the rate of color change across the pixel. Thus, the gradient may provide information about sharp edges or changes in the input image. At operation 445B, the gradient magnitude within the temporarily identified pixel region is compared to thresholds for both the mean and SD across each channel. Regions (or portions of the temporarily identified regions) where the threshold conditions are met for all three channels (e.g., the mean and standard deviation exceed the respective thresholds) may be considered as "cyan texture" pixel regions. For cyan texture pixel regions, the temporary identification may be converted to a permanent identification (e.g., identified until operation 462B or operation 480B is completed). For temporarily identified regions that are not considered cyan texture pixel regions, the temporary identification may be erased to discard the region.
[0060] Operation 450B includes operation 451B, which converts the input image to HSV color space. In implementations in which both operations 430B and 450B are performed, operations 431B and 451B may be combined to avoid unnecessary repetition. Operation 450 includes operation 452B, which compares each of the H, S, and V channels of the converted image to a threshold condition. Pixels that satisfy the threshold condition for each channel may be identified as candidate pixels. In the example below, pixels that may not be "true" cyan pixels are eliminated by thresholding. This may be done, for example, to identify pixels that correspond to bacteria having a particular concentration, or to limit to pixels that correspond to a particular bacterial species. The thresholding operation may be performed for all pixels in each successive channel (i.e., all pixels for H, all pixels for S, all pixels for V), or for all channels in each successive pixel (i.e., H, S, and V for the top left pixel, followed by H, S, and V for the right adjacent pixel, and so on). In one example of operation 432B, the hue threshold condition may be that pixels having an H parameter between 160° and 200° are considered to be “cyan candidate” pixels, the saturation threshold condition may be that pixels having an S parameter greater than 75% are considered to be “cyan candidate” pixels, and the value threshold condition may be that pixels having a V parameter greater than 75% are considered to be “cyan candidate” pixels. Operation 452B may also include identifying regions of the transformed image (or original fluorescent image) where each channel satisfies a corresponding threshold condition. Thus, pixels that are cyan candidates in each of the three channels may be considered to be “cyan colored” pixels. In some implementations, operation 452B may be performed only on the regions permanently identified in operation 445B, or conversely, operation 445B may be performed only on the regions identified in operation 452B (in some implementations where the order of operations 440B and 450B is reversed). In other implementations, both operations 445B and 452B may be performed separately and their results added or combined together. Thus, a pixel that is both a cyan color pixel and a cyan texture pixel may be considered a "cyan" pixel, or may be considered a pixel indicative of a particular bacterial species.
[0061] After operations 440B and 450B, another operation may be performed at 462B to remove small regions. Operation 462B may include comparing the size of each identified region to a size threshold (which may be in terms of a two-dimensional area and / or a number of adjacent pixels) and discarding identified regions of the transformed image having a size smaller than the size threshold. Such small regions are below a minimum acceptable area for indicating a bacterial infection. In implementations in which operations 430B, 440B, and 450B are performed, operations 461B and 462B may be combined to avoid unnecessary repetition.
[0062] In operation 470B, a red contour is drawn on the input image based on the results of operations 430B and 461B. In operation 480B, a cyan contour is drawn on the input image based on the results of operations 440B, 450B, and 462B. This allows an overlay image to be generated. In an implementation in which operation 430B is performed but not operations 440B or 450B, operation 480B may be omitted or skipped. Similarly, in an implementation in which operations 440B and 450B are performed but not operation 430B, operation 470B may be omitted or skipped. Operation 470B and / or 480B may be performed on the entire input image or on only a portion of the input image. For example, the contour may be drawn only in a central region of the input image (e.g., within a predetermined rectangular or other shaped region) to reduce color discrepancies when detecting thresholds between the center and edges of the image. Operation 470B and / or 480B may include drawing a boundary of the central region on the input image. In some implementations, the contour drawn in act 470B and the contour drawn in act 480B may be drawn using different colors. For example, the contour drawn in act 470B may be drawn in red and the contour drawn in act 480B may be drawn in cyan. In other implementations (e.g., in a color-blind mode), the contours drawn in acts 470B and 480B may be drawn in the same color and may be labeled (e.g., labeled "red" and "cyan" and / or labeled with the name of the bacterial species). In other implementations, the contours may be filled with color or replaced with other forms (e.g., partially transparent color overlay, cross-hatching, etc.) that highlight the relevant area with an arrow or other pointer or pointing icon. In act 490B, the overlay image is output (e.g., displayed to an operator, stored in memory, sent to another device, etc.). Compared to image capture, the overlay image output may be performed in substantially real-time (e.g., displayed as a live view or preview on the same device that captures the image) or near real-time (e.g., within 0.5 seconds).
[0063] The above operations in FIG. 4B include exemplary operations in specific color spaces (e.g., RGB for image capture, HSV for color analysis conversion, and L*a*b* for texture analysis conversion), but these color spaces are merely exemplary and not limiting. As described above with respect to FIG. 3, any of the operations can be performed in any desired color space, including, but not limited to, HSV, HSL, CIEXYZ, CIELUV, L*a*b*, RGB, YCbCr, YUV, LCh, CMY, CMYK, and custom color spaces. Thus, certain conversion operations shown in FIG. 4B (e.g., zero, one, or more of operations 431B, 441B, 443B, and 451B) may be combined with one another or omitted entirely if the image is already in the target color space.
[0064] 5A-5B show example images input to or resulting from the operations of FIG. 4B. FIG. 5A shows an input fluorescent image 510 including a red region. FIG. 5B shows a corresponding overlay image 520 including a boundary of a central region 521 and a contour 522 highlighting the red region, for example, as determined by the operations of FIG. 4B. FIG. 5B may be the result of only the red operation of FIG. 4B (e.g., a path including operation 430B) or may be the result of both the red and cyan operations being performed on an input image that does not have a cyan region. FIG. 6A shows an input fluorescent image 610 including a cyan region. FIG. 6B shows a corresponding overlay image 620 including a boundary of a central region 621 and a contour 622 highlighting the cyan region, for example, as determined by the operations of FIG. 4B. FIG. 6B may be the result of only the cyan operation of FIG. 4B (e.g., a path including operations 440B and 450B) or may be the result of both the red and cyan operations being performed on an input image that does not have a cyan region. In another example not explicitly shown, the overlay image may include contours corresponding to red regions at the same time as contours corresponding to cyan regions.
[0065] Thus, the above-described systems, devices, and methods can provide information to accurately, reliably, and / or easily identify potential areas of concern (e.g., bacterial infections) in fluorescence images, thereby improving treatment and reducing errors.
[0066] The exemplary systems, devices, and methods described herein may be performed under the control of a processing system that executes computer readable code embodied on a computer readable recording medium or communication signals transmitted via a transitory medium. The computer readable recording medium may be any data storage device capable of storing data readable by the processing system, and may include both volatile and non-volatile media, removable and non-removable media, as well as media readable by databases, computers, and various other network devices.
[0067] Examples of computer readable recording media include, but are not limited to, read only memory (ROM), random access memory (RAM), erasable electrically programmable ROM (EEPROM), flash memory or other memory technology, holographic media or other optical disk storage, magnetic storage including magnetic tape and magnetic disk, and solid state storage. The computer readable recording media may also be distributed over network coupled computer systems such that the computer readable code is stored and executed in a distributed manner. The communication signal transmitted over the transitory medium may include, for example, a modulated signal transmitted over a wired or wireless transmission path.
[0068] Illustrative examples of the systems, methods, and devices described herein are provided below. Embodiments of the systems, methods, and / or devices described herein can include any one or more, and any combination, of the provisions described below.
[0069] Clause 1. An imaging device configured to receive a fluorescent image, a memory, and a processor, the method comprising: converting the fluorescent image into an HSV color space to generate a first converted image; comparing a saturation parameter of the first converted image to a first saturation threshold condition for a first range of hue parameters of the first converted image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescent signature; and comparing a value parameter of the first converted image to a first value threshold condition for the first range of hue parameters. and a processor configured to: perform a color analysis of the fluorescent image including comparing, where a first value threshold condition is based on a correlation between a first color and a first bacterial fluorescence signature; identifying, for a first range of hue parameters, regions of the fluorescent image where the saturation parameter satisfies the first saturation threshold condition and the value parameter satisfies the first value threshold parameter; discarding the identified regions that are smaller than a size threshold; and outlining the remaining identified regions on the fluorescent image to generate an overlay image.
[0070] Clause 2. The device described in Clause 1, wherein the color analysis further includes: comparing a saturation parameter of the first converted image to a second saturation threshold condition for a second range of hue parameters of the first converted image, the second saturation threshold condition being based on a correlation between a second color corresponding to the second range of hue parameters and a second bacterial fluorescence signature; comparing a value parameter of the first converted image to a second value threshold condition for the second range of hue parameters, the second value threshold condition being based on a correlation between the second color and the second bacterial fluorescence signature; and identifying areas of the fluorescent image where, for the second range of hue parameters, the saturation parameter satisfies the second saturation threshold condition and the value parameter satisfies the second value threshold condition.
[0071] Clause 3. The device of clause 3, wherein the first color is cyan and the second color is red or a color having red as a predominant hue.
[0072] Clause 4. The apparatus of any one of clauses 1 to 3, wherein the processor is configured to perform texture analysis of the fluorescence image, comprising: converting the fluorescence image to grayscale, thereby generating a second transformed image; comparing, for each pixel, an intensity parameter of the second transformed image with an intensity threshold; temporarily identifying regions within a predetermined pixel distance of pixels whose intensity parameter exceeds the intensity threshold; generating a third transformed image by converting the fluorescence image to L*a*b* color space; determining, in each channel of the third transformed image, respective gradients of the temporarily identified regions; and permanently identifying one of the temporarily identified regions in which the respective gradient of each channel of the third transformed image exceeds the gradient threshold.
[0073] Clause 5. First bacterial fluorescence signature is 10 4 5. The device according to any one of clauses 1 to 4, corresponding to a bacterial concentration of cfu / g or more.
[0074] Article 6. First Bacterial Fluorescence Signatures for Pseudomonas aeruginosa, Escherichia coli, Proteus mirabilis, Proteus vulgaris, Enterobacter cloacae, Serratia marcescens, Acinetobacter baumannii, Klebsiella pneumoniae, Klebsiella oxytoca, Morganella morganii, Stenotrophomonas maltophilia, Citrobacter koseri, Citrobacter freundii, Aeromonas hydrophilia, Alcaligenes faecalis, Pseudomonas putida 6. The device according to any one of clauses 1 to 5, wherein the device corresponds to at least one of the following bacteria: Staphylococcus aureus, Staphylococcus epidermidis, Staphylococcus lugdunensis, Staphylococcus capitis, Corynebacterium striatum, Bacillus cereus, Listeria monocytogenes, Bacteroides fragilis, Clostridium perfringens, Peptostreptococcus anaerobius, Propionibacterium acnes, and / or Veillonella parvula.
[0075] Clause 7. The device of any one of clauses 1 to 6, further comprising an excitation light source configured to irradiate the target with excitation light.
[0076] Clause 8. The apparatus of clause 7, wherein the excitation light comprises a wavelength of 400 nm to 450 nm. Clause 9. An apparatus according to any one of clauses 1 to 8, wherein the imaging device includes a first image sensor configured to detect wavelengths between 500 nm and 550 nm.
[0077] Clause 10. An apparatus according to any one of clauses 1 to 9, wherein the imaging device includes a second image sensor configured to detect wavelengths between 600 nm and 660 nm.
[0078] Clause 11. The apparatus of any one of clauses 1 to 10, further comprising a display device configured to output a fluorescence image and / or an overlay image.
[0079] Clause 12. The apparatus of clause 11, wherein the display device is configured to output the overlay image in real time or near real time.
[0080] Clause 13. An apparatus described in any one of clauses 1 to 12, wherein the processor is configured to ensure accuracy of white balance of the fluorescent image.
[0081] Clause 14. The apparatus of clause 13, wherein ensuring white balance accuracy includes determining whether a previous white balance correction has been applied based on metadata included in the fluorescence image.
[0082] Clause 15. Apparatus according to clause 13 or 14, comprising performing a correction undoing operation.
[0083] Clause 16. An imaging device comprising a display device, an excitation light source configured to illuminate a target area with excitation light, and an image sensor configured to capture a fluorescence image responsive to a response of the target area to the excitation light, a housing, and a processing circuit disposed within the housing, the processing circuit including a processor, the processor converting the fluorescence image into an HSV color space, thereby generating a first converted image, and comparing a saturation parameter of the first converted image to a first saturation threshold condition for a first range of hue parameters of the first converted image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature. and a processing circuit configured to: perform a color analysis of the fluorescent image, including comparing a value parameter of the first transformed image to a first value threshold condition for a first range of hue parameters, the first value threshold condition being based on a correlation between the first color and a first bacterial fluorescence signature; identifying regions of the fluorescent image where a saturation parameter satisfies the first saturation threshold condition and where a value parameter satisfies the first value threshold parameter for the first range of hue parameters; discarding identified regions that are smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions on the fluorescent image.
[0084] Clause 17. A method of fluorescence image interpretation comprising: receiving a fluorescence image; converting the fluorescence image into an HSV color space, thereby generating a first transformed image; comparing a saturation parameter of the first transformed image to a first saturation threshold condition, for a first range of hue parameters of the first transformed image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature; comparing a value parameter of the first transformed image to a first value threshold condition, for the first range of hue parameters, the first value threshold condition being based on a correlation between the first color and the first bacterial fluorescence signature; identifying regions of the fluorescence image whose saturation parameter satisfies the first saturation threshold condition and whose value parameter satisfies the first value threshold parameter, for the first range of hue parameters; discarding the identified regions that are smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions on the fluorescence image.
[0085] Clause 18. A portable handheld device comprising: an imaging device configured to receive a fluorescence image; a memory; and a processor configured to perform texture analysis of the fluorescence image, including: generating a first transformed image by converting the fluorescence image to grayscale; comparing, for each pixel, an intensity parameter of the first transformed image to an intensity threshold; temporarily identifying regions within a predetermined pixel distance of the pixels whose intensity parameter exceeds the intensity threshold; generating a second transformed image by converting the fluorescence image to a predetermined color space; determining, in each channel of the second transformed image, respective gradients of the temporarily identified regions; and permanently identifying one of the temporarily identified regions where the respective gradient of each channel of the second transformed image exceeds the gradient threshold.
[0086] Clause 19. The apparatus of clause 18, wherein the predetermined color space is an HSV color space, a CIEXYZ color space, a CIELUV color space, an RGB color space, a YCbCr color space, a YUV color space, an LCh color space, an L*a*b* color space, a CMY color space, a CMYK color space, or a custom color space.
[0087] Clause 20. The apparatus according to clause 18, wherein the predetermined color space is the L*a*b* color space. Clause 21. A method of interpreting clinical images, comprising: receiving, on a handheld imaging device, an input image; performing color and / or texture analysis on the input image to identify areas of clinical interest in the input image; generating an overlay, the overlay including one or more colors, highlighting, arrows, or other highlighting methods that, when placed on the input image, provide an indication of one or more clinical features of interest; and displaying, in real time, an output image including the input image and the overlay, wherein the clinical features of interest on the input image are highlighted to a user of the imaging device by the overlay.
[0088] Clause 22. The method of clause 21, wherein receiving an input image includes receiving a fluorescent image of the wound area.
[0089] Clause 23. The method of clause 21 or 22, wherein performing colour and / or texture analysis on the input image to identify regions of clinical interest in the input image comprises performing colour analysis to identify locations of colours corresponding to clinical features present in the input image.
[0090] Clause 24. The method of clause 23, wherein performing color analysis to identify locations of colors corresponding to clinical features present in the input image includes identifying areas in the input image where the color corresponding to the clinical feature exceeds a threshold amount.
[0091] Article 25. The threshold amount is 10 4 25. The method according to clause 24, corresponding to a bacterial concentration of cfu / g or more.
[0092] Clause 26. The method of clause 24, wherein the threshold amount corresponds to one or more of the presence, location, or abundance of a particular bacterial species.
[0093] Clause 27. The method of clause 21 or 22, wherein performing colour and / or texture analysis on the input image to identify regions of clinical interest in the input image comprises performing texture analysis to identify locations of clinically not interest in the input image.
[0094] Clause 28. The method of any one of clauses 21-27, wherein displaying the output image including the input image and the overlay in real time includes displaying the output image on a touch screen display of a handheld imaging device.
[0095] Clause 29. The method of clause 28, further comprising switching between displaying the input image and displaying the output image on a display of the handheld imaging device.
[0096] Clause 30. An imaging device configured to receive an input image, a memory, and a processor, the imaging device configured to: convert the input image into a predetermined color space including a first channel corresponding to a first parameter, a second channel corresponding to a second parameter, and a third channel corresponding to a third parameter, thereby generating a converted image; comparing a second parameter of the first converted image to a threshold condition for a first range of the first parameter of the converted image, the first threshold condition being based on a correlation between the first range of the first parameter and a bacterial signature; and comparing a third parameter of the converted image for the first range of the first parameter to a threshold condition. and a processor configured to: perform a color analysis of the input image, including comparing the parameter to a second threshold condition, the second threshold condition being based on a correlation between a first range of the first parameter and a bacterial signature; identifying, for the first range of the first parameter, regions of the input image where the second parameter satisfies the first threshold condition and where a third parameter satisfies the second threshold condition; discarding identified regions that are smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions of the input image.
[0097] Clause 31. The apparatus of clause 30, wherein the predetermined color space is an HSV color space, a CIEXYZ color space, a CIELUV color space, an RGB color space, a YCbCr color space, a YUV color space, an LCh color space, an L*a*b* color space, a CMY color space, a CMYK color space, or a custom color space.
[0098] Clause 32. The apparatus of clause 30, wherein the predetermined color space is the HSV color space. Clause 33. The apparatus of clause 30, wherein the processor is further configured to fill the outlined region with a predetermined color.
[0099] The above description and the associated figures teach the best mode of the invention and are intended to be illustrative rather than limiting. Many embodiments and applications other than the examples provided will be apparent to those skilled in the art upon reading the above description. The scope should be determined not with reference to the above description, but with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled. It is anticipated and intended that future developments will occur in the technology discussed herein, and that the disclosed systems and methods will be incorporated into future embodiments. In short, it should be understood that applications are subject to modification and change.
[0100] All terms used in the claims are intended to be given their broadest reasonable construction and their ordinary meaning as understood by one skilled in the art described herein, unless expressly indicated to the contrary herein. In particular, the use of singular articles such as "a," "the," "said," etc., should be read as describing one or more of the indicated elements, unless the claims recite an express limitation to the contrary.
[0101] The Abstract is provided to enable the reader to quickly ascertain the nature of the technical disclosure. It is submitted with the understanding that it will not be used to interpret or limit the scope or meaning of the claims. Moreover, in the foregoing Detailed Description, it will be seen that various features have been grouped together in various embodiments for the purpose of streamlining the disclosure. This method of disclosure should not be interpreted as reflecting an intention that the claimed embodiments require more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive subject matter lies in less than all features of a single disclosed embodiment. Thus, the following claims are incorporated into the Detailed Description, with each claim standing on its own as separately claimed subject matter.
Claims
1. 1. A portable handheld device, comprising: an imaging device configured to receive a fluorescence image; Memory, 1. A processor comprising: converting the fluorescence image into HSV color space to generate a first converted image; comparing a saturation parameter of the first transformed image to a first saturation threshold condition for a first range of hue parameters of the first transformed image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature; comparing a value parameter of the first transformed image to a first value threshold condition for the first range of the hue parameter, the first value threshold condition being based on the correlation between the first color and the first bacterial fluorescence signature; identifying regions of the fluorescent image where the saturation parameter satisfies the first saturation threshold condition and the value parameter satisfies the first value threshold condition for the first range of the hue parameter; performing a color analysis of the fluorescent image, comprising: discarding the identified regions that are smaller than a size threshold; and outlining the remaining identified regions on the fluorescent image to generate an overlay image; A processor configured to execute An apparatus comprising:
2. The color analysis may include comparing the saturation parameter of the first transformed image to a second saturation threshold condition for a second range of the hue parameter of the first transformed image, the second saturation threshold condition being based on a correlation between a second color corresponding to the second range of the hue parameter and a second bacterial fluorescence signature; comparing the value parameter of the first transformed image to a second value threshold condition for the second range of the hue parameter, the second value threshold condition being based on a correlation between the second color and the second bacterial fluorescence signature; identifying regions of the fluorescent image where the saturation parameter satisfies the second saturation threshold condition and the value parameter satisfies the second value threshold condition for the second range of the hue parameter; The apparatus of claim 1 further comprising:
3. 3. The apparatus of claim 2, wherein the first color is cyan and the second color is red or a color having red as a predominant hue.
4. The processor, converting the fluorescent image to grayscale, thereby generating a second converted image; comparing, for each pixel, an intensity parameter of the second transformed image with an intensity threshold; temporarily identifying a region within a predetermined pixel distance of pixels in which said intensity parameter exceeds said intensity threshold; converting the fluorescence image into an L*a*b* color space to generate a third converted image; determining a gradient for each of the temporarily identified regions within each channel of the third transformed image; permanently identifying one of the temporarily identified regions in which the respective gradient of each channel of the third transformed image exceeds a gradient threshold; The apparatus according to any one of claims 1 to 3, configured to perform a texture analysis of the fluorescence image, comprising:
5. The first bacterial fluorescence signature is 10 4 The device according to any one of claims 1 to 4, which corresponds to a bacterial concentration of cfu / g or more.
6. The first bacterial fluorescence signature is selected from the group consisting of Pseudomonas aeruginosa, Escherichia coli, Proteus mirabilis, Proteus vulgaris, Enterobacter cloacae, Serratia marcescens, Acinetobacter baumannii, Klebsiella pneumoniae, Klebsiella oxytoca, Morganella morganii, Stenotrophomonas maltophilia, Citrobacter koseri, Citrobacter freundii, Aeromonas hydrophilia, Alcaligenes faecalis, Pseudomonas putida, 6. The device of claim 1 , which is adapted to at least one of the following: Staphylococcus aureus, Staphylococcus epidermidis, Staphylococcus lugdunensis, Staphylococcus capitis, Corynebacterium striatum, Bacillus cereus, Listeria monocytogenes, Bacteroides fragilis, Clostridium perfringens, Peptostreptococcus anaerobius, Propionibacterium acnes, and / or Veillonella parvula.
7. The apparatus of any one of claims 1 to 6, further comprising an excitation light source configured to irradiate the target with excitation light.
8. The apparatus of claim 7 , wherein the excitation light comprises a wavelength between 400 nm and 450 nm.
9. The apparatus of any one of claims 1 to 8, wherein the imaging device comprises a first image sensor configured to detect wavelengths between 500 nm and 550 nm.
10. The apparatus of any one of claims 1 to 9, wherein the imaging device includes a second image sensor configured to detect wavelengths between 600 nm and 660 nm.
11. An apparatus according to any one of claims 1 to 10, further comprising a display device configured to output the fluorescence image and / or the overlay image.
12. The apparatus of claim 11 , wherein the display device is configured to output the overlay image in real time or near real time.
13. An apparatus according to any preceding claim, wherein the processor is configured to ensure correct white balance of the fluorescence image.
14. The apparatus of claim 13 , wherein ensuring the accuracy of the white balance comprises determining whether a previous white balance correction has been applied based on metadata included in the fluorescence image.
15. 15. Apparatus according to claim 13 or 14, comprising performing a correction undo operation.
16. A display device; an imaging device comprising an excitation light source configured to irradiate a target area with excitation light and an image sensor configured to capture a fluorescence image responsive to a response of the target area to the excitation light; Housing and processing circuitry disposed within the housing, the processor comprising: converting the fluorescence image into HSV color space, thereby generating a first converted image; comparing a saturation parameter of the first transformed image to a first saturation threshold condition for a first range of hue parameters of the first transformed image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature; comparing a value parameter of the first transformed image to a first value threshold condition for the first range of the hue parameter, the first value threshold condition being based on the correlation between the first color and the first bacterial fluorescence signature; identifying regions of the fluorescent image where the saturation parameter satisfies the first saturation threshold condition and the value parameter satisfies the first value threshold condition for the first range of the hue parameter; performing a color analysis of the fluorescent image, comprising: discarding the identified regions that are smaller than a size threshold; and generating an overlay image by outlining remaining identified regions on the fluorescence image; a processing circuit configured to execute A system comprising:
17. 1. A method for fluorescence image interpretation comprising: Receiving a fluorescent image; converting the fluorescence image into HSV color space, thereby generating a first converted image; comparing a saturation parameter of the first transformed image to a first saturation threshold condition for a first range of hue parameters of the first transformed image, the first saturation threshold condition being based on a correlation between a first color corresponding to the first range of hue parameters and a first bacterial fluorescence signature; comparing a value parameter of the first transformed image to a first value threshold condition for the first range of the hue parameter, the first value threshold condition being based on the correlation between the first color and the first bacterial fluorescence signature; identifying regions of the fluorescent image where the saturation parameter satisfies the first saturation threshold condition and the value parameter satisfies the first value threshold condition for the first range of the hue parameter; performing a color analysis of the fluorescent image, comprising: discarding the identified regions that are smaller than a size threshold; and generating an overlay image by outlining remaining identified regions on the fluorescence image; A method comprising:
18. 1. A portable handheld device, comprising: an imaging device configured to receive a fluorescence image; Memory, converting the fluorescence image to grayscale to generate a first converted image; comparing, for each pixel, an intensity parameter of the first transformed image with an intensity threshold; temporarily identifying a region within a predetermined pixel distance of pixels in which said intensity parameter exceeds said intensity threshold; generating a second transformed image by transforming the fluorescence image into a predetermined color space; determining a gradient for each of the temporarily identified regions within each channel of the second transformed image; permanently identifying one of the temporarily identified regions in which the respective gradient of each channel of the second transformed image exceeds a gradient threshold; a processor configured to perform texture analysis of the fluorescence image, the processor comprising: An apparatus comprising:
19. 20. The apparatus of claim 18, wherein the predetermined color space is an HSV color space, a CIEXYZ color space, a CIELUV color space, an RGB color space, a YCbCr color space, a YUV color space, an LCh color space, an L*a*b* color space, a CMY color space, a CMYK color space, or a custom color space.
20. The apparatus of claim 18 , wherein the predetermined color space is the L*a*b* color space.
21. Receiving an input image on a handheld imaging device; performing color and / or texture analysis on the input image to identify regions of clinical interest within the input image; generating an overlay, the overlay including one or more colors, highlighting, arrows, or other highlighting methods that, when placed over the input image, provides an indication of one or more clinical features of interest; displaying in real time the input image and an output image including the overlay, where clinical features of interest on the input image are highlighted by the overlay to a user of the imaging device; A method for interpreting clinical images, including:
22. 22. The method of claim 21, wherein receiving an input image comprises receiving a fluorescent image of the wound.
23. 23. The method of claim 21 or 22, wherein performing colour and / or texture analysis on the input image to identify regions of clinical interest in the input image comprises performing colour analysis to identify locations of colours corresponding to clinical features present in the input image.
24. 24. The method of claim 23, wherein performing a color analysis to identify locations of colors present in the input image corresponding to clinical features comprises identifying regions in the input image where the colors corresponding to the clinical features exceed a threshold amount.
25. The threshold amount is 10 4 25. The method of claim 24, corresponding to a bacterial concentration of cfu / g or more.
26. 25. The method of claim 24, wherein the threshold amount corresponds to one or more of the presence, location, or abundance of a particular bacterial species.
27. 23. The method of claim 21 or 22, wherein performing a colour and / or texture analysis on the input image to identify regions of clinical interest in the input image comprises performing a texture analysis to identify locations of clinically not interest in the input image.
28. 28. The method of claim 21, wherein displaying the input image and the output image including the overlay in real time comprises displaying the output image on a touch screen display of the handheld imaging device.
29. 30. The method of claim 28, further comprising switching between displaying the input image and displaying the output image on the display of the handheld imaging device.
30. 1. A portable handheld device, comprising: an imaging device configured to receive an input image; Memory, 1. A processor comprising: transforming the input image into a predetermined color space including a first channel corresponding to a first parameter, a second channel corresponding to a second parameter, and a third channel corresponding to a third parameter, thereby generating a transformed image; comparing the second parameter of the first transformed image to a threshold condition for a first range of the first parameter of the transformed image, the first threshold condition being based on a correlation between the first range of the first parameter and a bacterial signature; comparing the third parameter of the transformed image to a second threshold condition for the first range of the first parameter, the second threshold condition being based on a correlation between the first range of the first parameter and the bacterial signature; identifying regions of the input image where, for the first range of the first parameter, the second parameter satisfies the first threshold condition and the third parameter satisfies the second threshold condition; performing a color analysis of the input image, comprising: discarding the identified regions that are smaller than a size threshold; and generating an overlay image by outlining the remaining identified regions of the input image; A processor configured to execute An apparatus comprising:
31. 31. The apparatus of claim 30, wherein the predetermined color space is an HSV color space, a CIEXYZ color space, a CIELUV color space, an RGB color space, a YCbCr color space, a YUV color space, an LCh color space, an L*a*b* color space, a CMY color space, a CMYK color space, or a custom color space.
32. 31. The apparatus of claim 30, wherein the predetermined color space is the HSV color space.
33. 31. The apparatus of claim 30, wherein the processor is further configured to fill the outlined region with a predetermined color.
