Allergen test patches for automated allergic contact dermatitis patch testing

Abstract

An allergen test patch includes a patch substrate having a top surface and a bottom surface, at least one allergen well coupled to the bottom surface of the patch substrate, and a border removably coupled to and surrounding the patch substrate. The border has a top surface and a bottom surface. An adhesive is disposed on the bottom surface of the border to adhere the border to a skin surface of a subject, thereby removably coupling the patch substrate to the skin surface of the subject. When the border is adhered to the skin surface of the subject, removing the patch substrate from the border retains the border on the skin surface of the subject. This design allows for improved allergen testing by maintaining a defined test area on the skin after removal of the allergen-containing substrate.

Description

Mayo 2023-378630666.01678ALLERGEN TEST PATCHES FOR AUTOMATED ALLERGIC CONTACT DERMATITIS PATCH TESTING BACKGROUND

[0001] Allergic contact dermatitis is a leading cause of dermatologic morbidity. The prevalence of allergic reactions to single allergens is up to 20% of the general population. More importantly, it has been recently established that over 63% of people are allergic to at least one substance when tested in a clinical setting with a comprehensive allergen panel (Zawawi et al, 2023). However, the current practice of extended patch testing is inconvenient for patients as it requires at least 3 clinic visits within the same week thereby greatly limiting the number of patients that have access to diagnostic testing.SUMMARY OF THE DISCLOSURE

[0002] According to an aspect of the present disclosure, an allergen test patch is provided. The allergen test patch includes a patch substrate having a top surface and a bottom surface, at least one allergen well coupled to the bottom surface of the patch substrate, and a border removably coupled to and surrounding the patch substrate. The border has a top surface and a bottom surface. An adhesive is disposed on the bottom surface of the border to adhere the border to a skin surface of a subject, thereby removably coupling the patch substrate to the skin surface of the subject. The adhesive retains the border on the skin surface of the subject when removing the patch substrate from the border.

[0003] According to another aspect of the present disclosure, an allergen test patch may include a patch substrate having atop surface and a bottom surface, an allergen well coupled to the bottom surface of the patch substrate, a self-inking marker coupled to the bottom surface of the patch substrate, such that when the patch substrate is adhered to a skin surface of a subj ect the self-inking marker transfers ink to mark the skin surface of the subj ect with a marker that identifies a tested allergen, and at least one adhesive patch coupled to the bottom surface of the patch substrate to removably couple the patch substrate to the skin surface of the subject.

[0004] According to still another aspect of the present disclosure, an allergen test patch may include a patch substrate having a top surface and a bottom surface, a plurality of allergen wells coupled to the bottom surface of the patch substrate, and at least one control fiber coupled to the bottom surface of the patch substrate, the at least one control fiber being composed of a control material to test as a control for allergen reactions.1QB\630666.016781100992613.2Mayo 2023-378630666.01678

[0005] According to yet another aspect of the present disclosure, an allergen test patch may include an annular patch substrate having a top surface and a bottom surface, a plurality of allergen wells coupled to the bottom surface of the annular patch substrate, wherein the plurality of allergen wells are radially distributed around the annular patch substrate, and a fiducial support spanning a central aperture of the annular patch substrate, the fiducial support being shaped to indicate an orientation of the annular patch substrate.

[0006] According to another aspect of the present disclosure, a method for processing patch test images to detect allergen reactions is provided. The method includes accessing, with a computer system, image data including at least one patch test image depicting a test site on a skin surface of a subject, the test site having been exposed to at least one allergen. The method further includes accessing, with the computer system, a trained machine learning model configured to detect presence of allergen reactions in patch test images. The method also includes inputting the image data to the trained machine learning model to generate classified feature data indicative of a detected presence of an allergen reaction at the test site. The method additionally includes at least one of displaying the classified feature data to a user or storing the classified feature data for later use.

[0007] According to another aspect of the present disclosure, a computer-implemented method for manufacturing a patient-specific allergen test patch is provided. The method includes receiving, with a computer system, patient data including at least one of medical history data, environmental exposure data, or suspected allergen reaction data associated with a patient. The method further includes processing the patient data with a recommender algorithm to generate a ranked list of allergens, where the ranked list of allergens identifies allergens having an increased likelihood of causing an allergic reaction for the patient based on the patient data. The method also includes selecting, based on the ranked list of allergens, a patient-specific allergen panel including a subset of allergens from the ranked list of allergens. The method additionally includes manufacturing an allergen test patch including a plurality of allergen wells, where each allergen well of the plurality of allergen wells is loaded with a respective allergen from the patient-specific allergen panel.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] FIG. 1 A is an example allergen test patch configured as a single- allergen test patch.2QB\630666.01678\100992613.2Mayo 2023-378630666.01678

[0009] FIG. IB is an example allergen test well that can be used with the allergen test patches described in the present disclosure.

[0010] FIG. 1C is an obverse side of the example allergen test well shown in FIG. IB.

[0011] FIG. 2 A is an example allergen test patch configured as a single- allergen test patch with a bandage-like form factor.

[0012] FIG. 2B is another example allergen test patch configured as a single-allergen test patch with a bandage-like form factor.

[0013] FIG. 3A is another example allergen test patch configured as a single-allergen test patch with a bandage-like form factor.

[0014] FIG. 3B show s an obverse side of the example allergen test patch of FIG. 3 A.

[0015] FIG. 4 is an example allergen test patch configured as a multi-allergen test patch.

[0016] FIG. 5A is an example allergen test patch configured as a multi-allergen test patch having an annular patch substrate, a top surface of the allergen test patch is illustrated.

[0017] FIG. 5B is an example allergen test patch configured as a multi-allergen test patch having an annular patch substrate, a bottom surface of the allergen test patch is illustrated.

[0018] FIG. 5C shows another example allergen test patch configured as a multiallergen test patch having an annular patch substrate.

[0019] FIG. 5D is an example allergen test patch configured as a multi-allergen test patch having an annular patch substrate, in which a first sequence of skin-inking regions on the bottom surface of the allergen test patch are filled with ink to indicate a first patch identifier.

[0020] FIG. 5E is an example allergen test patch configured as a multi-allergen test patch having an annular patch substrate, in which a second sequence of skin-inking regions on the bottom surface of the allergen test patch are filled with ink to indicate a second patch identifier.

[0021] FIG. 5F is an example allergen test patch configured as a multi-allergen test patch having an annular patch substrate and coupled to an adhesive backing.

[0022] FIG. 5G is an example of a top surface of the adhesive backing illustrated in FIG. 5F.

[0023] FIG. 5H is an example allergen application cradle for retaining a radial allergen test patch while applying allergens to the allergen wells.

[0024] FIG. 6 is another example allergen test patch configured as a multi-allergen test patch.3QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0025] FIGS. 7A-7E show another example allergen test patch configured as a multiallergen test patch. FIG. 7A illustrates the individual components of the allergen test patch. FIG. 7B illustrates applying a patch substrate to a patch border. FIG. 7C illustrates a patch substrate having been applied to a patch border (e.g., via hook-and-loop or other fasteners). FIG. 7D illustrates an allergen test patch in which the patch substrate has been removed from the patch border and replaced with a protective cover. FIG. 7E illustrates allergen test patch in which the patch substrate has been removed and replaced with a fiducial marker template.

[0026] FIG. 8 illustrates an example automated allergen patch testing system.

[0027] FIG. 9 shows a patch test image captured under full spectrum light with no filtering.

[0028] FIG. 10 shows a patch test image of the same patient shown in FIG. 9 but captured under ultraviolet light.

[0029] FIG. 11 shows a patch test image captured under ultraviolet light with iridescent ink marking on the skin surface for localization.

[0030] FIG. 12 shows example patch test images in which allergic contact dermatitis is more visible under ultraviolet light than visible spectrum light.

[0031] FIG. 13 is a flowchart of an example method for processing patch test images to generate classified feature data that indicate the detected presence and / or severity of allergen reactions in the patch test images.

[0032] FIG. 14 is a flowchart of an example method for training a machine learning model to generate classified feature data from patch test images.

[0033] FIG. 15 is a block diagram of an example automated allergen testing system.

[0034] FIG. 16 is a block diagram of example components that can implement the system of FIG. 15.DETAILED DESCRIPTION

[0035] Described here are systems and methods for automated allergic contact dermatitis patch testing. The disclosed technology may provide improved accessibility, efficiency, and accuracy in diagnosing allergic contact dermatitis compared to traditional patch testing methods. In general, the disclosed systems and methods provide integrated technologies to facilitate allergen test site localization, allergen identification, test site protection, and reaction classification during allergic contact dermatitis screening (i.e., '‘patch testing”).4QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0036] In some cases, the present disclosure provides an automated patch testing system that may include specialized allergen test patch designs, image capture devices, reaction detection algorithms, and data management workflows. These components may work together to enable remote or at-home allergen patch testing with automated analysis and interpretation of results.

[0037] The automated patch testing system may allow for more widespread screening of allergic contact dermatitis, potentially increasing early detection and treatment of this condition. By automating aspects of the testing and analysis process, the automated patch testing system may reduce the need for multiple in-person clinic visits and expert visual assessment of allergen patch test results.

[0038] In some implementations, the automated patch testing system may incorporate machine learning algorithms to improve reaction detection accuracy over time. The system may also enable adaptive testing protocols, where subsequent allergen selections are informed by previous test results and patient-specific factors. The disclosed automated patch testing system may provide a more accessible, efficient, and data-driven approach to diagnosing allergic contact dermatitis compared to conventional methods. This may lead to improved patient outcomes and expanded knowledge about allergen prevalence and reactions across diverse populations.

[0039] In some cases, the allergen testing process may begin with one or more allergens that have been selected fortesting. Additionally or alternatively, the testing process may begin with a recommender algorithm for personalized patch testing. This algorithm may analyze patient data, such as medical history, environmental exposures, and previous reactions, to generate a ranked list of allergens most likely to cause a reaction for the specific patient. The algorithm may use this information to recommend which allergens should be included in the patch test for that individual. In some implementations, the recommender algorithm may process data from a Patient Journey instrument, which may be a computer-adaptive, comprehensive patient history' form configured to systematically gather patient-provided information to guide individualized testing for allergens. The Patient Journey instrument may be configured as multiple journeys or sections, such as: a background section for gathering demographic and ancestry' information; a medical history section for documenting relevant health conditions; a work and life environments section for identifying potential allergen exposures; a section for items the patient suspects have caused allergic reactions; and a treatments section for documenting treatments the patient has used for skin conditions. The5QB\630666.01678X100992613.2Mayo 2023-378630666.01678Patient Journey instrument may be implemented as a database-driven application to allow greater flexibility in expanding the depth of information gathered while minimizing application development costs.

[0040] In some implementations, the patient data processed by the recommender algorithm may be received from a Patient Journey instrument. The Patient Journey instrument may be a computer-adaptive, comprehensive patient history form configured to systematically gather patient-provided information to guide individualized testing for allergens. The Patient Journey instrument may be configured as multiple journeys or sections, such as: a background section for gathering demographic and ancestry information; a medical history7section for documenting relevant health conditions; a work and life environments section for identifying potential allergen exposures; a section for items the patient suspects have caused allergic reactions; and a treatments section for documenting treatments the patient has used for skin conditions. The Patient Journey instrument may be implemented as a database-driven application to allow greater flexibility in expanding the depth of information gathered while minimizing application development costs.

[0041] In some implementations, the recommender algorithm may comprise a naive Bayes classifier configured to incorporate a probabilistic structure to rank order allergens in terms of recommendations. Additionally or alternatively, the recommender algorithm may comprise a collaborative filtering algorithm configured to identify allergens associated with high reactivity based on patient histories and observed reactions from other patients with similar histories. The collaborative filtering approach may leverage the observation that patients with similar histories may yield similar patch test results. The naive Bayes classifier may extend the collaborative filtering approach by incorporating a probabilistic structure to the recommendations, allowing allergens to be ranked in terms of likelihood of causing a reaction for the specific patient. In some cases, the recommender algorithm may also utilize known crosswalks of allergens and common products, as well as databases that cross-reference allergens with common products and co-reactivity with other allergens, to inform the allergen recommendations.

[0042] In some implementations, the system may support adaptive, sequential allergen testing where results from a previous allergen test performed on the patient are received and used to update the ranked list of allergens. For example, if a patient shows a reaction to a particular allergen in a previous test, the recommender algorithm may update the ranked list to include allergens related to the allergen that caused the reaction. The selection of related6QB\630666.01678X100992613.2Mayo 2023-378630666.01678allergens may be based on known co-reactivity or cross-reactivity databases that crossreference allergens with other allergens having similar reaction profiles. This adaptive approach may allow for more efficient testing by focusing subsequent tests on allergens with increased likelihood of causing reactions based on the patient’s demonstrated sensitivities.

[0043] The patch designs described in the present disclosure may incorporate features to facilitate accurate image capture and analysis. In some implementations, the patch may include fiducial markers, computer-readable barcodes, or other identifiers to aid in image registration and patch identification during the automated analysis process.

[0044] Once the patch is applied to the patient’s skin, the image capture system may be used to document the test sites at various time points. The image capture system may include specialized lighting and positioning features to ensure consistent, high-quality images suitable for subsequent analysis of the patch test images.

[0045] The captured images may then be processed by reaction detection algorithms. These algorithms may analyze the images to identify and classify potential allergic reactions at each test site. In some cases, the algorithms may use machine learning techniques to detect subtle changes in skin appearance that may indicate a reaction. Additionally or alternatively, the patch test images may be processed to determine a severity of an allergic reaction.

[0046] The data workflow provided by the disclosed systems and methods may manage the flow of information throughout the testing process. Patient data, patch configuration information, captured images, and algorithm results may be integrated into a cohesive dataset. In some implementations, the data workflow may also facilitate secure data storage and retrieval, enabling healthcare providers to access and review test results remotely.

[0047] As mentioned above, in some cases an adaptive, sequential allergen testing workflow may be implemented to further personalize and optimize the testing process. In this approach, the results of initial allergen tests may be used to inform subsequent testing decisions. For example, if a patient shows a strong reaction to a particular allergen, the system may recommend additional testing with related allergens or adjust the concentrations of subsequent tests. The adaptive workflow may work as follows. First, an initial set of allergens may be selected based on a recommender algorithm’s suggestions. The patch with these allergens may then be applied to the patient’s skin. Images may be captured at predetermined intervals using the image capture system. The reaction detection algorithms may then analyze these images to identify any reactions. Based on the detected reactions (or lack thereof), the7QB\630666.01678X100992613.2Mayo 2023-378630666.01678system may generate recommendations for the next round of testing. This process may be repeated, with each iteration refining the allergen selection based on previous results.

[0048] In some cases, the system may automatically generate a report summarizing the test results, including identified reactions, their severity, and potential allergen cross-reactivity. This report may be made available to healthcare providers through a secure interface, allowing for rapid interpretation and communication of results to the patient.

[0049] The integration of these features enables a more efficient and accurate allergic contact dermatitis testing process. By leveraging personalized recommendations, adaptive testing strategies, and automated image analysis, the system may provide a comprehensive approach to identifying allergens while minimizing patient discomfort and reducing the need for repeated clinic visits.

[0050] As described in more detail below with reference to FIGS. 1-7, the allergen test patch 10 described in the present disclosure may include various configurations to facilitate allergic contact dermatitis testing.

[0051] In some embodiments, the allergen test patch 10 may be configured as a singleallergen test patch having a single allergen well or chamber, as illustrated in FIGS. 1-3. In these instances, the allergen test patch 10 will include a single allergen well or chamber. Advantageously, this design may allow' for more targeted testing of individual allergens and / or sequential application of different allergens over time.

[0052] Alternatively, the allergen test patch 10 may be configured as a multi-allergen test patch, as illustrated in FIGS. 4-7. This design may incorporate multiple allergen wells or chambers on a single patch. The allergen wells or chambers may be arranged in a specific pattern on the allergen test patch 10, or may be arbitrarily distributed on the allergen test patch 10. Advantageously, the multi-allergen test patch may be suitable for comprehensive simultaneous testing of multiple potential allergens.

[0053] The allergen wells or chambers within the allergen test patch 10 may be constructed from inert materials, such as aluminum, titanium, medical-grade plastics, or the like. In some implementations, the allergen wells may include Finn Chambers or similar structures. The base matenal of the patch substrate 12 may include ScanPor tape or other suitable hypoallergenic adhesive tape materials whose performance in patch testing is well established. These materials may help prevent unwanted reactions or interactions with the allergens being tested. The size and shape of the allergen wells or chambers may be optimized for proper allergen delivery and skin contact. In some embodiments, the allergen wells may8QB\630666.01678X100992613.2Mayo 2023-378630666.01678have a single wall construction. In still other embodiments, the allergen wells may have a double wall construction. The double-walled allergen wells may include two ridges, or walls, surrounding the well or chamber that receives the allergen. This double walled construction can provide additional barrier from water intrusion into the test site when the allergen test patch 10 is applied to the skin surface of the patient, subject, or other user of the allergen test patch 10.

[0054] In some implementations, the allergen test patch 10 may be configured as a modular two-component system including an allergen carrier and an allergen carrier mount. The allergen carrier may be a single allergen welled structure that can store allergens regardless of viscosity through use of filter paper fitted within the allergen well. The allergen carriers may be mass produced and subsequently adhered to the allergen carrier mount to administer the allergen to the skin surface of the patient. The allergen carrier may include a double rim structure to retain the allergen and limit allergen bleed beyond the intended contact area. A two-part adhesive system may be used to secure the allergen carrier to the allergen carrier mount, with a first part of the adhesive disposed on the allergen carrier and a second part of the adhesive disposed on the allergen carrier mount.

[0055] The allergen carrier mount may be configured to remain on the skin surface of the patient for the entire test duration, such as five days. The allergen carrier mount may serve as a template to guide allergen carrier placement, ensuring consistent positioning of the allergen wells relative to the skin surface. The allergen carrier mount may include color calibration markers 26 and fiducial markings 24 to facilitate image capture and processing. The allergen carrier mount may also serve as a substrate to attach protective coverings during activities such as showering. In some cases, the allergen carrier mount may include an adhesive structure and a tear-away section for allergen retention and removal, allowing the allergen carrier to be removed while the mount remains adhered to the skin surface.

[0056] In certain designs, the allergen test patch 10 may include a removable section containing allergen wells and a retained border piece. The removable section may be detached after a predetermined testing period, while the border piece remains on the skin to aid in documenting test sites at later time points. In some implementations, the allergen test patch 10 may include perforations or tear-away sections to facilitate partial removal or access to specific allergen sites. This feature may allow for staged removal of allergens or inspection of individual test sites without disturbing the entire allergen test patch 10.9QB\630666.01678\100992613.2Mayo 2023-378630666.01678

[0057] In some cases, the allergen test patch 10 may incorporate fiducial markers to enable image recognition and scaling. These fiducial markers may be printed or embedded on the surface of the allergen test patch 10. The fiducial markers may be arranged in a specific pattern or arrangement.

[0058] Additionally or alternatively, the allergen test patch 10 may incorporate a computer-readable barcode for allergen identification. This barcode may be printed or embedded on the patch surface and can contain encoded information about the specific allergens used in the test. In some cases, the barcode may be scanned or imaged (e.g., with the image capture device 805) to automatically record allergen data.

[0059] To enhance visibility under different lighting conditions, the allergen test patch 10 may include ink or markings visible under ultraviolet (UV) and / or infrared (IR) light. These specialized inks may be used for fiducial markers, barcodes, or other identifying features on the allergen test patch 10. The UV and / or IR visibility' allows for improved image capture and analysis in various lighting environments. In some embodiments, the ink or markings may be primarily or only visible to UV or IR light. These embodiments may’ be desirable in some instances because the ink and / or markings will be less visible / noticeable to unaided vision.

[0060] The allergen test patch 10 may include hook-and-loop fasteners, or other suitable fasteners, along the border of the allergen test patch 10. In some implementations, these fasteners allow for the attachment of protective barriers around the test site of the allergen test patch 10. The protective barriers may help prevent contamination or disturbance of the allergen application area. In some cases, the protective coverings may’ include vinyl or other clear plastic films bordered with hook-and-loop fasteners that can be affixed to the test site by the patient prior to showering or other water exposure. The protective coverings may be applied before exposure to water and removed after drying.

[0061] The allergen test patch 10 may utilize a hypoallergenic adhesive backing to secure it to the skin. The adhesive may be formulated to minimize skin irritation while providing sufficient adherence for the duration of the test period. In some cases, the adhesive may be designed to allow for easy removal of the allergen test patch 10 without causing skin damage.

[0062] The dimensions of the allergen test patch 10 may’ vary depending on the specific design and intended application. For example, in a multi-allergen test patch configuration the allergen test patch 10 may measure approximately 6 cm x 6 cm as one example, while in a single-allergen test patch configuration the allergen test patch 10 may be smaller, such as 1.510QB\630666.01678X100992613.2Mayo 2023-378630666.01678cm x 3 cm. These dimensions may be adjusted based on factors such as the number of allergens being tested and the anatomical location of patch application.

[0063] The allergen test patch 10 design may also incorporate features to enhance user comfort and wearability. This may include breathable materials, flexible components, or ergonomic shapes that conform to body contours. Such features may improve patient compliance and reduce the likelihood of premature patch removal or displacement during the testing period.

[0064] In some implementations, the allergen test patch 10 may be provided as part of a kit. The kit configurations may vary depending on the intended use and number of allergens to be tested. For example, a single allergen kit may include a single patch, a coverslip, a pull tab, and patient instructions for a simple home test. A moderate panel kit may include a multiwell patch configured to test four to five allergens, along with a coverslip, a pull tab, and patient instructions. A full panel kit may include a multi-well patch configured to test ten to twelve allergens, along with a coverslip, a pull tab, and patient instructions. The kit may also include supplies to clean the arm, assemble a patch application jig, install a mobile application, and perform the patch testing procedure. In some cases, the kit may include printed instructions and necessary supplies to facilitate self-administered patch testing by the patient in an at-home setting.

[0065] As described above, in some embodiments the allergen test patch 10 is configured as a single-allergen test patch having a single allergen well or chamber. Referring now to FIG. 1 A, an example allergen test patch 10 configured as a single-allergen test patch is illustrated. The allergen test patch 10 generally includes a patch substrate 12 and a border 14. The patch substrate 12 is removably coupled to the border 14. In use, the patch substrate 12 and border 14 are placed over a test site on the skin surface of a patient, subject, or other user of the allergen test patch 10. After a predetermined period of time (e g., 48 hours, 60 hours, 72 hours, 84 hours, 96 hours, or the like), the patch substrate 12 may be removed while the border 14 remains adhered to the skin surface of the patient. This allows for the allergens to be removed from the test site and the reaction of the patient to the allergen(s) to be evaluated while maintaining a physical border (via border 14) around the test site, which supports imaging and placement of protective coverings to protect the test site. Alternatively, some embodiments may be adapted to record an application time and actual removal time of the allergen test patch 10 on the test site. In these cases, the calculated statistics (described in more detail below) may11QB\630666.01678X100992613.2Mayo 2023-378630666.01678be automatically corrected based on a calculated difference between the application time and actual removal time.

[0066] The patch substrate 12 defines an allergen test region that includes an allergen well 16, one or more skin-inking regions 18, and a patch identifier 20. The patch substrate 12 may be separable from the border 14. For example, as described above, the patch substrate 12 may be removably coupled to the border 14 via perforations, or the like, which enable the patch substrate 12 to be separated and removed from the border 14 after a period of time during which the patient is exposed to the allergen in the allergen well 16. The patch substrate 12 may include a pull tab 52 to facilitate separating the patch substrate 12 from the border 14.

[0067] The patch substrate 12 and border 14 each have a top surface and a bottom surface. The bottom surface is a skin-facing surface that is configured to be placed into contact with the skin surface of the patient. The top surface is opposite the bottom surface and is intended to be visible to the patient and / or clinician when the allergen test patch 10 is adhered to the skin surface of the patient. In some implementations, as shown in FIGS. IB and 1C, the allergen well 16 may be fitted with filter paper to retain the allergen regardless of viscosity. In some implementations, the allergen well 16 may include a double-rimmed structure to retain the allergen and limit allergen bleed beyond the intended contact area. In some cases, a barcode, QR code, or the like, may be disposed on the obverse side of the allergen well 16 to facilitate identification of the allergen well 16, such as allergen lot information, kit identification, expiration date, or other relevant data.

[0068] The border 14 is removably coupled to the patch substrate 12. The skin-facing surface of the border 14 may be coated with an adhesive to adhere the allergen test patch 10 to the skin surface of the patient. As a non-limiting example, the adhesive may include an adhesive tape. The adhesive is preferably a hypoallergenic adhesive so as not to confound the allergen testing. A fastener region 22 surrounds the patch substrate 12 and is coupled to the top surface of the border 14. As a non-limiting example, the fastener region 22 may be composed of a hook-and-loop fastener. The fastener region 22 allows for removable test site coverage during activities that may compromise the test site once the patch substrate 12 is removed (e.g., showering). Matching waterproof or breathable materials can be removably coupled to the fastener region 22 to offer a range of protection to the test site after the patch substrate 12 removed.

[0069] The allergen well 16 is coupled to the skin-facing surface of the patch substrate 12. As a non-limiting example, the allergen well 16 may be composed of aluminum or another12QB\630666.01678X100992613.2Mayo 2023-378630666.01678generally inert metal (e.g., titanium) or material (e.g., medical-grade plastic). The allergen test well 16 contains an allergen during the test period. The size and shape of the allergen test well 16 may be optimally selected for each patch form factor.

[0070] The skin-inking regions 18 on the patch substrate 12 may include cutouts or other open regions formed in the patch substrate 12 where ink can be applied to the skin surface of the patient, such that when the separable patch substrate 12 is removed from the border 14 the inked regions will remain on the patient’s skin. The patient’s skin can be inked using one or more inks that include colors that can be seen under visible light, UV light, and / or IR light. These non-toxic ink markings may cooperate to create a computer readable barcode on the skin to identify allergens and / or patch form factor used.

[0071] The patch identifier 20 may include markings, indicia, or the like, that enable identification of the patch substrate 12 after it has been removed from the border 14, and which allow for the patch substrate 12 to be matched with the patient, the allergens tested, and other unique information. Patch identification may be implemented using a combination of computer readable formats (e.g., barcodes. QR codes) and human readable text (e.g., Ni for nickel). With a combination of unique patch test IDs, each patch substrate 12 can be linked to unique lot numbers and / or serial numbers for quality control.

[0072] One or more fiducial markers 24 are arranged around the patch substrate 12. In the illustrated example, the fiducial markers 24 are arranged in the fastener region 22 of the border 14. In the illustrated example, the fiducial markers 24 are arranged on the allergen test patch 10 such that when the patch substrate 12 is removed from the allergen test patch 10 the fiducial markers 24 will still be coupled to the patient, enabling patient identification, localization for imaging, and other information. By way of example, fiducial markers 24 on the retained border 14 may have their properties (e.g., number, patterns, and colors) varied to allow for locations and identification of allergens across full range of patch designs. Consistent with the inking on the skin, the fiducial markers 24 may contain colorings to be seen under all lighting conditions (e.g., visible light, UV light, IR light). In some implementations, the fiducial markers 24 may include emitters to be used to allow for precise positioning of a camera or other image capture device by triangulation of radio signals. In some other cases, the fiducial markers 24 may emit unique ID numbers for identification of patch and allergens. In some implementations, the allergen test patch 10 may include a binary marking system for patch identification. For example, five binary markings may be arranged along one side of the patch substrate 12, such that different combinations of inked and non-inked markings provide13QB\630666.01678X100992613.2Mayo 2023-378630666.01678multiple permutations for unique patch identification. This binary' marking system may allow for up to 32 unique patch identifiers using five binary positions.

[0073] One or more calibration charts 26 may also be coupled to the allergen test patch 10. In the illustrated example, a calibration chart 26 is coupled to the border 14 of the allergen test patch 10. In general, the calibration chart 26 includes color, size, and / or exposure calibration markings that allow for automated post-processing of images of the allergen test patch 10. These calibration markers create a computer readable barcode on the skin that can be used to identify allergens and patch form factor used. In some implementations, the calibration chart 26 may include a graduated grey scale yvith a middle grey and solid white for exposure mapping and white balance correction. Additionally or alternatively, the calibration chart 26 may include representative tiles for each of the six Fitzpatrick skin types to assist with skin fype assessment during image processing. The calibration chart 26 may be designed to be modular, allowing for different calibration configurations to be used depending on the specific testing requirements. Additional markings related to a color chart may be used to capture color fidelity and compensate for image processing that may be applied by the patient's camera prior to upload.

[0074] Referring now to FIGS. 2A and 2B, an example of another allergen test patch 10 configured as a single-allergen test patch is illustrated. In this configuration, the allergen test patch is generally sized and shaped like an adhesive bandage or patch. For example, the allergen test patch 10 may have dimensions of about 1.5 cm x 3 cm.

[0075] As illustrated, the patch substrate 12 accommodates a single allergen well 16. The allergen well 16 may be, for example, a 1 cm disc having a lip that contains the allergen to the test site on the skin surface of the patient. For instance, the allergen well 16 may be a hypoallergenic circular aluminum disc that is designed with a beveled edge to prevent elution of the allergen that is pre-applied to the skin exposed side of the allergen well 16.

[0076] The patch substrate 12 may be removably coupled to the skin surface of the patient by one or more adhesive patches 30 coupled to the bottom (i.e., skin-facing) surface of the patch substrate 12. The adhesive patches 30 may be covered with cover slips that can be removed to expose the adhesive before applying the allergen test patch 10 to the skin surface of the patient.

[0077] A border 32 surrounds an allergen test region 34 of the patch substrate 12, which contains the allergen well 16. The allergen test region 34 may be composed of a waterproof membrane that mitigates ink leakage and water intrusion into the test site. The border 32 may14QB\630666.01678X100992613.2Mayo 2023-378630666.01678be a double-lined silicon boundary that helps protect the allergen test region 34 from water intrusion. In some instances, the border 32 may include boundary inking, such that when the allergen test patch 10 is removed from the skin surface an inking that delineates the area of the allergen test region 34 will be retained on the skin surface.

[0078] A patch identifier 20 may be applied to the allergen test region 34 of the patch substrate 12 as a self-inking marker. The self-inking patch identifier 20 may include text that indicates the allergen contained in the allergen well 16. In these cases, the text is preferably mirrored, such that when the allergen test patch 10 is applied to the skin surface the self-inking patch identifier 20 will apply inking to the skin surface as non-mirrored text that is readable to the patient and / or clinician. In the illustrated example, the allergen test patch 10 is configured to test a nickel allergy. In this case, the self-inking patch identifier 20 contains a minor image of the text “Ni” to identify the allergen as nickel. As a non-limiting example, for the nickel allergen an ointment 2.5% concentration may be retained in the allergen well 16.

[0079] The allergen may be pre-applied to the allergen well 16. In the embodiment illustrated in FIG. 2A. the allergen test region 34 of the patch substrate 12 is covered by a dissolvable coating 36. The dissolvable coating 36 may be composed of a heat-sensitive material that dissolves at skin temperature. In this way, the allergen can be retained in the allergen well 1 until the allergen test patch 10 is adhered to the skin surface of the patient. At that time, the dissolvable coating 36 may dissolve, allowing exposure of the allergen to the skin surface.

[0080] In the embodiment illustrated in FIG. 2B, a removable cover slip 38 is instead used to cover the allergen test region 34. The patch substrate 12 is coupled to a backing 40 that may facilitate placement of the allergen test patch 10 on the skin surface. Before adhering the allergen test patch 10 to the skin surface, the cover slip 38 is removed from the allergen test patch 10. The allergen test patch 10 is then adhered to the skin surface and the backing 40 is removed, retaining the patch substrate 12 on the skin surface. In some instances, the allergen test patch 10 illustrated in FIG. 2B may not be pre-loaded with an allergen in the allergen well 16. After the cover slip 38 is removed, an allergen can be applied to the allergen well 16 before adhering the allergen test patch 10 to the skin surface. To facilitate proper use and compliance, the allergen well 16 may be stamped with '‘apply allergen here” to indicate that an allergen should be applied to the allergen well 16 prior to adhering the allergen test patch 10 to the skin surface of the patient.15QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0081] The patch design as a fiducial marker provides a number of advantages. As one advantage, the design enables the allergen test patch 10 to serve as a registration marker for an image capture device. Additionally or alternatively, the allergen test patch 10 can provide a size calibration marker for image analysis. As will be described in more detail below, this feature and advantage may be particularly desirable for use as an at-home and self-administered allergy test, as the user may have difficulty precisely aligning an at-home image capture device (e.g., a personal smartphone) with the test site.

[0082] Referring now to FIGS. 3A and 3B, another example of an allergen test patch 10 configured as a single-allergen test patch is illustrated. FIG. 3 A illustrates a reverse side (i.e., skin-facing surface) of the allergen test patch 10 and FIG. 3B illustrates an obverse side (i.e., top surface) of the allergen test patch 10.

[0083] As illustrated in FIG. 3 A, the allergen test patch 10 includes a patch substrate 12 having a top surface and a bottom surface. The bottom surface of the patch substrate 12 is a skin-facing surface that is configured to be placed into contact with the skin surface of the patient. An adhesive backing is disposed on the bottom surface of the patch substrate 12 to adhere the allergen test patch 10 to the skin surface of the patient. The adhesive backing may be a hypoallergenic adhesive so as not to confound the allergen testing. The patch substrate 12 includes a separable region 50 that contains an allergen well 16. The separable region 50 is configured to be removable from the patch substrate 12. such that after a predetermined period of time the separable region 50 containing the allergen well 16 can be separated and removed from the patch substrate 12 while the remainder of the patch substrate 12 remains adhered to the skin surface of the patient. The perforations may be arranged around the perimeter of the separable region 50 to facilitate separation of the separable region 50 from the patch substrate 12. A precut hole is formed in the patch substrate 12 within the separable region 50. The precut hole is sized and positioned to allow an allergen carrier to be delivered through the precut hole to the allergen well 16. In some implementations, the allergen carrier may be inserted through the precut hole to apply an allergen to the allergen well 16 prior to adhering the allergen test patch 10 to the skin surface of the patient.

[0084] As illustrated in FIG. 3B, the top surface of the patch substrate 12 includes a pull tab 52 coupled to the separable region 50. The pull tab 52 facilitates removal of the separable region 50 from the patch substrate 12 after the allergen test patch 10 has been adhered to the skin surface of the patient for a predetermined period of time. The pull tab 52 may extend beyond the outer periphery of the patch substrate 12 to allow the patient to grasp the pull tab16QB\630666.01678X100992613.2Mayo 2023-378630666.0167852 and separate the separable region 50 from the patch substrate 12. One or more fiducial markers 24 are arranged on the top surface of the patch substrate 12. The fiducial markers 24 may be positioned around the separable region 50 such that when the separable region 50 is removed from the patch substrate 12 the fiducial markers 24 will remain on the retained portion of the patch substrate 12. The fiducial markers 24 may be used to facilitate image capture and processing of the test site, as described in more detail below. One or more calibration charts, or color calibration markers, 26 are also coupled to the top surface of the patch substrate 12. The calibration charts 26 may include color calibration markers that allow for automated postprocessing of images of the allergen test patch 10 and the test site. In some implementations, the calibration charts 26 may include a graduated grey scale with a middle grey and solid white for exposure mapping and white balance correction. The calibration charts 26 may be positioned on the retained portion of the patch substrate 12 such that when the separable region 50 is removed the calibration charts 26 remain visible for image capture and analysis.

[0085] Referring now to FIG. 4, an example of an allergen test patch 10 configured as a multi-allergen test patch is illustrated. The allergen test patch 10 includes a patch substrate 12 having a plurality of allergen wells 16 coupled to the bottom, skin-facing surface of the patch substrate 12. The patch substrate 12 may be composed of a suitable biocompatible material that is preferably inert or otherwise hypoallergenic. As a non-limiting example, the patch substrate 12 may be composed of silicon. In some embodiments, the patch substrate 12 may be composed of a transparent or otherwise translucent material, such that the allergen test site may be observed and imaged while the allergen test patch 10 is adhered to the skin surface of the patient.

[0086] In the illustrated example, the patch substrate 12 may have dimensions of about 6 cm x 6 cm, with each allergen well 16 having dimensions of about 1.5 cm x 1.5 cm. The allergen wells 16 can advantageously be used to contain different allergens for testing, different concentrations of allergens for testing, or combinations thereof. In some cases, one of the allergen wells 16 can be used as a control (e.g., by containing a control substance). Although the illustrated example includes four allergen wells 16, in other embodiments the allergen test patch 10 may include more or fewer allergen wells 16. In some cases, the allergen wells 16 may be preloaded with allergens, similar to the allergen test patch 10 described above with respect to FIG. 2A. In these instances, each of the allergen wells 16 may be preloaded with a different allergen. Alternatively, one or more of the allergen wells 16 may be preloaded with different concentrations of the same allergen. The allergens preloaded in the allergen wells 1617QB\630666.01678X100992613.2Mayo 2023-378630666.01678may be selected from a test group of allergens. The number of allergens in the test group may be greater than the number of allergen wells 16, such that the test group of allergens is adapted to allow identification of a reaction to a specific allergen in the test group of allergens.

[0087] In some embodiments, the multi-allergen test patch 10 may include multiple individually removable allergen wells 16. In these configurations, each allergen well 16 may be contained within its own perforated, or otherwise separable, region formed in the patch substrate 12, similar to the separable region 50 described above with respect to FIGS. 3A and 3B. Perforations may be arranged around the perimeter of each separable region 50 to facilitate selective separation and removal of individual allergen wells 16 from the patch substrate 12 while other allergen wells 16 remain coupled to the patch substrate 12 and in contact with the skin surface of the patient. Each separable region 50 may include an individual pull tab (e.g., similar to pull tab 52 in FIG. 3B) coupled to the top surface of the patch substrate 12 to facilitate removal of the corresponding allergen well 16. The pull tabs 52 may extend beyond the outer periphery of the respective separable regions 50 to allow the patient to grasp each pull tab 52 independently. This design may provide several advantages. For example, the individually removable allergen wells 16 may allow for staged removal of allergens at different time points, which may be desirable when testing allergens that require different exposure durations. Additionally or alternatively, the individually removable allergen wells 16 may allow for sequential removal based on observed reactions, such that if a strong reaction is observed at a particular test site the corresponding allergen well 16 can be removed while other allergen wells 1 remain in place to continue testing. In some cases, the individually removable allergen wells 16 may also facilitate inspection of individual test sites without disturbing the entire allergen test patch 10. The separable regions 50 may be arranged in an array corresponding to the arrangement of the allergen wells 16, and precut holes similar to those described above with respect to FIG. 3 A may be formed within each separable region 50 to allow allergen carriers to be delivered to the respective allergen wells 16.

[0088] By way of non-limiting example, when testing for a nickel allergy, the allergen wells 16 may contain different concentrations of a nickel containing solution. For instance, three concentrations of nickel sulfate hexahydrate (e.g., 0.5%, 1.0%, 1.5%) may be used to assess the dose-response reaction to nickel while one allergen well 16 can be dedicated as a control test site containing a control solution, such as 0% nickel sulfate hexahydrate. In some configurations, one or more control fibers 28 may be integrated into the patch substrate 12. For instance, one of the control fibers 28 may be a single aluminum fiber to rule out reactions to18QB\630666.01678X100992613.2Mayo 2023-378630666.01678either the allergen wells 16 themselves or the process of physical testing. Additional control fibers 28 may include a common zinc alloy (e.g., to test costume jewelry) and 304 stainless steel fiber (typical “medical grade” stainless steel, like used in implants and surgical instruments), which can be introduced to further assess the severity of the reaction to nickel. By monitoring for potential skin reactions to the control-containing allergen well 16 and / or control fibers 28. irritant reactions may be further ruled out. If any of the controls are positive, all testing may be compromised.

[0089] The patch substrate 12 may include one or more fiducial markers 24. For example, a 0.5 cm x 1.5 cm fiducial marker 24 can be arranged on the top surface of the patch substrate 12 to be used as a scale for images of the test site in post-processing. The patch substrate 12 may also include a patch identifier 20, such as an icon, text, or marker that indicates the allergens used in the allergen test patch 10. The patch identifier 20 may be arranged on the top surface of the patch substrate 12. Additionally or alternatively, the fiducial marker 24 and / or patch identifier 20 may be arranged on the bottom, skin-facing surface of the patch substrate 12. In these instances, the fiducial marker 24 and / or patch identifier 20 may include ink markings that are transferred to the skin surface upon removal of the patch substrate 12 from the skin surface of the patient.

[0090] Referring now to FIGS. 5A-5F, another example of an allergen test patch 10 configured as a multi-allergen test patch is illustrated. In the illustrated example, the allergen test patch is configured as a radial test patch. The radial test patch includes a patch substrate 12 having an annular shape. In this case, the patch substrate 12 generally has an annular shape that circumscribes a central aperture 42. The patch substrate 12 may be composed of a biocompatible material that is preferably inert and / or hypoallergenic, such as silicon. A fiducial support 44 spans the central aperture 42. One or more fiducial markers 24 may be coupled to the fiducial support 44. In some configurations, the fiducial support 44 may be shaped to indicate an orientation of the allergen test patch 10. In the illustrated example, the fiducial support 44 includes a v-shape that indicates an orientation of the allergen test patch 10.

[0091] A plurality of allergen wells 16 are coupled to the patch substrate 12. The allergen wells 16 may be radially distributed around the annular patch substrate 12. The allergen wells 16 may be uniformly distributed around the annular patch substrate 12, or in some other embodiments the allergen wells 16 may be non-uniformly distributed around the annular patch substrate 12. The allergen wells 16 may in some instances be fitted with Finn chambers and cotton fibers to retain allergens.19QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0092] The patch substrate 12 has formed therein one or more skin-inking regions 18. The skin-inking regions 18 may include holes formed in the patch substrate 12 that allow for ink to be applied to the skin surface of the patient when the allergen test patch 10 is adhered to the skin surface, or may be recessed regions of the patch substrate 12 that can be fdled with ink prior to applying the allergen test patch 10 to the skin surface of the patient. The skin-inking regions 18 may be arranged along the outer periphery 46 of the patch substrate 12. The skininking regions 18 may be radially distributed around the outer periphery 46 of the annular patch substrate 12. The skin-inking regions 18 may be uniformly distributed around the annular patch substrate 12, or in some other embodiments the skin-inking regions 18 may be non-uniformly distributed around the annular patch substrate 12.

[0093] The skin-inking regions 18 can be selectively used to mark the skin surface of the patient with ink markings that provide a unique patch identifier. In this way, after the allergen test patch 10 is removed from the skin surface, the retained ink markings can identify which allergen test patch 10 was applied to the skin surface. As a non-limiting example, the allergen test patch 10 may include five skin-inking regions 18 that are distributed around the outer periphery 46 of the annular patch substrate 12, as illustrated in FIGS. 5D and 5E. Each skin-inking region 18 can be used to form a unique identifier for the allergen test patch 10. For instance, inking one of the skin-inking regions 18 can indicate a ‘ ’ in the unique identifier and leaving a skin-inking region 18 free of ink can indicate a “0’?in the unique identifier. The skin-inking regions 18 can be ordered, such that a consistent identifier can be indicated by selectively inking the skin-inking regions 18. For example, the skin-inking regions 18 can be indexed in a counter-clockwise manner starting with a skin-inking region 18 to which the fiducial support 44 is pointing. A counter-clockwise indexing can be used such that when the allergen test patch 10 is applied to the skin surface and subsequently removed, the retained ink markings will be ordered with a clockwise sequence. Table 1 below shows an example of unique identifiers that can be generated using the skin-inking regions 18.Table 1: Patch IDs Created Using Skin-Inking RegionsPatch Ink Sequence Patch InkID (0=No ink; l=lnk Present) ID Sequence1 0-0-0-0-0 17 0-0-0-0-12 1-0-0-0-0 18 1-0-0-0-13 0-1-0-0-0 19 0-1-0-0-1 4 1-1-0-0-0 20 1-1-0-0-1 5 0-0-1-0-0 21 0-0-1-0-120QB\630666.01678\100992613.2Mayo 2023-378630666.016786 1-0-1-0-0 22 l-O-l-O-l 7 0-1-1-0-0 23 O-l-l-O-l 8 1-1-1-0-0 24 l-l-l-O-l 9 0-0-0-1-0 25 O-O-O-l-l 10 1-0-0-1-0 26 l-O-O-l-l 11 0-1-0-1-0 1 O-l-O-l-l 12 1-1-0-1-0 28 l-l-O-l-l 13 0-0-1-1-0 29 O-O-l-l-l 14 1-0-1-1-0 30 l-O-l-l-l 15 0-1-1-1-0 31 O-l-l-l-l 16 1-1-1-1-0 32 l-l-l-l-l

[0094] In some embodiments, an adhesive backing 48 can be coupled to the allergen test patch 10 to facilitate adhering the allergen test patch 10 to the skin surface of the patient, as illustrated in FIG. 5F. In use, the annular patch substrate 12 is coupled to the adhesive backing 48 and then the adhesive backing 48 is applied to the skin surface of the patient, thereby coupling the allergen test patch 10 to the skin surface of the patient. The bottom surface of the adhesive backing 48 has an adhesive disposed thereon. As illustrated in FIG. 5G, the top surface of the adhesive backing 48 may include one or more fiducial markers 24, calibration charts 26, or other markings that can be retained on the skin surface after removing the annular patch substrate 12 from the skin surface. The adhesive backing 48 can include a separable region 50 that can be removed to reveal the annular patch substrate 12, such that the annular patch substrate 12 can be removed from the skin surface while retaining the rest of the adhesive backing 48. A pull tab 52 may be coupled to the separable region 50 to facilitate removal of the separable region 50. The separable region 50 may be removably coupled to the adhesive backing via perforations, or the like.

[0095] The radial patch design illustrated in FIGS. 5A-5F accommodates use with various computer vision and other image processing techniques. In some applications, the shape and size of the fiducial support 44 can be used to scale images of the test site accordingly. As described below in more detail, images of the test site can also be processed to perform object detection (e.g., reaction detection) across the allergen test sites defined by the allergen wells 16. The skin-inking regions 18 can also be used to identify a unique patch identifier. For example, FIG. 5D illustrates an allergen test patch 10 with skin-inking regions 18 filled with ink to form a patch identifier l-l-l-l-l (e.g., Patch 32) and FIG. 5E illustrates an allergen test patch 10 with skin-inking regions 18 filled with ink to form a patch identifier 0-1-0-0-1 (e.g.,21QB\630666.01678X100992613.2Mayo 2023-378630666.01678Patch 19). These patch identifiers can be used to reference images and patient records to retrieve the unique allergens assigned to the patient.

[0096] As shown in FIG. 5H, an allergen application cradle 54 can be used to retain the annular patch substrate 12 to facilitate loading the allergen wells 16 with respective allergens for testing. Instructions can direct the patient to arrange and place each allergen tube around the radial patch according to numbering, coloring, or the like. The test site can be imaged with the image capture device after preparation to document and verify the allergen locations.

[0097] Referring now to FIG. 6, yet another example of an allergen test patch 10 configured as a multi-allergen test patch is illustrated. Like the allergen test patch 10 illustrated in FIG. 1A, the allergen test patch 10 illustrated in FIG. 6 includes a patch substrate 12 that is removably coupled to a border 14. The patch substrate 12 defines an allergen test region that includes a plurality7of allergen wells 16, one or more skin-inking regions 18, and one or more fiducial markers 24.

[0098] The patch substrate 12 may be separable from the border 14. For example, as described above, the patch substrate 12 may be removably coupled to the border 14 via perforations, or the like, which enable the patch substrate 12 to be separated and removed from the border 14 after a period of time during which the patient is exposed to the allergen(s) in the allergen wells 16. The patch substrate 12 may include a pull tab 52 to facilitate separating the patch substrate 12 from the border 14.

[0099] The patch substrate 12 and border 14 each have a top surface and a bottom surface. The bottom surface is a skin-facing surface that is configured to be placed into contact with the skin surface of the patient. The top surface is opposite the bottom surface and is intended to be visible to the patient and / or clinician when the allergen test patch 10 is adhered to the skin surface of the patient.

[0100] The border 14 is removably coupled to the patch substrate 12. The skin-facing surface of the border 14 may be coated with an adhesive to adhere the allergen test patch 10 to the skin surface of the patient. As a non-limiting example, the adhesive may include an adhesive tape. The adhesive is preferably a hypoallergenic adhesive so as not to confound the allergen testing. A fastener region 22 surrounds the patch substrate 12 and is coupled to the top surface of the border 14. As a non-limiting example, the fastener region 22 may be composed of a hook-and-loop fastener. The fastener region 22 allows for removable test site coverage during activities that may compromise the test site once the patch substrate 12 is removed (e.g., showering). Matching waterproof or breathable materials can be removably coupled to the22QB\630666.01678X100992613.2Mayo 2023-378630666.01678fastener region 22 to offer a range of protection to the test site after the patch substrate 12 removed.

[0101] The allergen well 16 is coupled to the skin-facing surface of the patch substrate 12. As a non-limiting example, the allergen well 16 may be composed of aluminum or another generally inert metal (e.g., titanium) or material (e.g., medical-grade plastic). The allergen test well 16 contains an allergen during the test period. The size and shape of the allergen test well 16 may be optimally selected for each patch form factor.

[0102] The skin-inking regions 18 may include holes in the patch substrate 12 such that the skin surface can be marked with ink in a pattern defined by the holes, such as a binary code that indicates a unique identifier for the allergen test patch 10. Additionally or alternatively, the skin-inking regions 18 may be ink wells or the like that can be filled with ink before applying the allergen test patch 10 to the skin surface of the patient. Like the radial test patch described above, inked regions can indicate a “1” in the unique identifier and non-inked regions can indicate a "0" in the unique identifier. The skin-inking regions 18 can be ordered, such that a consistent identifier can be indicated by selectively marking the skin surface via the skininking regions 18. For example, the skin-inking regions 18 can be indexed in a top-down manner. Table 2 below- shows an example of unique identifiers that can be generated using the skin-inking regions 18.Table 2: Patch IDs Created Using Removable Patch IdentifiersPatch Hole Patch HoleID Sequence ID Sequence1 0-0-0-0-0 17 0-0-0-0-12 1-0-0-0-0 18 1-0-0-0-13 0-1-0-0-0 19 0-1-0-0-14 1-1-0-0-0 20 1-1-0-0-15 0-0-1-0-0 21 0-0-1-0-16 1-0-1-0-0 22 1-0-1-0-17 0-1-1-0-0 23 0-1-1-0-18 1-1-1-0-0 24 1-1-1-0-19 0-0-0-1-0 25 0-0-0-1-110 1-0-0-1-0 26 1-0-0-1-111 0-1-0-1-0 T1 0-1-0-1-112 1-1-0-1-0 28 1-1-0-1-113 0-0-1-1-0 29 0-0-1-1-114 1-0-1-1-0 30 1-0-1-1-115 0-1-1-1-0 31 0-1-1-1-123QB\630666.01678\100992613.2Mayo 2023-378630666.0167816 1-1-1-1-0 32 l-l-l-l-l

[0103] One or more fiducial markers 24 are arranged around the patch substrate 12. In the illustrated example, the fiducial markers 24 are arranged in the comers of the patch substrate 12. The fiducial markers 24 may have a rotationally variant shape, such that the fiducial markers 24 can indicate the orientation of the allergen test patch 10. In some implementations, the fiducial markers 24 may have their properties (e g., number, patterns, and colors) varied to allow for locations and identification of allergens across full range of patch designs. The fiducial markers 24 may also contain colorings to be seen under all lighting conditions (e.g., visible light, UV light, IR light). In some implementations, the fiducial markers 24 may include emitters to be used to allow for precise positioning of a camera or other image capture device by triangulation of radio signals. In some other cases, the fiducial markers 24 may emit unique ID numbers for identification of patch and allergens.

[0104] The allergen test patch 10 may have a defined direction, height, and width, which can be demarked by the fiducial markers 24. Additionally or alternatively, the fiducial markers 24 may facilitate calculating relative scaling of the allergen test patch 10 in images of the allergen test patch 10 acquired with an image capture device. This can account for camera location and angle at the time of image capture. As indicated in FIG. 6, the vertical pixel range (i.e., yl to y2) may have its limits determined based on the relative image size and resolution after perspective correction. Likewise, the horizontal pixel range (i.e., xl to x2) may have its limits determined based on the relative image size and resolution after perspective correction. The square image tile 56 bounded by coordinates (xl, yl, x2, y2) may be extracted from the entire patch image to facilitate documentation of allergic reaction.

[0105] FIGS. 7A-7E illustrate another example of an allergen test patch 710 configured as a multi-allergen test patch. In this example, the allergen test patch 710 includes a patch substrate 712, a border 714, a protective cover 770, and a fiducial marker template 780, as illustrated in FIG. 7A. In the illustrated example, the patch substrate 712 is constructed similarly to the patch substrate 12 shown in FIG. 6.

[0106] The patch substrate 712 defines an allergen test region 734 that includes a plurality of allergen wells 716, one or more skin-inking regions 718, and one or more fiducial markers 724. The patch substrate 712 includes a fastener region 758 to removably couple the patch substrate to the border 714. The fastener region 758 may circumscribe the allergen test region 734 of the patch substrate 712. As a non-limiting example, the fastener region 758 may24QB\630666.01678\100992613.2Mayo 2023-378630666.01678include a hook-and-loop fastener or other suitable fastener that allows for removably coupling the patch substrate 712 to the border 714.

[0107] The border 714 includes a substrate having overall dimensions (e.g., height, width) that are slightly larger than the patch substrate 712. The border 714 includes a central opening 762 that is sized to match the allergen test region 734 of the patch substrate 712, such that when the border 714 is adhered to the skin surface of the patient and the patch substrate 712 is coupled to the border 714 the allergen well 716 in the allergen test region 734 of the patch substrate 712 will come into contact with the skin surface of the patient.

[0108] The border 714 includes a fastener region 768 to removably couple the patch substrate 712, protective cover 770, and fiducial marker template 780 to the border 714. The fastener region 768 may circumscribe the central opening 762 of the border 714. As a nonlimiting example, the fastener region 768 may include a hook-and-loop fastener or other suitable fastener that allows for removably coupling components of the allergen test patch 710 to the border 714. By way of example, when the fastener region 758 of the patch substrate 712 includes a hook-and-loop fastener, the fastener region 768 of the border 714 may include a matching hook-and-loop fastener. For instance, the fastener region 758 of the patch substrate 712 may include the hook components of the hook-and-loop fastener and the fastener region 768 of the border 714 may include the corresponding loop components of the hook-and-loop fastener. When the border 714 is adhered to the skin surface of the patient, the various components of the allergen test patch 710 (e.g., patch substrate 712, protective cover 770, fiducial marker template 780) may be selectively removed and / or applied to the border 714, as described below in more detail.

[0109] The protective cover 770 may be composed of a transparent or translucent material such that the test site on the skin surface of the patient may be observed when the protective cover 770 is coupled to the border 714. Advantageously, the protective cover 770 may be composed of a waterproof material, such that when the protective cover 770 is coupled to the border 714 the patient may wash (e.g., take a shower) without the integrity of the test site being compromised. In some cases, the protective cover 770 includes a fastener region 778, which may include a border of the protective cover 770 that is covered in one or more fasteners or fastener materials. By way of example, the fastener region 778 of the protective cover 770 may include a hook-and-loop fastener (e.g., the matching components of the hook-and-loop fastener of the fastener region 768 of the border 714).25QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0110] The fiducial marker template 780 may be removably coupled to the border 714, such that when the border 714 is adhered to the skin surface of the patient the fiducial marker template 780 can be coupled to the border 714 to provide fiducial markers 724, calibration charts 726, and other information that can be imaged together with the test site on the skin surface of the patient. Like the border 714, the fiducial marker template 780 has a central opening 782 that is sized and shaped to match the allergen test region 734 of the patch substrate 712, such that when the fiducial marker template 780 is coupled to the border 714 the test site on the skin surface of the patient will remain visible.

[0111] The fiducial marker template 780 has a top surface and a bottom surface. One or more fiducial markers 724 are coupled to the top surface of the fiducial marker template 780. Additionally or alternatively, one or more calibration charts 726 may be coupled to the top surface of the fiducial marker template 780. A fastener region (not shown) is coupled to the bottom surface of the fiducial marker template 780. For instance, the fastener region may circumscribe the central opening 782 of the fiducial marker template 780 on the bottom surface of the fiducial marker template 780, such that the fastener region of the fiducial maker template 780 may match with the fastener region 768 on the border 714. By way of example, the fastener region of the fiducial marker template 780 may include a hook-and-loop fastener (e.g., the matching components of the hook-and-loop fastener of the fastener region 768 of the border 714).

[0112] In use, the allergen test patch 710 illustrated in FIGS. 7A-7E may be used to perform an allergen test (e g., an allergic contact dermatitis test), which may be carried out by a patient in an at-home setting, or by a clinician in a clinical setting. The allergen test may be carried out over a period of time such as several hours, days, or the like. The border 714 is adhered to a skin surface of the patient. As an example, the border 714 may be applied to the volar aspect of the patient’s forearm. One or more allergens are supplied to the allergen wells 716 of the patch substrate 712. The self-inking regions 718 and / or fiducial markers 724 on the patch substrate 712 may be selectively filled with ink, as described above. As illustrated in FIG. 7B, the patch substrate 712 is then flipped over and removably coupled to the border 714 by mating the fastener region 758 of the patch substrate 712 with the fastener region 768 of the border 714.

[0113] The patch substrate 712 is then kept in place on the skin surface of the patient for a period of time, as illustrated in FIG. 7C. As a non-limiting example, the patch substrate 712 may be kept in place on the skin surface of the patient for a duration of three days.26QB\630666.01678X100992613.2Mayo 2023-378630666.01678Alternatively, the patch substrate 712 may be kept in place on the skin surface of the patient for other durations of time, such as one day, two days, four days, five days, and the like. In some other cases, the patch substrate 712 may be kept in place on the skin surface of the patient for a duration of time measured in hours, such as a duration of hours in the range of 6-60 hours, such as 6 hours, 12 hours, 18 hours, 24 hours, 30 hours, 36 hours, 42 hours, 48 hours, 54 hours, 60 hours, and so on.

[0114] When the duration of time has passed, the patch substrate 712 is removed from the border 714, exposing the test site on the skin surface of the patient. The patient or another individual then takes a photograph of the test site using an image capture device, which in some instances may be the camera on a smartphone. Additional images of the test site may be captured on subsequent days (e.g., day four, day five, etc.). The smartphone may execute an app that allows for image capture and automated processing of the images. In the illustrated example shown in FIGS. 7D and 7E, allergens supplied in the allergen wells 716 have caused allergic reactions with the skin surface of the patient, leaving reaction sites 790 on the skin surface after removal of the patch substrate 712. As also illustrated in FIGS. 7D and 7E, ink applied to the skin-inking regions 718 and fiducial markers 724 have left ink markings 792 and 794, respectively, on the skin surface of the patient. As described above, the ink markings 792 from the self-inking regions 718 may define a patch identification code and the ink markings 794 from the fiducial markers 724 may provide information about the orientation, scaling, etc., of the applied patch substrate 712, which can be used for automated processing of images of the test site. The image of the test site may be uploaded automatically for offline analysis. A report with interpretation may then be returned to the patient’s smartphone.

[0115] As illustrated in FIG. 7D, after the patch substrate 712 has been removed from the border 714. the protective cover 770 may be removably coupled to the border 714 to protect the test site on the skin surface of the patient. In the illustrated example, the protective cover 770 is transparent or otherwise translucent such that images of the test site can be captured without having to remove the protective cover 770. Alternatively, as illustrated in FIG. 7E, the protective cover 770 can be removed from the border 714 and replaced with the fiducial marker template 780 when taking images of the test site. In this way, the fiducial markers 724, calibration charts 726, or other information conveyed via the fiducial marker template 780 can be depicted in images of the test site to facilitate automated processing of the images.

[0116] Referring now to FIG. 8, an automated patch testing system 800 in accordance with some embodiments is illustrated. The automated patch testing system 800 includes an27QB\630666.01678X100992613.2Mayo 2023-378630666.01678allergen test patch 10(e.g., the allergen test patch 10 of any one of FIGS. 1-7), an image capture device 805, a computing device 810 and / or a server 815. The image capture device 805 may be in communication with the computing device 810 and / or server 815 either directly or indirectly via a network 820.

[0117] Referring again to FIG. 8, the image capture device 805 may include any suitable device for capturing, acquiring, or otherwise recording images. In some cases, the image capture device 805 may include one or more cameras. In some implementations, the image capture device 805 may include a smartphone, tablet computer, or other mobile device having a camera. For example, the image capture device 805 may include one or more cameras on a smartphone. FIGS. 9-12 show example patch test images that can be acquired using the image capture device 805, including images in both the visible light spectrum and UV spectrum. The patch test images depict allergen reaction sites in addition to ink marking retained on the skin surface by use of the allergen test patches 10 described in the present disclosure.

[0118] The image capture device 805 may include both hardware and software components designed to standardize and optimize the capture of patch test images. In some cases, the image capture device 805 may incorporate a smartphone camera hood to provide consistent lighting conditions and prevent unwanted shadows. This hood may be attachable to various smartphone models and may include built-in light sources to ensure uniform illumination of the patch test area.

[0119] The image capture device 805 may also include an automated image capture process implemented through a smartphone application. This application may guide users through the image capture process, ensuring proper framing and focus of the patch test area. In some cases, the application may utilize the smartphone’s camera and sensors to automatically detect when the patch is properly positioned and in focus, triggering image capture without user intervention.

[0120] An advantageous component of the image capture device 805 is that patch recognition software may be integrated into the smartphone application. This software may use computer vision algorithms to detect and identify the fiducial markers (e.g., fiducial markers 24) on the allergen test patch 10, allowing for automatic scaling and orientation of the captured image. In some cases, the software may be capable of recognizing different patch designs and configurations, adapting its analysis accordingly. In some other embodiments, the smartphone application may be primarily used to capture and transmit images of the test site, and the patch28QB\630666.016781100992613.2Mayo 2023-378630666.01678recognition software may be executing on an internet accessible computer (e.g., server 815 or another cloud server).

[0121] The image capture device 805 may also incorporate a full spectrum lighting camera apparatus. This apparatus may allow for the capture of images under various lighting conditions, including visible light, UV light, near-infrared light, IR light, and so on. In some cases, this multi-spectral imaging capability of the image capture device 805 may enhance the detection and analysis of patch test reactions, particularly in cases where reactions may be more visible under certain lighting conditions.

[0122] The image capture device 805 may be designed to standardize the image capture process across different users and environments. This standardization may be achieved through a combination of hardware design, such as a fixed geometry' of the camera hood, and software controls, such as automatic exposure and white balance adjustments. In some cases, the image capture device 805, computing device 810, and / or server 815 may include calibration features to account for variations in ambient lighting conditions. In some instances, the captured images can be processed to detect calibration information (e.g., calibration charts 26 on the allergen test patch 10) that can improve the ability for automatic exposure and white balance adjustments or other adjustments to compensate for variations in ambient lighting conditions.

[0123] To further enhance image quality and consistency, the image capture device 805, computing device 810. and / or server 815 may incorporate advanced image processing techniques. These techniques may include noise reduction, contrast enhancement, and / or color correction algorithms applied in real-time or during post-processing. In some cases, the image capture device 805 may capture multiple images in rapid succession and combine them to produce a single high-quality' image with improved dynamic range and reduced noise.

[0124] The image capture device 805 may also include features to ensure proper documentation and organization of captured images. This may include automatic tagging of images yvith relevant metadata such as date, time, patch type, and patient identifier. In some cases, the system may integrate with secure cloud storage solutions (e.g., via server 815) for immediate backup and sharing of images with healthcare providers.

[0125] The computing device 810 and / or server 815 receive images from the image capture device 805 where they are processed to detect regions on the skin surface that have reacted to one or more allergens, to quantity' reaction severity', etc., in accordance with some embodiments described in the present disclosure.29QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0126] The network 820 may be a long-range wireless network such as the Internet, a local area network (LAN), a wide area network (WAN), or a combination thereof. In other embodiments, the network 820 may be a short-range wireless communication network, and in yet other embodiments, the network 820 may be a wired network using, for example, universal serial bus (USB) cables. In some embodiments, the network 820 may include both wired and wireless devices and connections. Similarly, the server 815 may transmit information to the computing device 810.

[0127] In some embodiments, the image capture device 805 communicates directly with the computing device 810. For example, the image capture device 805 can transmit data (e.g., images) to the computing device 810. Similarly, the image capture device 805 can receive data (e.g., settings, firmware updates, etc.) from the computing device 810.

[0128] In some other embodiments, the image capture device 805 bypasses the computing device 810 to access the network 820 and communicate with the server 815 via the network 820. In some embodiments, the image capture device 805 is equipped with a long-range transceiver instead of or in addition to a short-range transceiver. In such embodiments, the image capture device 805 communicates directly with the server 815 or with the server 815 via the network 820 (in either case, bypassing the computing device 810).

[0129] In some embodiments, the image capture device 805 may communicate directly with both the server 815 and the computing device 810. In such embodiments, the computing device 810 may, for example, generate a graphical user interface to facilitate control and programming of the image capture device 805 while the server 815 may store and analyze larger amounts of data (e.g., training data, trained machine learning models and parameters, images) for future programming or operation of the image capture device 805. In other embodiments, the image capture device 805 may communicate directly with the server 815 without utilizing a short-range communication protocol with the computing device 810.

[0130] In the illustrated embodiment, the image capture device 805 communicates with the computing device 810. The computing device 810 may include, for example, a smartphone, a tablet computer, a laptop computer, a desktop computer, a smart watch, another wearable device, and the like. The image capture device 805 communicates with the computing device 810, for example, to transmit at least a portion of the images or other data collected or generated by the image capture device 805.

[0131] In some embodiments, the computing device 810 may include a short-range transceiver to communicate with the image capture device 805, and a long-range transceiver to30QB\630666.01678X100992613.2Mayo 2023-378630666.01678communicate with the server 815. In the illustrated embodiment, the image capture device 805 can also include a transceiver to communicate with the computing device 810 via. for example, a short-range communication protocol such as Bluetooth®. In some embodiments, the computing device 810 bridges the communication between image capture device 805 and the server 815. That is, the image capture device 805 transmits data to the computing device 810, and the computing device 810 forwards the data from image capture device 805 to the server 815 over the network 820.

[0132] The server 815 includes a server electronic control assembly having a server electronic processor, a server memory, and a transceiver. The transceiver allows the server 815 to communicate with the image capture device 805, the computing device 810, or both. The server electronic processor receives images or other data collected with or generated by the image capture device 805, and stores the received data in the server memory. The server 815 may maintain a database (e.g., on the server memoiy ) for containing images, training data, trained machine learning controls (e.g., trained machine learning models and / or algorithms), artificial intelligence controls (e.g., rules and / or other control logic implemented in an artificial intelligence model and / or algorithm), and the like.

[0133] Although illustrated as a single device, the server 815 may be a distributed device in which the server electronic processor and server memory are distributed among two or more units that are communicatively coupled (e.g., via the network 820).

[0134] As an example, a camera of the image capture device 805 can be used to record an image of the allergen test patch 10 and the computing device 810 and / or server 815 can receive and process the image to detect regions on the skin surface that have reacted to the tested allergens. For example, the image can be processed using computer vision techniques, object detection techniques, other machine learning-based image processing techniques, or other suitable image processing techniques, to detect regions on the skin surface that have reacted to the tested allergens.

[0135] The present disclosure encompasses various algorithms and techniques for detecting allergic contact dermatitis reactions in patch test images. These techniques may include image segmentation, using CNNs or other machine learning models for reaction detection, and region-of-interest (ROI)-based reaction detection.

[0136] In some cases, the machine learning model may implement image segmentation to isolate the patch test area from the surrounding skin. As a non-limiting example, the image segmentation model may be based on a CNN implementing a U-Net architecture. The U-Net31QB\630666.01678X100992613.2Mayo 2023-378630666.01678architecture may include an encoder path that downsamples the input image and a decoder path that upsamples the features, with skip connections between corresponding encoder and decoder layers to preserve spatial information. A CNN-based model may be used to detect reactions within the segmented patch test areas. This model may be trained on a dataset of patch test images with expert-annotated reactions.

[0137] Additionally or alternatively, an ROI-based reaction detection algorithm may be utilized to provide more flexibility in detecting reactions across various patch configurations. This technique may be based on a You Only Look Once (YOLO) model architecture or other object detection model architecture. The YOLO model may be trained to detect and localize positive reactions within patch test images by drawing bounding boxes around reaction sites. The ROI-based reaction detection algorithm may involve preprocessing the input image to normalize brightness, contrast, distance, angle(s) / orientation of the image capture device relative to the test site, and so on. The preprocessed image may then be input to the YOLO model to generate bounding box predictions for potential reaction sites. Nonmaximum suppression may be applied to eliminate overlapping bounding boxes and retain the most confident predictions. The retained bounding boxes may then be post-processed to refine their coordinates and assign confidence scores. In some implementations, a second-generation object detection model based on YOLO or similar architectures may be used to provide additional flexibility compared to first-generation convolutional neural network models. The second-generation model may be conducive to video capture, alternative patch designs, and less restrictive image cropping and camera orientation requirements. This flexibility7may allow the system to accommodate various patch configurations and imaging conditions without requiring a fixed patch layout or fully cropped images of the test site.

[0138] In some implementations, the Y OLO model may be trained on a dataset of patch test images with expert-annotated bounding boxes around positive reaction sites. The training process may involve data augmentation techniques such as random rotations, flips, and brightness adjustments to improve the generalization of the model.

[0139] The ROI-based reaction detection algorithm may offer several advantages, such as flexibility in handling various patch configurations and sizes, the ability' to detect multiple reactions within a single patch test image, and the provision of localized information about reaction sites, which may be useful for further analysis or visualization.32QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0140] The algorithms described herein may be implemented using various deep learning frameworks and may be deployed on different hardware configurations, including but not limited to graphics processing units (GPUs) for accelerated inference.

[0141] In some implementations, the images may also be processed to quantify the severity' of the reaction to one or more allergens. In these instances, the computing device 810 and / or server 815 can output processed images in addition to classified feature data that indicate a severity' of reaction to one or more allergens.

[0142] Additionally, the computing device 810 and / or server 815 can receive and process the image to detect patch identifiers (e.g., patch identifiers 20), fiducial markers (e.g., fiducial markers 24), and / or inked regions of the skin surface (e.g., regions inked via one or more skin-inking regions 18). These can be detected in the image and used to post-process the image (e.g., to scale the image, etc.), to identify the allergen test patch 10 used during testing, and to otherw ise link the image to the patient’s health records.

[0143] Images of the allergen test patch 10 can also be recorded and processed by the computing device 810 and / or server 815 to generate training data for a machine learning model. As one example, images of an allergen test patch 10 can be captured and stored to associate images of the allergen test patch 10 with a reaction to a tested allergen. In some instances, the images may be labeled with a class, categorization, or quantification of reaction severity.

[0144] Images can also be shared with clinicians or other healthcare providers. For example, images can be shared with the patient’s doctor, or the like. In some instances, the computing device 810 and / or server 815 can store images, processed images, and / or classified feature data to the patient’s electronic health record (EHR).

[0145] In some examples, the processing of images acquired with the image capture device 805 can be performed by a software application (e.g., an app) executed on the computing device 810 and / or server 815. By way of example, the computing device 810 may be a smartphone and the software application may be an app executed by the smartphone to provide automatic scaling and detection of allergen test patches based on the fiducial markers and binary math codes (i.e., patch identifiers). Captured images may be automatically uploaded to a database (e.g., a database residing on server 815). The app may include the automated acquisition as described in the present disclosure, or alternatively may' implement a full manual mode.

[0146] One of the advantages of the systems and methods described in the present disclosure is that they provide effective screening as they may enable testing at home by the33QB\630666.01678X100992613.2Mayo 2023-378630666.01678patient using existing hardware (e.g.. a personal smartphone). This can be measured by positive and negative predictive values in the intended use cohort. In some implementations, the detection methods may be configured to purposefully provide high sensitivity. A mathematical side effect of such a configuration is relatively lower specificity. The combination of high sensitivity with low specificity results in lower positive predictive value.

[0147] In some implementations, the mobile application may include two operating modes: a patient mode and a provider mode. In patient mode, the application may allow patients to access testing instructions, capture images of skin reactions at the test site, track their patient journey through the testing process, and receive testing reminders at appropriate intervals. The patient mode may include user profile integration for secure login and patch registration, guided instructions for patch testing procedures, and high-quality image acquisition features. In provider mode, the application may provide healthcare providers with access to patient data and images, testing and photographic documentation, Al-generated reaction detection results retrieved from the cloud, and clinical report review functionality. The provider mode may automatically generate test reports summarizing the detected reactions, their severity, and potential allergen cross-reactivity. In some cases, the application may advance the Patient Journey assessment to be database-driven, allowing greater flexibility to expand the depth of information gathered while minimizing the application development costs of a static mobile application.

[0148] To improve positive predictive value, increasing the prevalence (i.e., using allergens that generate a positive reaction in a higher percentage of patients, e.g., nickel) may improve the prevalence. Additionally, improving the specificity' (e.g., by reducing the number of false positive predictions) will also improve the positive predictive value. To provide additional improvements in specificity, two approaches may be adopted. As one example, images that are indicated as possibly containing reactions can be flagged for user review. In these instances, a subset of processed images may be flagged for subsequent review' by a clinician. These flagged images may be automatically routed to a clinician or otherwise added to a clinical review workflow. As another example, the computing device 810 can prompt the patient with a questionnaire to provide additional information that can be used to confirm a positive reaction. For example, if a suspected reaction is observed at the test site that tested nickel, the questionnaire could include questions such as: “Have you ever had a reaction to jewelry, particularly costume jewelry?” More generally, these questions would tie into34QB\630666.01678X100992613.2Mayo 2023-378630666.01678knowledge of common source of exposure of the allergen and would provide a guided question and answer format to further rule out false positive.

[0149] By way of example, an illustrative patch testing workflow using the allergen test patches 10 described in the present disclosure is now described. In this illustrative example, a multi-allergen test patch, such as the allergen test patch 10 illustrated in FIG. 6 and / or the allergen test patch 10 illustrated in FIGS. 7A-7E, may be used.

[0150] A standard clinical testing protocol will apply 15 or more patch tests to the entire surface of the back. The advantage here is a comprehensive test is performed at once, maximizing the five day investment that the patient makes. The approach, however, requires sophisticated compounding and storage approaches for the allergens, requires assistance of a bystander or medical professional to apply, document the patch locations, and remove the patches from the back. Such a model is only conducive to in-clinic testing by medical professionals.

[0151] A multi-allergen array (e g., 10 allergens) tested on the volar aspect of the arm allows for the person being tested to apply, remove, and document the testing process at home and without special training. Furthermore, having the ability to see the patch test as it is being conducted allows for maintenance of the patch during the three-day direct exposure period. Having this ability also allows for body care such as showering and other activities of daily living without the assistance of another person. Disadvantages are that the patch is in a nondiscrete location and located on skin surface that experiences regular motion and exposure to sun.

[0152] In an example testing scenario, the patient applies a prepared allergen test patch 10 to the volar aspect of the arm and uses an ink marker pen to color over the fiducial cutout (e.g.. skin-inking regions 18) as a stencil to leave an ink mark on the skin. The patch remains in place for three days and then is removed by the patient. The patient or another individual then takes a photograph of the skin patch site using an image capture device, which in some instances may be the camera on the patient’s smartphone. The smartphone may execute an app that allows for image capture and automated processing of the images. The ink markings on the skin and / or fiducial markers (e.g., fiducial markers 24) may be used as a reference in the images. The image is uploaded automatically for offline analysis (e.g., via a computing device 810 and / or server 815). A report with interpretation may then be returned to the patient’s smartphone.35QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0153] In another example, a single-allergen test patch may be used. The single allergen approach is a modification to the self-managed multi-allergen approach. One modification is that a single allergen test patch utilizes a much smaller, more discrete patch design, as described above. An advantage of this approach is that the allergens can be individually applied. Retesting of single allergens can also be economically conducted. The total start-to-finish time for testing a panel of allergens may take significantly longer with a single-allergen patch design than with a multi-allergen array, but in some instances more than one patch may be applied at a time.

[0154] In an example testing scenario with a single-allergen patch design, the patch is applied and the patient uses an ink marker pen and colors over the fiducial cutout (e.g., skininking regions 18) as a stencil to leave an ink mark on the skin. Additionally or alternatively, the allergen test patch 10 may include self-inking markers that can transfer ink markings from the allergen test patch 10 to the skin surface. The patch remains in place for three days and then is removed by the patient. The patient or another individual takes a photograph of the skin patch site using an image capture device and the ink markings on the skin and / or fiducial markers (e.g., fiducial markers 24) may be used as a reference in the images. The image is uploaded automatically for offline analysis (e.g., via a computing device 810 and / or server 815). A report with interpretation may then be returned to the patient’s smartphone. If no reaction is detected, an allergy to the specific allergen is ruled out (with likely probability) and a suggestion is made regarding the next allergen(s) to be tested. The images are stored for expert overread if needed / desired.

[0155] Advantageously, the allergen test patches described in the present disclosure enable adaptive, sequential allergen testing. This testing strategy leverages specialized knowledge of the likelihood of reactions to certain allergens. A predetermined suggested sequence of allergen testing is recommended based on known prevalence and ultimately individualized risk assessment. As a non-limiting example, the suggested sequence may be generated using an artificial intelligence and / or machine learning-based approach that processes input data such as an exposure questionnaire and / or genomic data such as genetic predisposition to allergens.

[0156] Additionally or alternatively, the recommender model may process data including detailed metadata, such as past reactions, environmental exposures, and medical history, to generate a ranked list of allergens most likely to cause an allergic reaction for participating patient. For example, a patient working in an environment with frequent plant36QB\630666.01678X100992613.2Mayo 2023-378630666.01678exposure who reports reactions to certain skincare products might be recommended to test for propolis as their top allergen. In some implementations, this data may be gathered through the Patient Journey instmment, which systematically collects patient-provided information through its five journeys or sections: a background section for gathering demographic and ancestry information; a medical history section for documenting relevant health conditions; a work and life environments section for identifying potential allergen exposures; a section for items the patient suspects have caused allergic reactions; and a treatments section for documenting treatments the patient has used for skin conditions. This personalized approach ensures that the diagnostic and treatment process for allergic contact dermatitis is not only data-driven, but also uniquely tailored to each patient’s individual experiences and needs.

[0157] In an example testing scenario, the patient may start with the single-allergen testing model. Before beginning the testing process, the patient may complete the Patient Journey assessment via the mobile application in patient mode, providing baseline information about their background, medical history, work and life environments, suspected allergens, and treatments used for skin conditions. This information may inform the initial allergen selection and subsequent testing recommendations. The patch is applied to the skin by the patient (e.g., volar aspect of the forearm) and used as described above. The results of the test for the most common allergen for the patient may be used to generate a recommendation for the second allergen to be tested. This next allergen could be automatically shipped to the patient upon confirmation of the results. The sequence of allergens may be customized to the individual person based on these results. The testing may be terminated at the patient’s request or based on a probability threshold for the likelihood of a reaction (e.g., less than 1 in a 1,000 chance a reaction would occur).

[0158] Referring now to FIG. 13. a flowchart is illustrated as setting forth the operations of an example method for processing patch test images using a machine learning model to detect allergen reactions in the patch test images and / or to generate classified feature data that indicate a severity of the allergen reaction. As will be described, the machine learning model takes patch test images as inputs and generates classified feature data as an output. As an example, the classified feature data can indicate the presence of an allergen reaction and / or the severity of a detected allergen reaction.

[0159] The method includes accessing image data with a computer system (e.g., computing device 810. server 815), as indicated at block 1302. Accessing the image data may include retrieving such data from a memory or other suitable data storage device or medium.37QB\630666.01678X100992613.2Mayo 2023-378630666.01678Additionally or alternatively, accessing the image data may include acquiring such data with a camera or other imaging system (e.g., image capture device 805) and transferring or otherwise communicating the data to the computer system.

[0160] A trained machine learning model is then accessed with the computer system, as indicated at block 1304. In general, the machine learning model is trained, or has been trained, on training data to detect the presence of allergen reactions in patch test images and / or to estimate a severity of an allergen reaction.

[0161] Accessing the trained machine learning model may include accessing model parameters (e.g., weights, biases, or both) that have been optimized or otherwise estimated by training the machine learning model on training data. In some instances, retrieving the machine learning model can also include retrieving, constructing, or otherwise accessing the particular model architecture to be implemented. For instance, data pertaining to the layers in a neural network architecture (e.g., number of layers, type of layers, ordering of layers, connections between layers, hyperparameters for layers) may be retrieved, selected, constructed, or otherwise accessed.

[0162] An artificial neural network generally includes an input layer, one or more hidden layers (or nodes), and an output layer. Typically, the input layer includes as many nodes as inputs provided to the artificial neural netw ork. The number (and the type) of inputs provided to the artificial neural network may vary based on the particular task for the artificial neural network.

[0163] The input layer connects to one or more hidden layers. The number of hidden layers varies and may depend on the particular task for the artificial neural network. Additionally, each hidden layer may have a different number of nodes and may be connected to the next layer differently. For example, each node of the input layer may be connected to each node of the first hidden layer. The connection between each node of the input layer and each node of the first hidden layer may be assigned a weight parameter. Additionally, each node of the neural network may also be assigned a bias value. In some configurations, each node of the first hidden layer may not be connected to each node of the second hidden layer. That is, there may be some nodes of the first hidden layer that are not connected to all of the nodes of the second hidden layer. The connections between the nodes of the first hidden layers and the second hidden layers are each assigned different weight parameters. Each node of the hidden layer is generally associated with an activation function. The activation function defines how the hidden layer is to process the input received from the input layer or from a previous38QB\630666.01678X100992613.2Mayo 2023-378630666.01678input or hidden layer. These activation functions may vary and be based on the type of task associated with the artificial neural network and also on the specific type of hidden layer implemented.

[0164] Each hidden layer may perform a different function. For example, some hidden layers can be convolutional hidden layers which can, in some instances, reduce the dimensionality of the inputs. Other hidden layers can perform statistical functions such as max pooling, which may reduce a group of inputs to the maximum value; an averaging layer; batch normalization; and other such functions. In some of the hidden layers each node is connected to each node of the next hidden layer, which may be referred to then as dense layers. Some neural networks including more than, for example, three hidden layers may be considered deep neural networks.

[0165] The last hidden layer in the artificial neural network is connected to the output layer. Similar to the input layer, the output layer typically has the same number of nodes as the possible outputs.

[0166] The image data are then input to the one or more trained machine learning models, generating output as classified feature data, as indicated at block 1306. For example, the classified feature data may include an indication that an allergen reaction has been detected in a patch image. Additionally or alternatively, the classified feature data may indicate a severity of an allergen reaction. For example, the classified feature data may include a severity score that quantifies a severity of an allergen reaction. In other embodiments, the classified feature data may include one or more statistics about the reaction, such as a size of the reaction and / or a height of each reaction.

[0167] In some cases after classified feature data are generated by inputting images from the image data to the one or more trained machine learning models, a determination may be made at decision block 1308 whether additional images from the image data should be processed. When additional images should be processed, the method may loop back to block 1306 to apply additional images to the trained machine learning model (s) to generate additional classified feature data. In some implementations, the image data and classified feature data may be used to retrain, fine-tune, or otherwise update the machine learning model(s) as indicated at optional block 1310. In these instances, the machine learning model(s) may first be updated and then the image data reapplied to the now updated machine learning model(s) to generate updated classified feature data.39QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0168] The classified feature data generated by inputting the image data to the trained machine learning model(s) can then be displayed to a user, stored for later use or further processing, or both, as indicated at block 1312. In some implementations, the classified feature data may be output as part of a report. The report may include images, text, or combinations thereof. The report may be presented to the patient (e.g., via the computing device 810), stored in the patient's EHR, sent to a clinical team, or combinations thereof.

[0169] In some embodiments, the classified feature data output by the model may be evaluated, such as by a physician. This evaluation, in turn, may be used to generate new models with improved performance. For instance, as described herein, images acquired with the image capture device and resulting classified feature data generate from those images may be stored for use as training data for training, retraining, fine-tuning, or otherwise updating machine learning models. By providing the images and / or classified feature data to a physician, the images and / or classified feature data may also be further labeled or annotated by the physician to provide additional information useful for training, retraining, fine-tuning, or otherwise updating machine learning models.

[0170] Referring now to FIG. 14, a flowchart is illustrated as setting forth the operations of an example method for training one or more machine learning models on training data, such that the one or more machine learning models are trained to receive patch test images as an input in order to generate classified feature data as an output, where the classified feature data are indicative of detecting the presence of allergen reactions in the patch test images, a severity of allergen reactions, or both.

[0171] In general, the machine learning model(s) can implement any number of different model architectures. For instance, the machine learning models may implement neural network(s), such as a CNN, a residual neural network, or the like. In still other examples, the machine learning model(s) may implement a YOLO model architecture. Alternatively, the machine learning model(s) could be replaced with other suitable machine learning or artificial intelligence algorithms, such as those based on supervised learning, unsupervised learning, deep learning, ensemble learning, dimensionality reduction, and so on.

[0172] The method includes accessing training data with a computer system, as indicated at block 1402. Accessing the training data may include retrieving such data from a memory or other suitable data storage device or medium. Alternatively, accessing the training data may include acquiring such data with an imaging system and transferring or otherwise communicating the data to the computer system.40QB\630666.01678X100992613.2Mayo 2023-378630666.01678

[0173] In general, the training data can include patch test images that have been annotated or otherwise labeled with information indicating the presence of an allergic reaction and / or the severity of the allergic reaction.

[0174] The method can include assembling training data from patch test images using a computer system. This operation may include assembling the patch test images into an appropriate data structure on which the neural network or other machine learning algorithm can be trained. Assembling the training data may include assembling patch test images, segmented patch test images, and other relevant data. For instance, assembling the training data may include generating labeled data and including the labeled data in the training data. Labeled data may include patch test images, segmented patch test images, or other relevant data that have been labeled as belonging to, or otherwise being associated with, one or more different classifications or categories. For instance, labeled data may include patch test images and / or segmented patch test images that have been labeled as being associated with an allergic reaction.

[0175] As one non-limiting example, the training data may include patch test images from a curated annotated database. In one example, training data may include a dataset of skin patch tests performed in a clinical setting. The data include clinical, documentation photographs taken with either a professional camera system or a smart phone at time of care. The images may utilize a standard clinical test patch setup (e.g., 10 Finn Chambers secured with Scanpor tape, arranged in a 5x2 array, test site boundary indicated with dashed lines around the border). Patch level photographs can be generated by image cropping when only a full back is available. The data set may also contains annotations on visual attributes of the patches (e.g., “hairy skin”, “photo is blurry ”). Additionally or alternatively, the data labels may include data pertaining to the allergen tested at each test site, labels for novel patch design, reaction grading as determined by a board-certified dermatologist and skin type labeled fair, moderate or dark, and so on.

[0176] In another example, the training data may include patch test images acquired from patients using a single patch with fiducial markings and a location-randomized combination of ten allergens tested on the volar aspect of the forearm in a group of participants. The images may be captured by a smartphone using resolution of 1080 x 1920 pixels. Images were captured on days 1, 3 and 5 using up to three camera configurations at several zoom levels. Data labels include data pertaining to the allergen tested at each test site, reaction41QB\630666.01678X100992613.2Mayo 2023-378630666.01678grading as determined by a board-certified dermatologist and skin type graded on Fitzpatrick Scale.

[0177] One or more machine learning models are trained on the training data, as indicated at block 1404. In general, the machine learning model can be trained by optimizing model parameters (e.g., weights, biases, or both) based on minimizing a loss function. As one non-limiting example, the loss function may be a mean squared error loss function.

[0178] By way of example, training a neural network may include initializing the neural network, such as by computing, estimating, or otherwise selecting initial network parameters (e.g., weights, biases, or both). During training, an artificial neural network receives the inputs for a training example and generates an output using the bias for each node, and the connections between each node and the corresponding weights. For instance, training data can be input to the initialized neural network, generating output as classified feature data. The artificial neural network then compares the generated output with the actual output of the training example in order to evaluate the quality of the classified feature data. For instance, the classified feature data can be passed to a loss function to compute an error. The current neural network can then be updated based on the calculated error (e.g., using backpropagation methods based on the calculated error). For instance, the current neural network can be updated by updating the network parameters (e.g., weights, biases, or both) in order to minimize the loss according to the loss function. The training continues until a training condition is met. The training condition may correspond to, for example, a predetermined number of training examples being used, a minimum accuracy threshold being reached during training and validation, a predetermined number of validation iterations being completed, and the like. When the training condition has been met (e.g., by determining whether an error threshold or other stopping criterion has been satisfied), the current neural network and its associated network parameters represent the trained neural network. Different ty pes of training processes can be used to adjust the bias values and the weights of the node connections based on the training examples. The training processes may include, for example, gradient descent, Newton's method, conjugate gradient, quasi-Newton, Levenberg-Marquardt, among others.

[0179] The artificial neural network can be constructed or otherwise trained based on training data using one or more different learning techniques, such as supervised learning, unsupervised learning, reinforcement learning, ensemble learning, active learning, transfer learning, or other suitable learning techniques for neural networks. As an example, supervised learning involves presenting a computer system with example inputs and their actual outputs42QB\630666.01678X100992613.2Mayo 2023-378630666.01678(e.g., categorizations). In these instances, the artificial neural network is configured to learn a general rule or model that maps the inputs to the outputs based on the provided example inputoutput pairs.

[0180] The one or more trained machine learning models are then stored for later use, as indicated at block 1406. Storing the machine learning model(s) may include storing model parameters (e.g., weights, biases, or both), which have been computed or otherwise estimated by training the machine learning model(s) on the training data. Storing the trained machine learning model(s) may also include storing the particular model architecture to be implemented. For instance, data pertaining to the layers in a neural network architecture (e.g., number of layers, type of layers, ordering of layers, connections between layers, hyperparameters for layers) may be stored.

[0181] FIG. 15 shows an example of a system 1500 for automated allergen patch testing in accordance with some embodiments described in the present disclosure. As shown in FIG.15, a computing device 1550 can receive one or more types of data (e.g., patch test images) from data source 1502. In some embodiments, computing device 1550 can execute at least a portion of an automated allergen patch testing system 1504 to detect the presence and severity of allergen reactions from data received from the data source 1502.

[0182] Additionally or alternatively, in some embodiments, the computing device 1550 can communicate information about data received from the data source 1502 to a server 1552 over a communication network 1554, which can execute at least a portion of the automated allergen patch testing system 1504. In such embodiments, the server 1552 can return information to the computing device 1550 (and / or any other suitable computing device) indicative of an output of the automated allergen patch testing system 1504.

[0183] In some embodiments, computing device 1550 and / or server 1552 can be any suitable computing device or combination of devices, such as a desktop computer, a laptop computer, a smartphone, a tablet computer, a wearable computer, a server computer, a virtual machine being executed by a physical computing device, and so on. The computing device 1550 and / or server 1552 can also reconstruct images from the data.

[0184] In some embodiments, data source 1502 can be any suitable source of data(e.g., patch test images, patient health data, patient questionnaire data, etc.), such as an image capture device 805, another computing device (e.g., a server storing patch test images, patient health data, patient questionnaire data, etc.), and so on. In some embodiments, data source 1502 can be local to computing device 1550. For example, data source 1502 can be incorporated with43QB\630666.01678X100992613.2Mayo 2023-378630666.01678computing device 1550 (e.g., computing device 1550 can be configured as part of a device for measuring, recording, estimating, acquiring, or otherwise collecting or storing data). As another example, data source 1502 can be connected to computing device 1550 by a cable, a direct wireless link, and so on. Additionally or alternatively, in some embodiments, data source 1502 can be located locally and / or remotely from computing device 1550, and can communicate data to computing device 1550 (and / or server 1552) via a communication network (e.g., communication network 1554).

[0185] In some implementations, the server 1552 may include a secure, HIPAA-compliant cloud infrastructure that serves as a central enabling resource for the automated allergen patch testing system. The cloud infrastructure may be configured to store patient information and data securely, including patch test images, patient health data, patient questionnaire data, and test results. The cloud infrastructure may also manage the technological logistics of executing the Al algorithms centrally, which may minimize the technical requirements for performing computations locally on the computing device 1550 (e.g., a smartphone) and associated applications. By centralizing the Al algorithm execution on the server 1552, the system may reduce the processing burden on patient devices while maintaining consistent algorithm performance across different device types. The cloud architecture may include credential and authorization token administration, secure databases, and application programming interfaces (APIs) for data input and export. APIs that facilitate communications between the frontend. backend, and other systems or microservices may be encrypted using TLS 1.3 or other suitable encryption protocols. Strict role-based access controls may be implemented for different layers of the system, and audit and compliance logs may be generated and maintained to ensure regulatory compliance.

[0186] In some embodiments, communication network 1554 can be any suitable communication network or combination of communication networks. For example, communication network 1554 can include a Wi-Fi network (which can include one or more wireless routers, one or more switches, etc.), a peer-to-peer network (e.g., a Bluetooth network), a cellular network (e.g., a 3G network, a 4G network, etc., complying with any suitable standard, such as CDMA, GSM, LTE, LTE Advanced, WiMAX, etc.), other types of wireless network, a wired network, and so on. In some embodiments, communication network 1554 can be a local area network, a wide area network, a public network (e.g., the Internet), a private or semi-private network (e.g., a corporate or university intranet), any other suitable type of network, or any suitable combination of networks. Communications links shown in FIG. 1544QB\630666.01678X100992613.2Mayo 2023-378630666.01678can each be any suitable communications link or combination of communications links, such as wired links, fiber optic links, Wi-Fi links. Bluetooth links, cellular links, and so on.

[0187] Referring now to FIG. 16, an example of hardware 1600 that can be used to implement data source 1502, computing device 1550, and server 1552 in accordance with some embodiments of the systems and methods described in the present disclosure is shown.

[0188] As shown in FIG. 16, in some embodiments, computing device 1550 can include a processor 1602, a display 1604, one or more inputs 1606, one or more communication systems 1608, and / or memory 1610. In some embodiments, processor 1602 can be any suitable hardware processor or combination of processors, such as a central processing unit (CPU), a graphics processing unit (GPU), and so on. In some embodiments, display 1604 can include any suitable display devices, such as a liquid crystal display (LCD) screen, a light-emitting diode (LED) display, an organic LED (OLED) display, an electrophoretic display (e.g., an “e-ink” display), a computer monitor, a touchscreen, a television, and so on. In some embodiments, inputs 1606 can include any suitable input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, and so on.

[0189] In some embodiments, communications systems 1608 can include any suitable hardware, firmware, and / or software for communicating information over communication network 1554 and / or any other suitable communication networks. For example, communications systems 1608 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, communications systems 1608 can include hardware, firmware, and / or software that can be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0190] In some embodiments, memory 1610 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by processor 1602 to present content using display 1604, to communicate with server 1552 via communications system(s) 1608, and so on. Memory 1610 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof. For example, memory 1610 can include random-access memory (RAM), read-only memory (ROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM), other forms of volatile memory, other forms of non-volatile memory, one or more forms of semi-volatile memory, one or more flash drives, one or more hard disks, one or more solid state45QB\630666.01678X100992613.2Mayo 2023-378630666.01678drives, one or more optical drives, and so on. In some embodiments, memory 1610 can have encoded thereon, or otherwise stored therein, a computer program for controlling operation of computing device 1550. In such embodiments, processor 1602 can execute at least a portion of the computer program to present content (e.g., images, user interfaces, graphics, tables), receive content from server 1552, transmit information to server 1552, and so on. For example, the processor 1602 and the memory’ 1610 can be configured to perform the methods described herein (e.g., the method of FIG. 13, the method of FIG. 14).

[0191] In some embodiments, server 1552 can include a processor 1612, a display 1614, one or more inputs 1616, one or more communications systems 1618, and / or memory 1620. In some embodiments, processor 1612 can be any suitable hardware processor or combination of processors, such as a CPU, a GPU, and so on. In some embodiments, display 1614 can include any suitable display’ devices, such as an UCD screen, UED display, OLED display, electrophoretic display, a computer monitor, a touchscreen, a television, and so on. In some embodiments, inputs 1616 can include any suitable input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, and so on.

[0192] In some embodiments, communications systems 1618 can include any suitable hardware, firmware, and / or software for communicating information over communication network 1554 and / or any other suitable communication networks. For example, communications systems 1618 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, communications systems 1618 can include hardware, firmware, and / or software that can be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0193] In some embodiments, memory' 1620 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by processor 1612 to present content using display 1614, to communicate with one or more computing devices 1550, and so on. Memory 1620 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof. For example, memory 1620 can include RAM, ROM, EPROM, EEPROM, other types of volatile memory, other ty pes of non-volatile memory7, one or more types of semi-volatile memory, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical drives, and so on. In some embodiments, memory 1620 can have encoded thereon a server program46QB\630666.01678X100992613.2Mayo 2023-378630666.01678for controlling operation of server 1552. In such embodiments, processor 1612 can execute at least a portion of the server program to transmit information and / or content (e.g., data, images, a user interface) to one or more computing devices 1550, receive information and / or content from one or more computing devices 1550, receive instructions from one or more devices (e.g., a personal computer, a laptop computer, a tablet computer, a smartphone), and so on.

[0194] In some embodiments, the server 1552 is configured to perform the methods described in the present disclosure. For example, the processor 1612 and memory' 1620 can be configured to perform the methods described herein (e.g., the method of FIG. 13, the method of FIG. 14).

[0195] In some embodiments, data source 1502 can include a processor 1622. one or more data acquisition systems 1624, one or more communications systems 1626, and / or memory 1628. In some embodiments, processor 1622 can be any suitable hardware processor or combination of processors, such as a CPU, a GPU, and so on. In some embodiments, the one or more data acquisition systems 1624 are generally configured to acquire data, images, or both, and can include a camera or other imaging system, such as an image capture device 805. Additionally or alternatively, in some embodiments, the one or more data acquisition systems 1624 can include any suitable hardware, firmware, and / or software for coupling to and / or controlling operations of a camera or other imaging system, such as an image capture device 805. In some embodiments, one or more portions of the data acquisition system(s) 1624 can be removable and / or replaceable.

[0196] Note that, although not shown, data source 1502 can include any suitable inputs and / or outputs. For example, data source 1502 can include input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, a trackpad, a trackball, and so on. As another example, data source 1502 can include any suitable display devices, such as an LCD screen, an LED display, an OLED display, an electrophoretic display, a computer monitor, a touchscreen, a television, etc., one or more speakers, and so on.

[0197] In some embodiments, communications systems 1626 can include any suitable hardware, firmware, and / or software for communicating information to computing device 1550 (and, in some embodiments, over communication network 1554 and / or any other suitable communication networks). For example, communications systems 1626 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, communications systems 1626 can include hardware, firmware, and / or software that can be used to establish a wired connection using any suitable port and / or47QB\630666.01678X100992613.2Mayo 2023-378630666.01678communication standard (e.g., VGA, DVI video, USB, RS-232, etc.), Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0198] In some embodiments, memory' 1628 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by processor 1622 to control the one or more data acquisition systems 1624, and / or receive data from the one or more data acquisition systems 1624; to generate images from data; present content (e.g., data, images, a user interface) using a display; communicate with one or more computing devices 1550; and so on. Memory 1628 can include any suitable volatile memory', non-volatile memory', storage, or any suitable combination thereof. For example, memory 1628 can include RAM, ROM, EPROM, EEPROM, other types of volatile memory, other types of non-volatile memory, one or more types of semi-volatile memory, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical drives, and so on. In some embodiments, memory' 1628 can have encoded thereon, or otherwise stored therein, a program for controlling operation of data source 1502. In such embodiments, processor 1622 can execute at least a portion of the program to generate images, transmit information and / or content (e.g., data, images, a user interface) to one or more computing devices 1550, receive information and / or content from one or more computing devices 1550, receive instructions from one or more devices (e.g., a personal computer, alaptop computer, a tablet computer, a smartphone, etc.), and so on.

[0199] In some embodiments, any suitable computer-readable media can be used for storing instructions for performing the functions and / or processes described herein. For example, in some embodiments, computer-readable media can be transitory' or non-transitory. For example, non-transitory computer-readable media can include media such as magnetic media (e.g., hard disks, floppy disks), optical media (e.g., compact discs, digital video discs, Blu-ray discs), semiconductor media (e g., RAM, flash memory, EPROM, EEPROM), any suitable media that is not fleeting or devoid of any semblance of permanence during transmission, and / or any suitable tangible media. As another example, transitory computer-readable media can include signals on networks, in wires, conductors, optical fibers, circuits, or any suitable media that is fleeting and devoid of any semblance of permanence during transmission, and / or any suitable intangible media.

[0200] As used herein in the context of computer implementation, unless otherwise specified or limited, the terms “component,” “system,” “module,” “framework,” and the like are intended to encompass part or all of computer-related systems that include hardware.48QB\630666.01678X100992613.2Mayo 2023-378630666.01678software, a combination of hardware and software, or software in execution. For example, a component may be, but is not limited to being, a processor device, a process being executed (or executable) by a processor device, an object, an executable, a thread of execution, a computer program, or a computer. By way of illustration, both an application running on a computer and the computer can be a component. One or more components (or system, module, and so on) may reside within a process or thread of execution, may be localized on one computer, may be distnbuted between two or more computers or other processor devices, or may be included within another component (or system, module, and so on).

[0201] In some implementations, devices or systems disclosed herein can be utilized or installed using methods embodying aspects of the disclosure. Correspondingly, description herein of particular features, capabilities, or intended purposes of a device or system is generally intended to inherently include disclosure of a method of using such features for the intended purposes, a method of implementing such capabilities, and a method of installing disclosed (or otherwise known) components to support these purposes or capabilities. Similarly, unless otherwise indicated or limited, discussion herein of any method of manufacturing or using a particular device or system, including installing the device or system, is intended to inherently include disclosure, as embodiments of the disclosure, of the utilized features and implemented capabilities of such device or system.

[0202] The present disclosure has described one or more preferred embodiments, and it should be appreciated that many equivalents, alternatives, variations, and modifications, aside from those expressly stated, are possible and within the scope of the invention.49QB\630666.01678X100992613.2

Claims

Mayo 2023-378630666.01678CLAIMS1. An allergen test patch, comprising:a patch substrate having a top surface and a bottom surface;at least one allergen well coupled to the bottom surface of the patch substrate;a border removably coupled to and surrounding the patch substrate, the border having a top surface and a bottom surface;an adhesive disposed on the bottom surface of the border to adhere the border to a skin surface of a subject, thereby removably coupling the patch substrate to the skin surface of the subject, wherein the adhesive retains the border on the skin surface of the subject when removing the patch substrate from the border.

2. The allergen test patch of claim 1, further comprising at least one fiducial marker coupled to the border.

3. The allergen test patch of claim 1, further comprising at least one patch identifier coupled to the patch substrate.

4. The allergen test patch of claim 3, wherein the at least one patch identifier comprises a computer readable barcode.

5. The allergen test patch of claim 4, wherein the computer readable barcode comprises a quick response (QR) code.

6. The allergen test patch of claim 1, further comprising at least one skin-inking region formed in the patch substrate, wherein the at least one skin-inking region comprises a hole defining a region where ink can be applied to the skin surface of the subject.

7. The allergen test patch of claim 1, further comprising a fastener region disposed on the top surface of the border, the fastener region to receive a protective cover that spans a space left when removing the patch substrate from the border.50QB\630666.01678X100992613.2Mayo 2023-378630666.016788. The allergen test patch of claim 7, wherein the fastener region comprises hook-and-loop fasteners.

9. The allergen test patch of claim 1 , wherein the patch substrate has a plurality of allergen wells coupled to the bottom surface of the patch substrate.

10. The allergen test patch of claim 9, wherein each of the plurality of allergen wells is preloaded with an allergen.

11. The allergen test patch of claim 10, wherein each of the plurality’ of allergen wells is preloaded with a different allergen selected from a test group of allergens, wherein a number of allergens in the test group of allergens is greater than a number of the plurality of allergen w ells and the allergens selected from the test group of allergens are adapted to allow identi fication of a reaction to a specific allergen in the test group of allergens.

12. The allergen test patch of claim 9, wherein the plurality of allergen wells are arranged in an array on the bottom surface of the patch substrate.

13. The allergen test patch of claim 12, wherein the array comprises a first column of allergen wells and a second column of allergen wells.

14. The allergen test patch of claim 13, wherein the first column of allergen wells is offset from the second column of allergen w ells along a length of the second column of allergen wells.

15. The allergen test patch of claim 9, further comprising a plurality' of fiducial markers coupled to the patch substrate.

16. The allergen test patch of claim 15, wherein the patch substrate has a rectangular shape and each of the plurality' of fiducial markers are arranged in a comer of the patch substrate.51QB\630666.01678X100992613.2Mayo 2023-378630666.0167817. The allergen test patch of claim 16, wherein each fiducial marker comprises a recessed region containing ink such that when the patch substrate is applied to the skin surface of the subj ect the ink in each fiducial marker will transfer to the skin surface of the subject.

18. The allergen test patch of claim 15, wherein each of the plurality of fiducial markers has a shape that indicates an orientation of the patch substrate.

19. The allergen test patch of claim 9, further comprising a plurality of skin-inking regions formed in the patch substrate.

20. The allergen test patch of claim 19, wherein the plurality of skin-inking regions are formed as recessed regions on the bottom surface of the patch substrate to receive and retain ink such that when the patch substrate is applied to the skin surface of the subject the ink retained in the recessed regions will transfer to the skin surface of the subject.

21. An allergen test patch, comprising:a patch substrate having a top surface and a bottom surface;an allergen well coupled to the bottom surface of the patch substrate;a self-inking marker coupled to the bottom surface of the patch substrate, such that when the patch substrate is adhered to a skin surface of a subject the self-inking marker transfers ink to mark the skin surface of the subject with a marker that identifies a tested allergen; andat least one adhesive patch coupled to the bottom surface of the patch substrate to removably couple the patch substrate to the skin surface of the subject.

22. The allergen test patch of claim 21, wherein the allergen well and self-inking marker are arranged within an allergen test region of the patch substrate.

23. The allergen test patch of claim 22, further comprising a dissolvable cover that spans the allergen test region of the patch substrate.52QB\630666.01678X100992613.2Mayo 2023-378630666.0167824. The allergen test patch of claim 23, wherein the allergen well is preloaded with an allergen.

25. The allergen test patch of claim 22, further comprising a removable cover slip spanning the allergen test region of the patch substrate.

26. The allergen test patch of claim 22, further comprising a border surrounding the allergen test region of the patch substrate.

27. The allergen test patch of claim 26, wherein the border comprises a doublelined silicon boundary.

28. The allergen test patch of claim 22, wherein the allergen test region of the patch substrate is composed of a waterproof membrane.

29. An allergen test patch, comprising:a patch substrate having a top surface and a bottom surface;a plurality' of allergen wells coupled to the bottom surface of the patch substrate; at least one control fiber coupled to the bottom surface of the patch substrate, the at least one control fiber being composed of a control material to test as a control for allergen reactions.

30. The allergen test patch of claim 29, wherein the at least one control fiber comprises a single control fiber composed of aluminum.

31. The allergen test patch of claim 29, wherein the at least one control fiber comprises a plurality of control fibers each composed of a plurality of different control materials.

32. The allergen test patch of claim 31, wherein the plurality of different control materials comprises at least two of aluminum, a zinc alloy, and a stainless steel.53QB\630666.01678\100992613.2Mayo 2023-378630666.0167833. The allergen test patch of claim 32, wherein the stainless steel is a 304 stainless steel.

34. The allergen test patch of claim 29, further comprising a patch identifier coupled to the patch substrate, the patch identifier identifying at least one allergen to test with the allergen test patch.

35. An allergen test patch, comprising:an annular patch substrate having a top surface and a bottom surface;a plurality of allergen wells coupled to the bottom surface of the annular patch substrate, wherein the plurality of allergen wells are radially distributed around the annular patch substrate; anda fiducial support spanning a central aperture of the annular patch substrate, the fiducial support being shaped to indicate an orientation of the annular patch substrate.

36. The allergen test patch of claim 35, further comprising at least one fiducial marker coupled to fiducial support.

37. The allergen test patch of claim 35, further comprising a plurality of skininking regions formed in the annular patch substrate.

38. The allergen test patch of claim 37, wherein the plurality of skin-inking regions are formed as recessed regions on the bottom surface of the annular patch substrate to receive and retain ink such that when the annular patch substrate is applied to a skin surface of a subject ink retained in the recessed regions will transfer to the skin surface of the subject.

39. The allergen test patch of claim 37, wherein the plurality of skin-inking regions are radially distributed around an outer periphery of the annular substrate.

40. A method for processing patch test images to detect allergen reactions, comprising:54QB\630666.01678X100992613.2Mayo 2023-378630666.01678accessing, with a computer system, image data comprising at least one patch test image depicting a test site on a skin surface of a subject, the test site having been exposed to at least one allergen;accessing, with the computer system, a trained machine learning model configured to detect presence of allergen reactions in patch test images;inputting the image data to the trained machine learning model to generate classified feature data indicative of a detected presence of an allergen reaction at the test site; andat least one of displaying the classified feature data to a user or storing the classified feature data for later use.

41. The method of claim 40, wherein the classified feature data further indicates a severity of the detected allergen reaction.

42. The method of claim 41, wherein the severity of the detected allergen reaction is indicated by a severity score that quantifies the severity of the allergen reaction.

43. The method of claim 40, wherein the trained machine learning model comprises a convolutional neural network.

44. The method of claim 43, wherein the convolutional neural network implements a U-Net architecture configured to perform image segmentation to isolate a patch test area from surrounding skin in the at least one patch test image.

45. The method of claim 40, wherein the trained machine learning model comprises an object detection model configured to detect and localize positive reactions within the at least one patch test image by generating bounding boxes around reaction sites.

46. The method of claim 45, wherein the object detection model comprises a You Only Look Once (Y OLO) model architecture.

47. The method of claim 40, further comprising:55QB\630666.01678X100992613.2Mayo 2023-378630666.01678detecting, from the at least one patch test image, at least one fiducial marker on an allergen test patch applied to the test site; andpreprocessing the at least one patch test image based on the detected at least one fiducial marker to normalize at least one of brightness, contrast, distance, or orientation of the at least one patch test image.

48. The method of any one of claims 40 or 47, further comprising: detecting, from the at least one patch test image, at least one calibration marker on the allergen test patch; andadjusting at least one of exposure or white balance of the at least one patch test image based on the detected at least one calibration marker.

49. The method of claim 40, further comprising:detecting, from the at least one patch test image, a patch identifier on an allergen test patch applied to the test site; andidentifying, based on the detected patch identifier, the at least one allergen to which the test site was exposed.

50. The method of claim 40, further comprising:generating a report comprising the classified feature data; andtransmitting the report to at least one of a computing device associated with the subject or a healthcare provider.

51. The method of claim 40, further comprising when the classified feature data indicates a detected presence of an allergen reaction, flagging the at least one patch test image for review by a clinician.

52. A computer-implemented method for manufacturing a patient-specific allergen test patch, comprising:receiving, with a computer system, patient data comprising at least one of medical history7data, environmental exposure data, or suspected allergen reaction data associated with a patient;56QB\630666.01678X100992613.2Mayo 2023-378630666.01678processing the patient data with a recommender algorithm to generate a ranked list of allergens, wherein the ranked list of allergens identifies allergens having an increased likelihood of causing an allergic reaction for the patient based on the patient data;selecting, based on the ranked list of allergens, a patient-specific allergen panel comprising a subset of allergens from the ranked list of allergens; and manufacturing an allergen test patch comprising a plurality of allergen wells, wherein each allergen well of the plurality of allergen wells is loaded with a respective allergen from the patient-specific allergen panel.

53. The method of claim 52, wherein the patient data is received from a computer-adaptive patient history form configured to systematically gather patient-provided information.

54. The method of claim 53, wherein the computer-adaptive patient history form comprises:a background section for gathering demographic and ancestry7information;a medical history section for documenting relevant health conditions;a work and life environments section for identifying potential allergen exposures; a section for items the patient suspects have caused allergic reactions; and a treatments section for documenting treatments the patient has used for skin conditions.

55. The method of claim 52, wherein the recommender algorithm comprises a naive Bayes classifier configured to incorporate a probabilistic structure to rank order allergens in terms of recommendations.

56. The method of claim 52, wherein the recommender algorithm comprises a collaborative filtering algorithm configured to identify allergens associated with high reactivity7based on patient histories and observed reactions from other patients with similar histories.

57. The method of claim 52, further comprising:57QB\630666.01678X100992613.2Mayo 2023-378630666.01678receiving results from a previous allergen test performed on the patient: and updating the ranked list of allergens based on the results from the previous allergen test.

58. The method of claim 57, wherein selecting the patient-specific allergen panel comprises selecting allergens related to an allergen that caused a reaction in the previous allergen test.

59. The method of claim 52, wherein the allergen test patch further comprises: at least one fiducial marker to facilitate image registration; andat least one calibration marker for image processing.58QB\630666.01678X100992613.2

Images