Technologies for utilizing retinal oximetry to measure blood oxygen levels

Abstract

Technologies for utilizing retinal oximetry to measure blood oxygen levels are disclosed. An example method includes capturing, by one or more imagers of a handheld imaging device, at least two images of a retinal vasculature. The example method further includes determining (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images; and determining an optical density ratio of the retinal vasculature based on the first optical density and the second optical density. The example method further includes determining a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio; and outputting the TBI likelihood value for display to a user.

Description

TECHNOLOGIES FOR UTILIZING RETINAL OXIMETRY TO MEASURE BLOOD OXYGEN LEVELSCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 516,835, filed July 31, 2023, and entitled “Technologies For Utilizing Retinal Oximetry To Measure Blood Oxygen Levels”, which is incorporated herein by reference in its entirety.FIELD OF THE DISCLOSURE

[0002] The present disclosure relates generally to blood oxygen measurement / management and, more particularly, to technologies for utilizing retinal oximetry to measure blood oxygen levels.BACKGROUND

[0003] The background description provided herein is for the purpose of generally presenting the context of the disclosure. Work of the presently named inventor, to the extent it is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.

[0004] There exist multiple neurologic conditions where monitoring brain oxygenation is of utmost importance. One such condition, traumatic brain injury (TBI), is known to alter brain oxygenation levels. As a result, maintenance of cerebral oxygenation is a mainstay in the treatment of TBI as patients recover in the intensive care unit. Thus, monitoring brain tissue oxygen tension (PBO2) is critical to the recovery of TBI patients.

[0005] Presently, there exist two conventional methods to measure brain PBO2, the optical luminescent method and the polarographic method. Briefly, the optical luminescent method involves the diffusion of oxygen molecules into a silicone matrix, which in turn, changes the color of a dye. That change in color is detected via light and converted into a partial pressure of oxygen. The polarographic method involves a battery-like electrode (e.g., a Clark electrode) in which oxygen diffuses through a membrane to reduce at the cathode, the rate of which is converted to a partial pressure of oxygen.

[0006] However, these conventional methods suffer from several notable drawbacks. For example, the brain tissue oxygen monitors required in both methods are necessarily placed within the brain parenchyma proper, 2.5 to 3 cm below the dura. Patients with significant TBI are generally in a precarious state, such that the placement of a monitor itself is not an insignificant procedure, requiring invasion of functional brain parenchyma and consistent monitoring due to its potential as a nidus for devastating infection. Consequently, these conventional methods pose a substantial health risk to patients suffering from TBI or other similar neurologic conditions.

[0007] Nevertheless, these conventional techniques may be used in tandem with other supplementary techniques to monitor blood oxygenation values, but these supplementary techniques also suffer from drawbacks causing misleading results. For example, another common technique to measure oxygen saturation values is dermal pulse oximetry. A dermal pulse oximeter is a device that is typically placed on a fingertip that uses light beams to estimate the oxygen saturation of the blood and a patient’s pulse rate. Such a dermal pulse oximeter may be used in tandem with the conventional brain PBO2 measurement methods to provide supplemental / additional data regarding blood oxygenation, but conventional dermal pulse oximeters suffer from racial bias. Namely, the original development of dermal pulse oximetry occurred in non-racially diverse populations, leading to skewed readings for racially diverse patients (e.g., black patients have nearly three times the frequency of occult hypoxemia that is undetected by pulse oximetry relative to white patients).

[0008] Therefore, there is a need for improved blood oxygenation measurement techniques that alleviate the health risks, eliminate potential biases, and mitigate other drawbacks posed by conventional procedures, and thereby improve the resulting monitoring and treatment of the underlying neurological conditions.SUMMARY OF THE INVENTION

[0009] According to an embodiment of the present disclosure, a handheld imaging device configured to determine retinal oximetry values is disclosed. The handheld imaging device may comprise: one or more imagers; one or more processors; and one or more memories storing thereon executable instructions. The instructions, when executed by the one or more processors, may cause the one or more processors to: capture at least two images of a retinal vasculatureusing the one or more imagers, determine (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images, determine an optical density ratio of the retinal vasculature based on the first optical density and the second optical density, determine a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio, and output the TBI likelihood value for display to a user.

[0010] In a variation of this embodiment, the handheld imaging device may further comprise: one or more illumination sources configured to emit illumination with wavelengths of at least 570 nm and 601 nm; and one or more bandpass filters configured to transmit illumination at wavelengths of at least 570 nm and 601 nm.

[0011] In another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: determine the TBI likelihood value by applying a trained mathematical model or a trained machine learning (ML) model to the optical density ratio value.

[0012] In yet another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: generate, based on the TBI likelihood value, at least one of: (i) a treatment recommendation, (ii) a predicted diagnosis, or (iii) a predicted prognosis; and output the treatment recommendation, the predicted diagnosis, or the predicted prognosis for display to the user.

[0013] In still another variation of this embodiment, the TBI likelihood value may be a first TBI likelihood value, and the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: transmit the at least two images to a remote server configured to analyze retinal images; receive a second TBI likelihood value from the remote server; compare the first TBI likelihood value to the second TBI likelihood value to determine a composite TBI likelihood value; and output the composite TBI likelihood value for display to the user.

[0014] In yet another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: determine that a user has misaligned the handheld device based on one or more images captured by the one or more imagers; determine one or more alignment instructions to help the user align the handhelddevice based on the one or more images; and output the one or more alignment instructions for display to the user.

[0015] In still another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: connect with a wearable device; receive, from the wearable device, a baseline blood oxygenation value; and determine the TBI likelihood value based on the optical density ratio and the baseline blood oxygenation value.

[0016] In yet another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: prior to capturing the at least two images, execute an image stabilization algorithm configured to stabilize the capturing of the at least two images.

[0017] In still another variation of this embodiment, the one or more imagers may include at least two imagers.

[0018] In yet another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: generate a retinal vasculature map that visually indicates relative oxygenation values of the retinal vasculature.

[0019] In still another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: determine a vessel diameter of the retinal vasculature based on the at least two images.

[0020] In yet another variation of this embodiment, the executable instructions, when executed by the one or more processors, may further cause the one or more processors to: connect to a dermal oximeter device; receive blood oxygen saturation data from the dermal oximeter device; calculate an oximetry calibration value based on the blood oxygen saturation data and the optical density ratio; and adjust subsequent data from the dermal oximeter device based on the oximetry calibration value.

[0021] In another embodiment of the present disclosure, a method for determining retinal oximetry values is disclosed. The method may comprise: capturing, by one or more imagers of a handheld imaging device, at least two images of a retinal vasculature; determining, by one ormore processors, (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images; determining, by the one or more processors, an optical density ratio of the retinal vasculature based on the first optical density and the second optical density; determining, by the one or more processors, a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio; and outputting, by the one or more processors, the TBI likelihood value for display to a user.

[0022] In a variation of this embodiment, the method may further comprise: emitting, by one or more illumination sources of the handheld imaging device, illumination with wavelengths of at least 570 nm and 601 nm during the capturing of the at least two images..

[0023] In another variation of this embodiment, the method may further comprise: filtering, by one or more bandpass filters of the handheld imaging device, received illumination to capture wavelengths of at least 570 nm and 601 nm during the capturing of the at least two images.

[0024] In still another variation of this embodiment, the method may further comprise: determining, by the one or more processors, the TBI likelihood value by applying a trained mathematical model or a trained machine learning (ML) model to the optical density ratio value.

[0025] In yet another variation of this embodiment, the method may further comprise: generating, by the one or more processors and based on the TBI likelihood value, at least one of: (i) a treatment recommendation, (ii) a predicted diagnosis, or (iii) a predicted prognosis; and outputting, by the one or more processors, the treatment recommendation, the predicted diagnosis, or the predicted prognosis for display to the user.

[0026] In still another variation of this embodiment, the TBI likelihood value may be a first TBI likelihood value, and the method may further comprise: transmitting, by the one or more processors, the at least two images to a remote server configured to analyze retinal images; receiving, at the one or more processors, a second TBI likelihood value from the remote server; comparing, by the one or more processors, the first TBI likelihood value to the second TBI likelihood value to determine a composite TBI likelihood value; and outputting, by the one or more processors, the composite TBI likelihood value for display to the user.

[0027] In yet another variation of this embodiment, the method may further comprise: determining, by the one or more processors, that a user has misaligned the handheld de-vice based on one or more images captured by the one or more imagers; determining, by the one or more processors, one or more alignment instructions to help the user align the handheld device based on the one or more images; and outputting, by the one or more processors, the one or more alignment instructions for dis-play to the user.

[0028] In yet another embodiment of the present disclosure, a non-transitory computer- readable storage medium is disclosed. The non-transitory computer-readable storage medium may have stored thereon a set of instructions, executable by at least one processor, for determining retinal oximetry values. The instructions may comprise: instructions for capturing, by one or more imagers of a handheld imaging device, at least two images of a retinal vasculature; instructions for determining (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images; instructions for determining an optical density ratio of the retinal vasculature based on the first optical density and the second optical density; instructions for determining a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio; and instructions for outputting the TBI likelihood value for display to a user.BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The figures described below depict various aspects of the system and methods disclosed herein. It should be understood that each figure depicts an embodiment of a particular aspect of the disclosed system and methods, and that each of the figures is intended to accord with a possible embodiment thereof. Further, wherever possible, the following description refers to the reference numerals included in the following figures, in which features depicted in multiple figures are designated with consistent reference numerals.

[0030] FIG. 1 illustrates an example environment for utilizing retinal oximetry to measure blood oxygen levels, in accordance with various aspects disclosed herein.

[0031] FIGs. 2A and 2B depict exemplary front-facing and rearward-facing perspectives of the handheld imaging device of FIG. 1, in accordance with various aspects disclosed herein.

[0032] FIGs. 3A and 3B illustrate image data capture and oxygenation analysis performed on the captured image data by the handheld imaging device of FIG. 1, in accordance with various aspects disclosed herein.

[0033] FIG. 4 illustrates an example method for utilizing retinal oximetry to measure blood oxygen levels, in accordance with various aspects disclosed herein.

[0034] FIG. 5 depicts an example user interface display enabling a user to view blood oxygen level mappings within the user’s retinal vasculature and potential TBI assessment, in accordance with various aspects disclosed herein.DETAILED DESCRIPTION

[0035] Retinal oximetry (RO) is a non-invasive imaging technique that simultaneously renders two images of a patient’s retina using two separate filters to measure the relative oxygen saturation (SO2) of retinal vessels by leveraging the spectrophotometric properties of oxygenated and deoxygenated blood. Specifically, the oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb) absorption spectra allow for the isolation of wavelengths where the absorptions of light are similar and different. HbO2 and Hb share an isosbestic wavelength, a wavelength of light that is equally absorbed by both media, at 570 nm. By contrast, the non-isosbestic wavelength that is typically referenced to differentiate between oxidized and deoxidized hemoglobin is 600-601 nm. At this wavelength, Hb absorbs approximately four times the amount of light relative to HbO2, thereby providing an avenue to parse the differences in SO2 between veins and arteries using retinal oximetry.

[0036] Studies evaluating the potential analysis / diagnosis efficacy of retinal oximetry for neurologic conditions have only genuinely emerged in the last decade. Consequently, the clinical adoption of retinal oximetry for neurologic evaluation has been relatively slow. Nevertheless, retinal oximetry has the potential to quickly deliver retinal SO2 and retinal diameter values that proxy cerebral oxygenation completely non-invasively, thereby circumventing the risks associated with conventional PBO2 monitoring techniques, as described above.

[0037] Thus, in accordance with the above, and with the disclosure herein, the present disclosure includes improvements to other technologies or technical fields at least because thepresent disclosure describes or introduces improvements in the field of diagnosing, treating, and monitoring neurological conditions. More specifically, the handheld imaging device(s) of the present disclosure directly advance / improve these fields, as the handheld imaging device(s) of the present disclosure enable non-invasive, unbiased, and efficient analyses of blood oxygen levels / saturation to improve the diagnosis, treatment, and monitoring of neurological conditions, such as TBI. The handheld imaging device(s) of the present disclosure also incorporate various applications / instructions configured to enable any user (e.g., non-medical professionals) to capture high-quality images of a patient’s retinal vasculature and to instruct the user how best to aid the patient. Overall, the handheld imaging dcvicc(s) of the present disclosure represent a non-invasive, portable, and reliable neurological evaluation tool that improves over conventional techniques at least because such conventional techniques lack the portability, reliability, and diagnostic analysis capabilities through non-invasive means. As a result, the handheld imaging device(s) of the present disclosure reduce the health risk and accelerate the treatment of neurological conditions relative to conventional techniques.

[0038] Moreover, the present disclosure includes effecting a transformation or reduction of a particular article to a different state or thing, e.g., transforming or reducing the diagnosis, monitoring, and treatment of TBI (and other neurological conditions) from a non-optimal or error-prone state to an optimal state by eliminating erroneous, biased, and / or otherwise irrelevant blood oxygenation level measurements.

[0039] Still further, the present disclosure includes specific features other than what is well- understood, routine, conventional activity in the field, or adding unconventional steps that demonstrate, in various embodiments, particular useful applications, e.g., determining (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of at least two images; determining an optical density ratio of the retinal vasculature based on the first optical density and the second optical density; and / or determining a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio.

[0040] Moreover, the systems and methods disclosed herein may be primarily described in reference to diagnosis, prognosis, and / or otherwise evaluation of a patient’s risk / likelihood of experiencing some degree or form of TBI. It should be appreciated that this description is forexample purposes only, and that the systems and methods disclosed herein may be implemented, applied, and / or otherwise utilized for the diagnosis, prognosis, and / or otherwise evaluation of a patient’s risk / likelihood of experiencing any suitable neurological condition(s) or combinations thereof.

[0041] To provide a general understanding of the system(s) / components utilized in the techniques of the present disclosure, FIG. 1 illustrates an example environment 100 for utilizing retinal oximetry to measure blood oxygen levels. It should be appreciated that the example environment 100 is merely an example and that alternative or additional components are envisioned.

[0042] In reference to FIG. 1, the example environment 100 may broadly be a hospital, an external environment (e.g., field of play, combat situation), and / or any other suitable location. In particular, the example environment 100 includes the handheld imaging device 101, an external server 106, a user device 108, and a dermal oximeter 109. Broadly, the handheld imaging device 101 may capture image data of a patient’s retinal vasculature and analyze that captured image data to output a TBI likelihood value. The handheld imaging device 101 may capture the image data using the imagers 104a, which may generally be any suitable imagers configured to capture image data at any suitable wavelengths. Moreover, the handheld imaging device 101 may also include one or more processors 104b, a memory 104c storing instructions corresponding to a TBI analysis algorithm 104c 1, a networking interface 104d, and an input / output (I / O) interface 104e. Generally, the one or more processors 104b may interface with the memory 104c to access / execute the instructions stored in the memory 104c, such as the TBI analysis algorithm 104cl, which may include instructions causing the imagers 104a to capture images of a patient’s retinal vasculature.

[0043] Broadly, the imagers 104a may include a digital camera and / or digital video camera for capturing or taking digital images and / or frames. Each digital image may comprise pixel data that may be analyzed in accordance with instructions comprising the TBI analysis algorithm 104c 1, as executed by the one or more processors 104b, as described herein. The digital camera and / or digital video camera of, e.g., the imagers 104a may be configured to take, capture, or otherwise generate digital images and, at least in some embodiments, may store such images in amemory (e.g., one or more memories 104c) of a respective device (e.g., handheld imaging device 101).

[0044] For example, the imagers 104b may be or include a photo-realistic camera (not shown) for capturing, sensing, or scanning 2D image data. The photo-realistic camera may be an RGB (red, green, blue) based camera for capturing 2D images having RGB-based pixel data. The imagers 104a may also process the 2D image data / datasets for use by other devices (e.g., the external server 106, the user device 108). For example, in some embodiments, the one or more processors 104b, 106a may process the image data or datasets captured by the imagers 104a. The processing of the image data may generate post-imaging data that may include metadata, simplified data, normalized data, result data, status data, or alert data as determined from the original captured image data. The image data and / or the post-imaging data may be analyzed by the one or more processors 104b executing, for example, the TBI analysis algorithm 104cl for evaluation, manipulation, and / or otherwise analysis. In other embodiments, the image data and / or the post-imaging data may be sent to a server (e.g., external server 106) for storage or for further analysis.

[0045] The imagers 104a may also be or include two separate imagers configured to capture image data at an isosbestic wavelength and a wavelength featuring notably different absorptions between oxygenated / deoxygenated blood. For example, a first imager may be configured to capture image data at an isosbestic wavelength (e.g., approximately 570 nm), and a second imager may be configured to capture image data at a wavelength featuring notably different absorptions between oxygenated / deoxygenated blood (e.g., approximately 601 nm). Of course, it should be appreciated that the imagers 104a described herein may be configured to capture image data at any suitable wavelength or combinations thereof.

[0046] As part of this data capture, the imagers 104a may also include one or more illumination sources 104al and one or more optical filters 104a2. The one or more illumination sources 104al may be or include light emitting diodes (LEDs) and / or any other suitable illumination source or combinations thereof that are configured to emit specific wavelengths of illumination. Similarly, the one or more optical filters 104a2 may be or include bandpass filters and / or any other suitable optical filters or combinations thereof configured to transmit specific wavelengths of illumination while filtering others. Further, the one or more illumination sources104al may be configured to emit illumination with wavelengths approximately equal to the isosbestic wavelength(s) (e.g., approximately 570 nm) and the wavelength(s) featuring notably different absorptions between oxygenated / deoxygenated blood (e.g., approximately 601 nm), and the one or more optical filters 104a2 may be configured to transmit illumination at these wavelengths (e.g., at least 570 nm and 601 nm). Moreover, it should be appreciated that the illuminations sources 104al and optical filters 104a2 described herein may be configured to emit / filter illumination at any suitable wavelength or combinations thereof.

[0047] In this manner, the handheld imaging device 101 may capture image data of retinal vasculature that is focused on isosbestic and non-isosbestic illumination wavelengths to clearly represent the oxygenation levels of a patient’s retinal vasculature. As previously mentioned, approximately 570 nm is an isosbestic wavelength shared by HbO2 and Hb, such that illumination with a wavelength of 570 nm is nearly equally absorbed by both media. By contrast, approximately 600-601 nm is a non-isosbestic wavelength where Hb absorbs approximately four times the amount of light relative to HbO2. The handheld imaging device 101 may leverage this substantial absorption difference between the two oxygenation levels at the two wavelengths to determine a ratio of oxygenated to deoxygenated blood within the patient’s retinal vasculature. In particular, the handheld imaging device 101 may utilize the TBI analysis algorithm 104c 1 to determine this ratio.

[0048] Generally speaking, the TBI analysis algorithm 104c 1 may include and / or otherwise comprise instructions configured to cause the processors 104b to determine optical densities of the retinal vasculature in the captured image data (e.g., taken at about 570 nm and about 601 nm), and further determine a TBI likelihood value based on these optical densities. More specifically, the TBI analysis algorithm 104c 1 may initially cause the processors 104b to determine a first optical density of the retinal vasculature represented in a first image of the captured image data and to determine a second optical density of the retinal vasculature represented in a second image of the captured image data. Consequently, the TBI analysis algorithm 104c 1 may also be or include instructions that comprise image analysis algorithms, such as image segmentation, edge detection, feature detection, object recognition, and / or any other suitable image analysis algorithms or combinations thereof.

[0049] For example, the first image may be captured (e.g., by the first imager of imagers 104a) at about 570 nm and the second image may be captured (e.g., by the second imager of imagers 104a) at about 601 nm, and / or the first image may be captured at about 601 nm and the second image may be captured at about 570 nm. In any event, the processors 104b executing the TBI analysis algorithm 104c 1 may determine these optical density values at both wavelengths (-570 nm and -601 nm) based on the relative intensities of the incident light emitted by the one or more illumination sources 104al and the intensities of the light passing through the one or more optical filters 104 l and received by the imagers 104a.

[0050] When the TBI analysis algorithm 104c 1 causes the one or more processors 104b to determine the optical densities for each captured image, the algorithm 104c 1 may then proceed to instruct the processors 104b to determine the TBI likelihood value. The TBI analysis algorithm 104c 1 may instruct the processors 104b to determine an optical density ratio of the retinal vasculature based on the first optical density and the second optical density. The optical density ratio may generally be a ratio of the first optical density to the second optical density (e.g., first:second) or the second optical density to the first optical density (e.g., second:first), depending on which wavelength is represented by which optical density value. For example, if the 570 nm wavelength is represented by the first optical density value and the 601 nm wavelength is represented by the second optical density value, then the optical density ratio may be the ratio of the first optical density value to the second optical density value. Of course, it should be appreciated that either mathematical representation of the optical density ratio (e.g., 570:601 and / or 601:570) may be and / or represent a value that the TBI analysis algorithm 104cl may cause the processors 104b to utilize to determine the TBI likelihood value.

[0051] Regardless, with the optical density ratio, the TBI analysis algorithm 104cl may cause the one or more processors 104b to determine the TBI likelihood value corresponding to the optical density ratio. The TBI likelihood value may generally be and / or represent a likelihood that the patient whose retinal vasculature is represented in the captured images (e.g., the captured image data) may be experiencing and / or otherwise suffering from some form and / or degree of TBI. In certain embodiments, the TBI likelihood value may be retrieved by the processors 104b from a reference table and / or other suitable source that lists TBI likelihood values based on optical density ratios and / or other suitable values. For example, the one or more processors 104b may determine a first optical density ratio, and the TBI analysis algorithm 104c 1 may cause theprocessors 104b to access a TBI likelihood reference table to determine a first TBI likelihood value that corresponds to the first optical density ratio.

[0052] In some embodiments, the TBI analysis algorithm 104c 1 may be, include, and / or otherwise utilize one or more machine learning (ML) models / techniques to cause the processors 104b to determine the TBI likelihood value based on the optical density ratio. For example, in these embodiments, the TBI analysis algorithm 104c 1 may be or include a ML model that is trained using a training data set consisting and / or otherwise including image data of retinal vasculature, training optical density values, and / or TBI likelihood values. Thus, when the imagers 104a capture the images of a patient’s retinal vasculature, the TBI analysis algorithm 104c 1 may cause the processors 104b to apply the trained ML model to the captured images to determine optical density values of the retinal vasculature, determine optical density ratios, determine TBI likelihood values based on the optical density ratios, and / or to determine a diagnosis based on the TBI likelihood values, optical densities, optical density ratios, images, and / or any other suitable data or combinations thereof.

[0053] More specifically, the TBI analysis algorithm 104c 1 may further include a training module (not shown) configured to utilize artificial intelligence and / or machine learning techniques to train a TBI analysis model (not shown). The training module may generally employ supervised or unsupervised machine learning techniques, which may be followed or used in conjunction with reinforced or reinforcement learning techniques. As noted above, in some embodiments, the handheld imaging device 101 or other computing device may be configured to implement machine learning, such that the device 101 “learns” to analyze, organize, and / or process data through the TBI analysis model as part of the TBI analysis algorithm 104c 1 without being explicitly programmed. Thus, the training module may train the TBI analysis model to automatically analyze captured image data of a patient’s retinal vasculature and determine a TBI likelihood value based on the captured image data.

[0054] In some embodiments, at least one of a plurality of machine learning methods and algorithms may be applied, which may include but are not limited to: linear or logistic regression, instance-based algorithms, regularization algorithms, decision trees, Bayesian networks, naive Bayes algorithms, cluster analysis, association rule learning, neural networks (e.g., convolutional neural networks (CNN), deep learning neural networks, combined learningmodule or program), deep learning, combined learning, reinforced learning, dimensionality reduction, support vector machines, k-nearest neighbor algorithms, random forest algorithms, gradient boosting algorithms, Bayesian program learning, voice recognition and synthesis algorithms, image or object recognition, optical character recognition, natural language understanding, and / or other ML programs / algorithms either individually or in combination. In various embodiments, the implemented machine learning methods and algorithms are directed toward at least one of a plurality of categorizations of machine learning, such as supervised learning, unsupervised learning, and reinforcement learning.

[0055] In one embodiment, the training module may employ supervised learning techniques, which involves identifying patterns in existing data to make predictions about subsequently received data. Specifically, the training module may “train” the TBI analysis model using training data, which includes example inputs (e.g., training image data of retinal vasculatures, training optical densities / ratios) and associated example outputs (e.g., training TBI likelihood values). Based upon the training data, the training module may cause the TBI analysis model to generate a predictive function which maps outputs to inputs and may utilize the predictive function to generate machine learning outputs based upon data inputs. The example inputs and example outputs of the training data may include any of the data inputs or machine learning outputs described above. In the example embodiment, a processing clement may be trained by providing it with a large sample of data with known characteristics or features.

[0056] In another embodiment, the training module may employ unsupervised learning techniques, which involves finding meaningful relationships in unorganized data. Unlike supervised learning, unsupervised learning does not involve user-initiated training based upon example inputs with associated outputs. Rather, in unsupervised learning, the training module may cause the TBI analysis model to organize unlabeled data according to a relationship determined by at least one machine learning method / algorithm employed by the training module. Unorganized data may include any combination of data inputs and / or machine learning outputs as described above.

[0057] In yet another embodiment, the training module may employ reinforcement learning techniques, which involves optimizing outputs based upon feedback from a reward signal. Specifically, the training module may cause the TBI analysis model to receive a user-definedreward signal definition, receive a data input, utilize a decision-making model to generate a machine learning output based upon the data input, receive a reward signal based upon the reward signal definition and the machine learning output, and alter the decision-making model so as to receive a stronger reward signal for subsequently generated machine learning outputs. Of course, other types of machine learning techniques may also be employed, including deep or combined learning techniques.

[0058] After training, the TBI analysis model and / or other machine learning programs (or information generated by such machine learning programs) may be used to evaluate additional data. Such data may be and / or may be related to captured image data and / or other data that was not included in the training dataset. The trained machine learning programs (or programs utilizing models, parameters, or other data produced through the training process) may accordingly be used for determining, assessing, analyzing, predicting, estimating, evaluating, or otherwise processing new data not included in the training dataset. Such trained machine learning programs (e.g., trained TBI analysis model) may, therefore, be used to perform part or all of the analytical functions of the methods described elsewhere herein.

[0059] It is to be understood that supervised machine learning and / or unsupervised machine learning may also comprise retraining, relearning, or otherwise updating models with new, or different, information, which may include information received, ingested, generated, or otherwise used over time. Further, it should be appreciated that, as previously mentioned, the training module may train the TBI analysis model to output TBI likelihood values, vessel diameters of vessels represented in the imaged retinal vasculature, and / or any other values or combinations thereof using artificial intelligence (e.g., a machine learning model of the TBI analysis algorithm 104cl) or, in alternative aspects, without using artificial intelligence.

[0060] Moreover, although the methods described elsewhere herein may not directly mention machine learning techniques, such methods may be read to include such machine learning for any determination or processing of data that may be accomplished using such techniques. In some aspects, such machine learning techniques may be implemented automatically upon occurrence of certain events or upon certain conditions being met. In any event, use of machine learning techniques, as described herein, may begin with training a machine learning program, or such techniques may begin with a previously trained machine learning program. Further, thesystems and methods described herein may additionally or alternatively leverage trained mathematical models to accomplish any determination or processing of data described herein.

[0061] Further, in certain embodiments, the TB1 analysis algorithm 104c 1 may include instructions that cause the one or more processors 104b to generate, based on the TBI likelihood value, at least one of: (i) a treatment recommendation, (ii) a predicted diagnosis, (iii) a predicted prognosis, (iv) a retinal vasculature map, and / or (v) a vessel diameter for one or more of the blood vessels represented in the captured images. For example, the TBI likelihood value may indicate that the patient is likely experiencing some form / degree of TBI, and should remain relatively stationary and seek medical attention. In this example, the TBI analysis algorithm 104c 1 may also cause the processors 104b to generate a predicted prognosis that, with appropriate medical attention, the patient will likely recover without developing / sustaining any long-term side effects.

[0062] Any and / or all of these values may also be output for display to the user / patient. Namely, the TBI analysis algorithm 104c 1 may include instructions causing the processors 104b to render the TBI likelihood value on a display of the handheld imaging device 101 (e.g., via the VO interface 104e). In some embodiments, the TBI analysis algorithm 104cl may also cause the processors 104b to render and / or otherwise output the treatment recommendation the predicted diagnosis and / or the predicted prognosis for display to the user via the VO interface 104e. Of course, it should be appreciated that the TBI analysis algorithm 104c 1 may include instructions that cause the one or more processors 104b to output any of the values and / or other outputs (e.g., optical density values, optical density ratios, TBI likelihood value, predicted diagnosis, predicted prognosis, treatment recommendation, retinal vasculature maps, vessel diameters, etc.).

[0063] However, in certain circumstances, the user operating the handheld imaging device 101 may not capture and / or otherwise be unable to capture images of the patient’s retinal vasculature. For example, the user may misalign the imagers 104a field of view (FOV) with the patient’s retina, such that the resulting images do not feature and / or otherwise do not include a sufficient representation of the patient’s retinal vasculature for the instructions executed by the processors 104b from the TBI analysis algorithm 104c 1 to generate some / all of the above-mentioned values / outputs. Consequently, the TBI analysis algorithm 104cl may also include instructions comprising an image alignment algorithm and / or an image stabilization algorithm.

[0064] The image alignment algorithm may be or include instructions configured to cause the processors 104b to analyze the captured images from the imagers 104a and determine that a user has misaligned the handheld device based on one or more images captured by the one or more imagers. The image alignment algorithm may also include instructions configured to cause the processors 104b to thereafter determine one or more alignment instructions to help the user align the handheld device based on the one or more images; and output the one or more alignment instructions for display to the user. For example, a user may misalign the handheld imaging device 101 near a patient’s eye, such that the user may need to hold the device 101 further to the left to capture a suitable image of the patient’s retinal vasculature. The image alignment algorithm may instruct the processors to analyze an image captured from this misaligned perspective, recognize that the user should adjust the position of the device 101 to the left, and may generate an instruction for display to the user indicating that the user should move the device 101 to the left. Such an instruction may be or include a textual indication on the device 101 display (e.g., via I / O interface 104e), a verbal indication made through a device 101 speaker (e.g., as part of I / O interface 104e), a haptic feedback indication, a graphical indication (e.g., a left-facing arrow), and / or any other suitable indication(s) or combinations thereof.

[0065] The image stabilization algorithm may be or include instructions configured to cause the processors 104b to analyze the imagers 104a FOV prior to image capture and / or the captured images. More particularly, the image stabilization algorithm may be configured to stabilize the capturing of images by the imagers 104a, and / or may determine whether additional images should be captured based on an assessed stability of the captured images. The image stabilization algorithm may cause the processors 104b to employ any suitable stabilization routine to stabilize image capture and / or to adjust the captured images to account for unstable image capture circumstances. For example, the image stabilization algorithm may include instructions that cause the processors 104b to analyze / evaluate the focus / clarity of images captured by the imagers 104a. If the processors 104b determine that the captured images are of insufficient clarity, then the image stabilization algorithm may cause the one or more processors 104b to adjust the captured image and / or generate instructions for the user to help the user capture a clearer subsequent image.

[0066] In any event, and as described herein, the handheld imaging device 101 may generally perform some / all of this image capture and analysis locally. However, in certain embodiments,the device 101 may utilize external devices to perform some / all of the actions described herein. For example, the handheld imaging device 101 may capture and analyze images of a patient’s retinal vasculature, and may also transmit the captured images to an external server 106 for additional, independent analysis. The external server 106 may include a processor 106a, a memory 106b, a networking interface 106c, and a version of the TBI analysis algorithm 106b 1. In these embodiments, the external server 106 may receive the captured images from the handheld imaging device 101, analyze the images, and output a second TBI likelihood value, in accordance with the instructions included as part of the TBI analysis algorithm 106bl. The external server 106 may then transmit the second TBI likelihood value to the handheld imaging device 101, where the device 101 may proceed to compare the TBI likelihood value determined by the device 101 (e.g., a “first” TBI likelihood value) to the second TBI likelihood value to determine a composite TBI likelihood value. The handheld imaging device 101 may then output the composite TBI likelihood value for display to the user. The composite TBI likelihood value may be any suitable mathematical composition of the first TBI likelihood value and the second TBI likelihood value, such as an average, a weighted average, and / or any other suitable value or combinations thereof. Further, in certain embodiments, the external server 106 may be a remote server that is part of a cloud-based platform. Moreover, as referenced herein a “remote” server, such as the external server 106, may be or include any server that is configured as part of a cloud-based platform, stand-alone architecture, and / or any other suitable architecture and / or combinations thereof.

[0067] In some embodiments, and as illustrated in FIG. 1, the handheld imaging device 101 may also be communicatively coupled with a user device 108 (e.g., via network 116). The user device 108 may generally be any suitable device, and may include a processor 108a, a memory 108b, a networking interface 108c, and an I / O interface 108d. More specifically, the user device 108 may be a wearable device that is configured to monitor and / or otherwise record the blood oxygen levels of the user wearing the device 108. For example, the user device 108 may be a smartwatch and / or other smart device that is configured to measure blood oxygen levels through any suitable measurement methodology. In these embodiments, the handheld imaging device 101 may connect to the user device 108, receive a baseline blood oxygenation value for a patient, and determine the TBI likelihood value for the patient based on the optical density ratio from the captured images and the baseline blood oxygenation value from the user device 108. Of course,in certain instances, the user device 108 may also be another device that may not be a wearable device and / or capable of measuring blood oxygen levels, such as a smartphone. In any case, the handheld imaging device 101 may additionally, or alternatively, transmit and cause the user device 108 to display any relevant values on a display of the user device 108 (e.g., via I / O interface 108d).

[0068] Moreover, in certain embodiments, the handheld imaging device 101 may be communicatively coupled with a dermal oximeter device 109. The dermal oximeter device 109 may generally be any suitable dermal / pulse oximeter device that is configured to determine blood oxygenation values via light emission through a patient’s skin. The handheld imaging device 101 may connect to the dermal oximeter device 109, receive blood oxygen saturation data from the dermal oximeter device 109, calculate an oximetry calibration value based on the blood oxygen saturation data and the optical density ratio, and adjust subsequent data from the dermal oximeter device 109 based on the oximetry calibration value. For example, the handheld imaging device 101 may determine an optical density ratio based on captured image(s), and the dermal oximeter device 109 may determine a blood oxygen saturation value based on the pulse oximetry techniques previously described. The handheld imaging device 101 may determine that the optical density ratio and the blood oxygen saturation value indicate different levels of apparent blood oxygenation, and that the level represented by the optical density ratio is more accurate. As a result, the device 101 may calculate the oximetry calibration value based on the optical density ratio, the blood oxygen saturation value, and / or the blood oxygenation values represented therein to adjust subsequent data received and / or otherwise output by the dermal oximeter device 109. By calculating / applying this oximetry calibration value, the handheld imaging device 101 may alleviate concerns related to racial bias in the corresponding blood oxygenation values output by the dermal oximeter device because the optical density ratio used to calculate the oximetry calibration value does not suffer from the same skin pigmentation bias that the blood oxygen saturation data output by the dermal oximeter device 109 does.

[0069] The memory 104c may include one or more forms of volatile and / or non-volatile, fixed and / or removable memory, such as read-only memory (ROM), electronic programmable readonly memory (EPROM), random access memory (RAM), erasable electronic programmable read-only memory (EEPROM), and / or other hard drives, flash memory, MicroSD cards, and others.

[0070] The I / O interface 104e may include a display screen (not shown) and I / O components (not shown) (e.g., ports, capacitive or resistive touch sensitive input panels, keys, buttons, lights, LEDs). According to some aspects, a user may access the handheld imaging device 101 and / or execute / perform any of the actions described herein via the I / O interface 104e to review outputs from execution of the TBI analysis algorithm 104c 1, make various selections, and / or otherwise interact with the handheld imaging device 101.

[0071] As previously mentioned, and in some embodiments, the handheld imaging device 101 may perform the functionalities as discussed herein as part of a “cloud” network or may otherwise communicate with other hardware or software components within the cloud to send, retrieve, or otherwise analyze data. Thus, it should be appreciated that the external server 106, or indeed any components of the example environment 100 may be in the form of a distributed cluster of computers, servers, machines, or the like. In this implementation, a user may utilize the distributed example environment 100 as part of an on-demand cloud computing platform. Accordingly, when the user interfaces with the example environment 100 (e.g., by interacting with an input component of the I / O interface 104e, 108d), the example environment 100 may actually interface with one or more of a number of distributed computers, servers, machines, or the like, to facilitate the described functionalities.

[0072] Moreover, as illustrated in FIG. 1, the handheld imaging device 101 may communicate and interface with various external devices (e.g., external server 106, user device 108, dermal oximeter device 109) via a network(s) (e.g., network 116). The network(s) used to connect the devices / components of the example environment 100 may support any type of data communication via any standard or technology (e.g., GSM, CDMA, TDMA, WCDMA, LTE, EDGE, OFDM, GPRS, EV-DO, UWB, Internet, IEEE 802 including Ethernet, WiMAX, Wi-Fi, Bluetooth, and others).

[0073] Additionally, it is to be appreciated that a computer program product in accordance with an aspect may include a computer usable storage medium (e.g., standard random access memory (RAM), an optical disc, a universal serial bus (USB) drive, or the like) having computer-readable program code embodied therein, wherein the computer-readable program code may be adapted to be executed by the processor(s) 104b to facilitate the functions as described herein. In this regard, the program code may be implemented in any desired language,and may be implemented as machine code, assembly code, byte code, interpretable source code or the like (e.g., via Golang, Python, Scala, C, C++, Java, Actionscript, Objective-C, Javascript, CSS, XML). In some aspects, the computer program product may be part of a cloud network of resources.

[0074] FIGs. 2A and 2B depict an exemplary front-facing perspective 200 and a rearwardfacing perspective 210 of the handheld imaging device 101 of FIG. 1, in accordance with various aspects disclosed herein. Namely, the exemplary front-facing perspective 200 and the exemplary rearward-facing perspective 210 depict various hardware components of the handheld imaging device 101 that enable the image capture functionality described herein. Of course, it should be appreciated, the handheld imaging device 101 depicted by the exemplary front-facing perspective 200 and the exemplary rearw ard-lacing perspective 210 is for the purposes of discussion only, and the handheld imaging device 101 may include any suitable hardware components in any suitable configuration, as discussed herein.

[0075] The front-facing perspective 200 depicts an imager 202, a filter 204, an illumination source 206, and a grip portion 208. A user may grip the grip portion 208 of the device and may interact with the imager 202 to initiate capture of images of a patient’s retinal vasculature. For example, the imager 202 illustrated in FIGs. 2A and 2B may be a smartphone or other device with an integrated imager / camera. In this example, the user may interact with the interface of the imager 202 to initiate an image capture sequence. As part of the image capture sequence, the imager 202 may execute instructions that cause the illumination source 206 to emit illumination primarily at specified wavelengths (e.g., 570 nm, 601 nm), and the imager 202 may capture this illumination as it reflects / returns from the patient’s retina.

[0076] However, prior to being captured by the imager 202, the illumination may pass through the filter 204. As mentioned, the filter 204 may be a bandpass filter and / or any other suitable type of filter, and may generally filter the illumination travelling towards the imager 202 to further refine the illumination emitted from the illumination source 206. For example, the illumination emitted from the illumination source 206 may initially have a range of wavelengths near the wavelengths of interest (e.g., 570nm and 601 nm). When the illumination reflects from the patient’s retinal vasculature and returns to the filter 204, the filter 204 may only allow a smaller range of wavelengths near the wavelengths of interest to pass through to the imager 202,thereby further clarifying the resulting images to feature the wavelengths of interest more predominantly.

[0077] In any event, a user may place their smartphone or other device that may serve as the imager 202 in connection with the filtering device 201 (e.g., a combination of the filter 204, illumination source 206, and the grip portion 208), and in certain embodiments, may communicatively couple the imager 202 to the filtering device 201. For example, the filtering device may also include a networking interface (not shown) that enables the filtering device 201 to receive / transmit data from / to at least the imager 202. The user may configure the imager 202 (e.g., through an application executable on the imager 202) to communicate with the filtering device 201, and thereby enable the imager 202 to initiate the image capture sequence. Of course, it should be appreciated that the image capture sequence may be initiated in any suitable manner, such as by an interactive button, switch, trigger, interface, and / or other means on the filtering device 201 and / or from any other device or combinations thereof.

[0078] FIGs. 3A and 3B illustrate image data capture and oxygenation analysis performed on the captured image data by the handheld imaging device 101 of FIG. 1, in accordance with various aspects disclosed herein. Broadly speaking, FIG. 3A illustrates an image data capture sequence 300, whereby the handheld imaging device 101 is positioned proximate to a patient’s eye and the handheld imaging device 101 captures image(s) of the patient’s retinal vasculature. FIG. 3B illustrates an oxygenation analysis sequence 330, in which the handheld imaging device 101 analyzes the captured image(s) from the image data capture sequence 300 to determine TBI likelihood values.

[0079] In reference to FIG. 3A, the example image data capture sequence 300 includes the handheld imaging device 101 capturing image(s) of a patient’s eye 302, and more particularly the retinal vasculature of the patient’s eye 302. The handheld imaging device 101 may be positioned proximate to the eye 302, and may be oriented to capture image(s) of the vasculature supplying blood to the eye 302 through the retina 308. As illustrated in FIG. 3 A, these captured images may feature a plurality of various blood vessels, which the TBI analysis algorithm 104cl may be configured to detect and analyze accordingly.

[0080] Briefly, the eye 302 is generally comprised of, inter alia, a sclera 304, a choroid 306, and a retina 308 that is pail of the optic nerve 324. The internal carotid artery (ICA) 303 hasmultiple smaller artery segments that pass through the retina 308, such that the handheld imaging device 101 may capture images of these artery segments. In particular, the artery segments that flow into the eye 302 may do so primarily through the lamina cribrosa 310. However, several other ICA 303 segments may also attach to and / or otherwise contact the eye 302. For example, the short posterior ciliary arteries 312, the long posterior ciliary artery 314, the lateral posterior ciliary artery 316, the central retinal artery 318, the medial posterior ciliary artery 319, the lacrimal artery 320, and / or the ophthalmic artery 322 may each provide some measure of blood supply to the eye 302 even if they do not pass through the lamina cribrosa 310.

[0081] In any event, the handheld imaging device 101 may capture image data (e.g., via the imagers 104a) of some / all of these arteries 303, 312-322 and / or other blood vessels within the patient’s eye 302. Practically speaking, the images captured by the imagers 104a of the handheld imaging device 101 may focus more on the blood vessels that pass through the lamina cribrosa 310, and may less clearly observe any arteries or other blood vessels located further from the eye 302 and closer to the ICA 303. Regardless, the TBI analysis algorithm 104cl may analyze these captured images to evaluate the oxygenation of the patient’ s blood, and thereby determine whether the patient is likely experiencing TBI and / or other medical conditions.

[0082] As illustrated in FIG. 3B, the handheld imaging device 101 may analyze captured images of a patient’s eye by analyzing the retinal vasculature represented in the captured images. The oxygenation analysis sequence 330 depicts the handheld imaging device 101 executing instructions included as part of the TBI analysis algorithm 104c 1 to analyze the oxygenated image 332 and the deoxygenated image 334. The oxygenated image 332 and the deoxygenated image 334 generally refers to the wavelength of light corresponding to the respective image. Namely, the oxygenated image 332 may represent an interior portion of the patient’s eye when imaged at approximately 570 nm, and the deoxygenated image 334 may represent the interior portion of the patient’s eye when imaged at approximately 600-601 nm.

[0083] As previously mentioned, these wavelengths are significant for Hb and HbO2, due to their respective isosbestic and non-isosbestic natures. Thus, the oxygenated image 332 appears to have more strongly / boldly (e.g., optically dense) represented arteries and / or blood vessels because oxygenated and deoxygenated blood absorb / remit light relatively similarly at approximately 570 nm. By contrast, the arteries and / or blood vessels in the deoxygenated image334 appear less strong / bold (e.g., less optically dense) than those same arteries and / or blood vessels in the oxygenated image 332, as the oxygenated blood absorbs approximately four times the amount of light at about 600-601 nm than the deoxygenated blood. Consequently, the optical density of the arteries and / or other blood vessels in the oxygenated image 332 is significantly higher than the optical density of the arteries and / or other blood vessels in the deoxygenated image 334 due to the relative lack of oxygenated blood represented in the deoxygenated image 334.

[0084] For example, the handheld imaging device 101 may capture the oxygenated image 332 by emitting illumination of approximately 570 nm from the illumination sources 104al, and may capture the deoxygenated image 334 by emitting illumination of approximately 600-601 nm from the illumination sources 104al. These image 332, 334 captures may be nearly simultaneous to avoid / minimize relative movement or displacement of the retinal vasculature in the two images 332, 334. The TBI analysis algorithm 104c 1 may instruct the processors 104b to analyze the captured image 332, 334 by analyzing a portion (e.g., portions 332a, 334a) of the respective images 332, 334 that features a dense collection of arteries and / or other blood vessels. The TBI analysis algorithm 104c 1 may then cause the processors 104b to determine an optical density of each individual artery and / or other blood vessel represented in the images 332, 334 and / or within the respective image portions 332a, 334a. These optical densities may generally represent an average optical density of the arteries and / or other blood vessels in the respective images 332, 334 and / or portions 332a, 334a thereof and / or may be a set of optical densities, in which the processors 104b determine an optical density for each distinct artery and / or blood vessel in the respective images 332, 334 and / or portions 332a, 334a thereof. In certain embodiments, the TBI analysis algorithm 104c 1 may also cause the processors 104b to determine a vessel diameter and / or other suitable dimension of the arteries and / or other blood vessels represented in the respective images 332, 334 and / or image portions 332a, 334a.

[0085] When the processors 104b have determined the optical density of the arteries and / or other blood vessels represented in the images 332, 334 and / or potions 332a, 334a thereof, the TBI analysis algorithm 104c 1 may further cause the processors 104b to compare the optical densities to determine an optical density ratio. For example, the processors 104b may determine that the arteries and / or other blood vessels represented in the oxygenated image 332 and / or the potion 332a thereof may have a first optical density, and that the arteries and / or other bloodvessels represented in the deoxygenated image 334 and / or the potion 334a thereof may have a second optical density. In this example, the processors 104b may compare the first optical density (or set of optical densities) with the second optical density (or set of optical densities) to determine an optical density ratio that may broadly represent the patient’s blood oxygenation levels. The TBI analysis algorithm 104c 1 may instruct the processors 104b to perform this optical density ratio analysis on an image-by-image basis (e.g., comparing aggregate optical densities of captured images), on an artery-by-artery basis (e.g., comparing optical densities of corresponding individual arteries / vessels), and / or in any other suitable manner.

[0086] It should be appreciated, that the exemplary analysis described herein in reference to FIGs. 3A, 3B, and / or elsewhere in the present disclosure may include any suitable analysis of any suitable number of images and / or portions thereof to analyze, track, and / or otherwise make determinations (e.g., diagnoses, prognoses, severity values) of any suitable condition, such as TBI.

[0087] For example, the exemplary analysis described herein may also apply to the evaluation of ample blood oxygenation levels and / or sufficient oxygen perfusion levels during cardiopulmonary resuscitation (CPR) and / or other suitable conditions. In this example, the user may utilize the device(s) described herein to capture images of a patient’s eye(s), and the device may proceed to determine blood oxygenation values and / or oxygen perfusion levels of the patient using the techniques / analysis described herein. Using the resulting blood oxygenation values and / or oxygen perfusion values, the device may also subsequently determine oxygen supply values corresponding to whether the patient is receiving a sufficient oxygen supply to their brain and / or other vital organs during the cardiac event (e.g., cardiac collapse, cardiac arrest, etc.). As an example, the TBI analysis algorithm / model 104cl may evaluate the captured images to determine optical density values / ratios, which the algorithm / model 104c 1 may further utilize to determine the oxygenation values, oxygen perfusion values, and / or the oxygen supply values.

[0088] Consequently, the device may also make subsequent determinations regarding the continuation and / or modification of any time-sensitive treatments, such as CPR, that may benefit the patient. Furthering the prior example, the device may output values and / or indications that the patient is not receiving an optimal oxygen supply to their brain based on the oxygenation andoxygen perfusion represented in the captured images of the patient’s eye(s). The device may subsequently generate and output treatment adjustment instructions for display to the user. The user may receive these treatment adjustment instructions from the device and may correspondingly decide to increase / decrease the rate, magnitude, and / or other parameters of the resuscitative and / or other treatment efforts to increase the oxygen supplied to the patient’s brain during the cardiac event.

[0089] In certain instances, the user may also capture subsequent images of the patient’s eye(s) following such modifications to the applied treatments (e.g., resuscitative efforts) to receive updated values from the devices described herein. The device may capture subsequent images of the patient’s eye(s) and may perform a subsequent analysis of the captured images and compare the resulting oxygenation values and / or oxygen perfusion values to the corresponding values of the prior analysis. The device may then output an indication of the relative effectiveness, or lack thereof, of the modifications to the applied treatments (e.g., CPR). For example, the device may output values and / or other indications that the user’s modifications to the applied treatments have been effective in increasing the oxygen supply to the patient’s brain and / or other vital organs, and the device may determine / output subsequent adjustment instructions.

[0090] FIG. 4 illustrates an example method 400 for utilizing retinal oximetry to measure blood oxygen levels, in accordance with various aspects disclosed herein. For ease of discussion, some of the various actions included in the method 400 may be optional, and the various actions included in the method 400 may be performed by, for example, the handheld imaging device 101, the external server 106, the user device 108, and / or any suitable components or combinations thereof.

[0091] The method 400 may include capturing at least two images of a retinal vasculature using one or more imagers (e.g., imagers 104a) (block 402). The method 400 may further include determining (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the two images (block 404). The method 400 may further include determining an optical density ratio of the retinal vasculature based on the first optical density and the second optical density (block 406).

[0092] The method 400 may further include determining a TBI likelihood value corresponding to the optical density ratio (block 408). The method 400 may further include outputting the TBI likelihood value for display to a user (block 410).

[0093] In some embodiments, the handheld imaging device may further comprise: one or more illumination sources configured to emit illumination with wavelengths of at least 570 nm and 601 nm; and one or more bandpass filters configured to transmit illumination at wavelengths of at least 570 nm and 601 nm.

[0094] In certain embodiments, the method 400 may further include determining the TBI likelihood value by applying a trained mathematical model and / or a trained machine learning (ML) model to the optical density ratio value.

[0095] In some embodiments, the method 400 may further include generating, based on the TBI likelihood value, at least one of: (i) a treatment recommendation, (ii) a predicted diagnosis, or (iii) a predicted prognosis; and outputting the treatment recommendation, the predicted diagnosis, or the predicted prognosis for display to the user.

[0096] In certain embodiments, the TBI likelihood value may be a first TBI likelihood value, and the method 400 may further include transmitting the at least two images to a remote server configured to analyze retinal images; receiving a second TBI likelihood value from the remote server; comparing the first TBI likelihood value to the second TBI likelihood value to determine a composite TBI likelihood value; and outputting the composite TBI likelihood value for display to the user.

[0097] In some embodiments, the method 400 may further include determining that a user has misaligned the handheld device based on one or more images captured by the one or more imagers; determining one or more alignment instructions to help the user align the handheld device based on the one or more images; and outputting the one or more alignment instructions for display to the user.

[0098] In certain embodiments, the method 400 may further include connecting with a wearable device; receiving, from the wearable device, a baseline blood oxygenation value; and determining the TBI likelihood value based on the optical density ratio and the baseline blood oxygenation value.

[0099] In some embodiments, the method 400 may further include: prior to capturing the at least two images, execute an image stabilization algorithm configured to stabilize the capturing of the at least two images.

[0100] In certain embodiments, the one or more imagers may include at least two imagers.

[0101] In some embodiments, the method 400 may further include generating a retinal vasculature map that visually indicates relative oxygenation values of the retinal vasculature.

[0102] In certain embodiments, the method 400 may further include determining a vessel diameter of the retinal vasculature based on the at least two images.

[0103] In some embodiments, the method 400 may further include connecting to a dermal oximeter device; receiving blood oxygen saturation data from the dermal oximeter device; calculating an oximetry calibration value based on the blood oxygen saturation data and the optical density ratio; and adjusting subsequent data from the dermal oximeter device based on the oximetry calibration value.

[0104] FIG. 5 depicts an example user interface display 500 enabling a user to view blood oxygen level mappings within the user’s retinal vasculature and potential TBI assessment, in accordance with various aspects disclosed herein. In particular, as illustrated in FIG. 5, the user device (e.g., handheld imaging device 101, user device 108) may render a display that includes a graphical display portion 502, a textual display portion 504, and an interactive button display portion 506.

[0105] The graphical display portion 502 may include retinal vasculature map that visually indicates relative oxygenation values of the retinal vasculature. The retinal vasculature map may feature various graphical indications (e.g., colors, patterns, markings, gradations, etc.) indicating which arteries and / or other blood vessels may have been analyzed as part of the execution of the TBI analysis algorithm (e.g., TBI analysis algorithm 104cl). The graphical display portion 502 may be interactive, such that a user may interact (e.g., click, tap, swipe, touch, gesture, etc.) with the graphical display portion 502, and the example user interface display 500 may display additional and / or otherwise different information than illustrated on the graphical display portion 502. For example, a user may interact with the graphical display portion 502 by tapping on an individual artery within the retinal vasculature map, and as a result, the processors may cause theexample user interface display 500 to display additional information concerning the individual artery, such as the individual oxygenation level within the artery, implications of the oxygenation level (e.g., potential medical implications, such as TBI), and / or any other suitable data / information or combinations thereof.

[0106] Further, each of the arteries, blood vessels, and / or any other suitable information / data / objects represented on the graphical display portion 502 may be and / or otherwise include any suitable type of text, symbols, patterns, colors, and / or any other suitable visual indicia. For example, as illustrated in FIG. 5, the recognized arteries / vessels of the patient’s eye that may have been analyzed as part of the execution of the TBI analysis algorithm may be marked with colors and / or patternings indicating relative oxygenation levels of blood flowing through those arteries / vessels. Moreover, each object represented on the graphical display portion 502 may be or include an image, video, and / or any other suitable visual display configuration.

[0107] Further, in certain embodiments, the graphical display portion 502 may include indications that represent strength or other gradient values corresponding to the data displayed in the graphical display portion 502. For example, the graphical indications associated with the arteries / vessels may be or include graphical relative strength indicators (e.g., colors, symbols, graphics, etc.) corresponding to a level of concern a patient should have as a consequence of the oxygenation level(s) represented by the optical densities and / or vessel diameters represented in the patient’s images, numerical representations of the level of concern, textual strength indicators (e.g., “seek immediate medical attention”, “benign”, etc.), and / or any other suitable indicator types or combinations thereof.

[0108] The textual display portion 504 may include a text-based message for a user that corresponds to the display within the graphical display portion 502. For example, as illustrated in FIG. 5, the textual display portion 504 includes text reading “Based on analysis of blood oxygen levels in your retinal vasculature, you may have suffered contact to your head resulting in a traumatic brain injury.” Thus, the text-based message within the textual display portion 504 may enable a user to understand the context of the display within the graphical display portion 502, and as a result, the user may make more informed decisions to seek medicalattention / advice regarding a potential TBI diagnosis. In this manner, the textual display portion 504 may enable the user to alleviate / mitigate risk associated with having / experiencing TBI.

[0109] The interactive button display portion 506 may generally enable a user to view additional information and / or initiate certain additional functionalities corresponding to the information presented in the example user interface display 500. For example, a user may interact with the interactive button display portion 506, and the processor may cause the example user interface display 500 to display relevant retinal vasculature that is present within a patient’s retinal images. These relevant retinal vasculatures may be or include, for example, the retinal vasculature responsible for the generation of the message featured within the textual display portion 504. Additionally, or alternatively, the interactive button display portion 506 may cause the processor to initiate functionality outside of a display application or other application / module where the example user interface display 500 is rendered in the event, for example, that a user / patient may desire to contact a physician’s office to discuss the results of their retinal image analysis and / or any other suitable additional functionality or combinations thereof. As another example, the user may interact with the interactive button display portion 506, and the processor may access the Internet to retrieve and display information related to any oxygenation values, artery / vessel diameters, optical densities, and / or any other values or combinations thereof that arc flagged and / or otherwise determined as relevant within the patient’s image data.ADDITIONAL CONSIDERATIONS

[0110] Throughout this specification, plural instances may implement components, operations, or structures described as a single instance. Although individual operations of one or more methods are illustrated and described as separate operations, one or more of the individual operations may be performed concurrently, and nothing requires that the operations be performed in the order illustrated. Structures and functionality presented as separate components in example configurations may be implemented as a combined structure or component. Similarly, structures and functionality presented as a single component may be implemented as separate components. These and other variations, modifications, additions, and improvements fall within the scope of the subject matter herein.

[0111] Additionally, certain embodiments are described herein as including logic or a number of routines, subroutines, applications, or instructions. These may constitute eithersoftware (e.g., code embodied on a machine-readable medium or in a transmission signal) or hardware. In hardware, the routines, etc., are tangible units capable of performing certain operations and may be configured or arranged in a certain manner. In example embodiments, one or more computer systems (e.g., a standalone, client or server computer system) or one or more hardware modules of a computer system (e.g., a processor or a group of processors) may be configured by software (e.g., an application or application portion) as a hardware module that operates to perform certain operations as described herein.

[0112] In various embodiments, a hardware module may be implemented mechanically or electronically. For example, a hardware module may comprise dedicated circuitry or logic that is permanently configured (e.g., as a special-purpose processor, such as a field programmable gate array (FPGA) or an application- specific integrated circuit (ASIC)) to perform certain operations. A hardware module may also comprise programmable logic or circuitry (e.g., as encompassed within a general-purpose processor or other programmable processor) that is temporarily configured by software to perform certain operations. It will be appreciated that the decision to implement a hardware module mechanically, in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations.

[0113] Accordingly, the term "hardware module" should be understood to encompass a tangible entity, be that an entity that is physically constructed, permanently configured (e.g., hardwired), or temporarily configured (e.g., programmed) to operate in a certain manner or to perform certain operations described herein. Considering embodiments in which hardware modules are temporarily configured (e.g., programmed), each of the hardware modules need not be configured or instantiated at any one instance in time. For example, where the hardware modules comprise a general-purpose processor configured using software, the general-purpose processor may be configured as respective different hardware modules at different times. Software may accordingly configure a processor, for example, to constitute a particular hardware module at one instance of time and to constitute a different hardware module at a different instance of time.

[0114] Hardware modules can provide information to, and receive information from, other hardware modules. Accordingly, the described hardware modules may be regarded as beingcommunicatively coupled. Where multiple of such hardware modules exist contemporaneously, communications may be achieved through signal transmission (e.g., over appropriate circuits and buses) that connects the hardware modules. In embodiments in which multiple hardware modules are configured or instantiated at different times, communications between such hardware modules may be achieved, for example, through the storage and retrieval of information in memory structures to which the multiple hardware modules have access. For example, one hardware module may perform an operation and store the output of that operation in a memory device to which it is communicatively coupled. A further hardware module may then, at a later time, access the memory device to retrieve and process the stored output. Hardware modules may also initiate communications with input or output devices, and can operate on a resource (e.g., a collection of information).

[0115] The various operations of the example methods described herein may be performed, at least partially, by one or more processors that are temporarily configured (e.g., by software) or permanently configured to perform the relevant operations. Whether temporarily or permanently configured, such processors may constitute processor-implemented modules that operate to perform one or more operations or functions. The modules referred to herein may, in some example embodiments, comprise processor- implemented modules.

[0116] Similarly, the methods or routines described herein may be at least partially processor- implemented. For example, at least some of the operations of a method may be performed by one or more processors or processor- implemented hardware modules. The performance of certain of the operations may be distributed among the one or more processors, not only residing within a single machine, but also deployed across a number of machines. In some example embodiments, the processor or processors may be located in a single location (e.g., within a home environment, an office environment or as a server farm), while in other embodiments the processors may be distributed across a number of locations.

[0117] The performance of certain of the operations may be distributed among the one or more processors, not only residing within a single machine, but also deployed across a number of machines. In some example embodiments, the one or more processors or processor-implemented modules may be located in a single geographic location (e.g., within a home environment, an office environment, or a server farm). In other example embodiments, the one or moreprocessors or processor-implemented modules may be distributed across a number of geographic locations.

[0118] Unless specifically stated otherwise, discussions herein using words such as "processing," "computing," "calculating," "determining," "presenting," "displaying," or the like may refer to actions or processes of a machine (e.g., a computer) that manipulates or transforms data represented as physical (e.g., electronic, magnetic, or optical) quantities within one or more memories (e.g., volatile memory, non-volatile memory, or a combination thereof), registers, or other machine components that receive, store, transmit, or display information.

[0119] As used herein any reference to "one embodiment" or "an embodiment" means that a particular element, feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment.

[0120] Some embodiments may be described using the expression "coupled" and "connected" along with their derivatives. For example, some embodiments may be described using the term "coupled" to indicate that two or more elements are in direct physical or electrical contact. The term "coupled," however, may also mean that two or more elements arc not in direct contact with each other, but yet still co-operate or interact with each other. The embodiments are not limited in this context.

[0121] As used herein, the terms "comprises," "comprising," "includes," "including," "has," "having" or any other variation thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Further, unless expressly stated to the contrary, "or" refers to an inclusive or and not to an exclusive or. For example, a condition A or B is satisfied by any one of the following: A is true (or present) and B is false (or not present), A is false (or not present) and B is true (or present), and both A and B are true (or present).

[0122] In addition, use of the "a" or "an" are employed to describe elements and components of the embodiments herein. This is done merely for convenience and to give a general sense ofthe description. This description, and the claims that follow, should be read to include one or at least one and the singular also includes the plural unless it is obvious that it is meant otherwise.

[0123] While the present invention has been described with reference to specific examples, which are intended to be illustrative only and not to be limiting of the invention, it will be apparent to those of ordinary skill in the art that changes, additions and / or deletions may be made to the disclosed embodiments without departing from the spirit and scope of the invention.

[0124] The foregoing description is given for clearness of understanding; and no unnecessary limitations should be understood therefrom, as modifications within the scope of the invention may be apparent to those having ordinary skill in the art.

Claims

WHAT IS CLAIMED:

1. A handheld imaging device configured to determine retinal oximetry values, comprising: one or more imagers; one or more processors; and one or more memories storing thereon executable instructions that, when executed by the one or more processors, cause the one or more processors to: capture at least two images of a retinal vasculature using the one or more imagers, determine (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images, determine an optical density ratio of the retinal vasculature based on the first optical density and the second optical density, determine a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio, and output the TBI likelihood value for display to a user.

2. The handheld imaging device of claim 1, further comprising: one or more illumination sources configured to emit illumination with wavelengths of at least 570 nm and 601 nm; and one or more bandpass filters configured to transmit illumination at wavelengths of at least 570 nm and 601 nm.

3. The handheld imaging device of either claim 1 or claim 2, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: determine the TBI likelihood value by applying a trained mathematical model or a trained machine learning (ML) model to the optical density ratio.

4. The handheld imaging device of any one of claim 1-3, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: generate, based on the TBI likelihood value, at least one of: (i) a treatment recommendation, (ii) a predicted diagnosis, or (iii) a predicted prognosis; and output the treatment recommendation, the predicted diagnosis, or the predicted prognosis for display to the user.

5. The handheld imaging device of any one of claim 1-4, wherein the TBI likelihood value is a first TBI likelihood value, and the executable instructions, when executed by the one or more processors, further cause the one or more processors to: transmit the at least two images to a remote server configured to analyze retinal images; receive a second TBI likelihood value from the remote server; compare the first TBI likelihood value to the second TBI likelihood value to determine a composite TBI likelihood value; and output the composite TBI likelihood value for display to the user.

6. The handheld imaging device of any one of claim 1-5, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: determine that a user has misaligned the handheld device based on one or more images captured by the one or more imagers; determine one or more alignment instructions to help the user align the handheld device based on the one or more images; and output the one or more alignment instructions for display to the user.

7. The handheld imaging device of any one of claim 1-6, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: connect with a wearable device; receive, from the wearable device, a baseline blood oxygenation value; anddetermine the TBI likelihood value based on the optical density ratio and the baseline blood oxygenation value.

8. The handheld imaging device of any one of claim 1-7, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: prior to capturing the at least two images, execute an image stabilization algorithm configured to stabilize the capturing of the at least two images.

9. The handheld imaging device of any one of claim 1-8, wherein the one or more imagers include at least two imagers.

10. The handheld imaging device of any one of claim 1-9, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: generate a retinal vasculature map that visually indicates relative oxygenation values of the retinal vasculature.

11. The handheld imaging device of any one of claim 1-10, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: determine a vessel diameter of the retinal vasculature based on the at least two images.

12. The handheld imaging device of any one of claim 1-11, wherein the executable instructions, when executed by the one or more processors, further cause the one or more processors to: connect to a dermal oximeter device; receive blood oxygen saturation data from the dermal oximeter device; calculate an oximetry calibration value based on the blood oxygen saturation data and the optical density ratio; andadjust subsequent data from the dermal oximeter device based on the oximetry calibration value.

13. A method for determining retinal oximetry values, the method comprising: capturing, by one or more imagers of a handheld imaging device, at least two images of a retinal vasculature; determining, by one or more processors, (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images; determining, by the one or more processors, an optical density ratio of the retinal vasculature based on the first optical density and the second optical density; determining, by the one or more processors, a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio; and outputting, by the one or more processors, the TBI likelihood value for display to a user.

14. The method of claim 13, further comprising: emitting, by one or more illumination sources of the handheld imaging device, illumination with wavelengths of at least 570 nm and 601 nm during the capturing of the at least two images.

15. The method of either claim 13 or claim 14, further comprising: filtering, by one or more bandpass filters of the handheld imaging device, received illumination to capture wavelengths of at least 570 nm and 601 nm during the capturing of the at least two images.

16. The method of any one of claim 13-15, further comprising: determining, by the one or more processors, the TBI likelihood value by applying a trained mathematical model or a trained machine learning (ML) model to the optical density ratio.

17. The method of any one of claim 13-16, further comprising:generating, by the one or more processors and based on the TBI likelihood value, at least one of: (i) a treatment recommendation, (ii) a predicted diagnosis, or (iii) a predicted prognosis; and outputting, by the one or more processors, the treatment recommendation, the predicted diagnosis, or the predicted prognosis for display to the user.

18. The method of any one of claim 13-17, wherein the TBI likelihood value is a first TBI likelihood value, and the method further comprises: transmitting, by the one or more processors, the at least two images to a remote server configured to analyze retinal images; receiving, at the one or more processors, a second TBI likelihood value from the remote server; comparing, by the one or more processors, the first TBI likelihood value to the second TBI likelihood value to determine a composite TBI likelihood value; and outputting, by the one or more processors, the composite TBI likelihood value for display to the user.

19. The method of any one of claim 13-18, further comprising: determining, by the one or more processors, an oxygen supply value based on the optical density ratio; determining, by the one or more processors, treatment adjustment instructions corresponding to a cardiac event based on the oxygen supply value; and outputting, by the one or more processors, the treatment adjustment instructions for display to the user.

20. A non-transitory computer-readable storage medium having stored thereon a set of instructions, executable by at least one processor, for determining retinal oximetry values, the instructions comprising: instructions for capturing, by one or more imagers of a handheld imaging device, at least two images of a retinal vasculature;instructions for determining (i) a first optical density of the retinal vasculature represented in a first image of the at least two images and (ii) a second optical density of the retinal vasculature represented in a second image of the at least two images; instructions for determining an optical density ratio of the retinal vasculature based on the first optical density and the second optical density; instructions for determining a traumatic brain injury (TBI) likelihood value corresponding to the optical density ratio; and instructions for outputting the TBI likelihood value for display to a user.

Images