System and method for determining skin lesion score for companion animal

Abstract

The present disclosure relates to a method for determining a skin lesion score for a companion animal using an evaluation module trained based on previously acquired metadata associated with the companion animal selected from a data set having at least animal data and skin area data, the method comprises at least the steps of: receiving metadata relating to a companion animal, including two or more items of animal data and two or more items of skin area data listed in a data set; encoding the metadata into a metadata vector; running the trained evaluation module on the metadata vector based on the analysis of the evaluation module on the metadata vector; generating a skin lesion score indicative of a skin condition of the companion animal; and assessing whether the companion animal has an atopic dermatitis condition based on a skin lesion score indicative of at least one skin condition of the companion animal.

Description

Cross Reference to Related Applications

[0001] This application claims the priority benefit of European Patent Application No. EP 23161761.4, filed March 14, 2023, the contents of which are incorporated herein in their entirety by this reference and priority to which is claimed. TECHNICAL FIELD

[0002] The present disclosure relates to companion animal health, and in particular to the study of atopic dermatitis conditions. Various embodiments of the present disclosure relate generally to machine learning based technology for monitoring companion animal health, and more particularly, to systems and methods for determining a skin lesion score of a companion animal suspected of having an atopic dermatitis condition. BACKGROUND

[0003] In most companion animals, healthy skin and hair are indicative of the animal’s overall good health. As the skin and hair condition of a companion animal has such a significant visual impact, particularly to its owner, it is a constant goal in the art to assess the skin condition of an animal.

[0004] It is the most common reason for companion animals to be taken to a veterinary clinic that they have a skin condition. Data shows that 15% of the workload of such veterinary clinics is directed to animals with skin conditions. One of the main skin conditions leading to this statistic is atopic dermatitis, a common, genetically predisposed, inflammatory, pruritic skin disease. The diagnosis of atopic dermatitis can be complicated by various variables, such as the diversity of its clinical presentation, genetic factors, lesion extent, disease stage, secondary infections, and similarity to other non-atopic related skin diseases.

[0005] It can be difficult to reliably classify skin lesions and efficiently guide the owner of a companion animal. Dermatology is a complex medical field, as the skin can reveal all external and internal diseases. A clinical examination is often required to diagnose a skin condition, which is a time-consuming process requiring in-depth knowledge of dermatology. Moreover, a symptomatic treatment is often directed to the skin condition only, which leads to the underlying skin condition not being treated.

[0006] A smartphone application called "Atopy index" was developed based on the Canine Atopic Dermatitis Lesion Index (CADLI) and the intensity of pruritus. This application helps veterinary practitioners to easily and quickly assess the extent and severity index of canine atopic dermatitis (CADESI-4) before and after treatment. The owner of the companion animal identifies the lesions by coloring them on the application, which is based on 20 different areas of the body surface and 3 types of lesions. This application allows to obtain the dermatological index, as well as the severity index, based on a validated severity scale in a few minutes, and to follow the evolution of the total score and severity of each affected area at each consultation. However, this application is not designed for the diagnosis of atopic dermatitis, but only as an aid for the management by the owner.

[0007] Favrot et al. in "Prospective study on the clinical features of chronic atopic dermatitis in dogs and their diagnostic value" (Veterinary Dermatology; 2010, vol. 21, no. 1, pages 23-31) describe a method to assess the skin status of companion animals using several criteria including skin area data. The user needs to use said criteria and further identify the pattern of the skin condition of the patient, which varies depending on the animal breed.

[0008] Therefore, there is still a need to improve the existing methods in the field of atopic dermatitis conditions and to provide an efficient, simple and reliable method to assist animal owners and / or veterinary practitioners to assess the skin status of companion animals.

[0009] Reference in this disclosure to any particular activity is for convenience only and is not intended to limit the disclosure. One of ordinary skill in the art will recognize that the concepts underlying the disclosed devices and methods can be used in any suitable activity. The disclosure can be understood with reference to the following description and drawings, in which like elements are referred to with identical reference numerals.

[0010] Although the terms used below are used in connection with the detailed description of certain specific examples of the disclosure, these terms can be interpreted in their most reasonable broadest way. In fact, certain terms can even be emphasized below; however, any term intended to be interpreted in any limiting way will be explicitly and specifically defined in the detailed description section. Both the foregoing general description and the following detailed description are exemplary and explanatory only, and are not restrictive of the claimed features. SUMMARY

[0011] According to certain aspects of the present disclosure, methods and systems for determining a skin lesion score of a companion animal are disclosed.

[0012] According to a first aspect, the present disclosure relates to a method for determining a skin lesion score of a companion animal suspected of having an atopic dermatitis condition, the method using an evaluation module pre-trained to learn features indicative of the atopic dermatitis condition of the companion animal based at least on a plurality of previously acquired metadata related to the companion animal selected in a data set (A) comprising at least animal data and skin region data, the method comprising at least the steps of: a) receiving metadata related to the companion animal, comprising two or more items of animal data and two or more items of skin region data listed in the data set (A), b) encoding the metadata into a metadata vector, c) running the trained evaluation module on the metadata vector, wherein the running comprises inputting the metadata vector into a prediction model, d) generating a skin lesion score indicative of at least one skin state of the companion animal based on an analysis of the metadata vector by the prediction model associated with the evaluation module, and e) assessing whether the companion animal has an atopic dermatitis condition based on the skin lesion score indicative of at least one skin state of the companion animal.

[0013] By crossing animal data and skin region data from the metadata, and by using the evaluation module pre-trained, a score indicative of at least one skin state of the companion animal can be obtained. Thus, the present disclosure can help the average veterinary practitioner to make a skin disease diagnosis. The method can come up with the skin disease diagnosis hypothesis that best matches the clinical signs and provide a more reliable and faster diagnosis and treatment opportunity for the companion animal.

[0014] In one embodiment, the metadata related to the companion animal selected in the data set (A) comprises at least animal data and skin region data, wherein: - the animal data comprises at least two or more of the following: (i) breed, (ii) species, (iii) gender, (iv) body weight, (v) spayed or neutered status, (vi) age, (vii) age of onset, (viii) body condition, (ix) health status, (x) lifestyle, (xi) living environment, (xii) coat information, (xiii) activity level, (xiv) biological values of biological samples, (xv) age of onset, and (xvi) history of dermatitis; and - the skin region data comprises at least two or more of the following: (i) inguinal region, (ii) axillary region, (iii) ventral thoracic region, (iv) perineum / genital region, (v) ventral cervical region, (vi) pinna region, (vii) periorbital region, (viii) perioral region, (ix) elbow flexion side region, (x) forefoot region, (xi) hock flexion side region, (xii) hind foot region, and (xiii) non-specific atopic region.

[0015] In one particular embodiment, the companion animal related metadata selected in said data set (A) comprises at least animal data and skin region data, wherein: - the animal data comprises at least (i) breed, (ii) species, (iii) lifestyle, (iv) age of onset, and (v) history of dermatitis, and can further comprise other data such as, but not limited to, gender, body weight, spayed or neutered status, age, body condition, health status, living environment, coat information, activity level, and biological values of biological samples; and - the skin region data comprises at least (i) inguinal region, (ii) axillary region, and (iii) non-specific atopic region, and can further comprise additional data such as, but not limited to, ventral thoracic region, perineum / genital region, ventral cervical region, pinna region, periorbital region, perioral region, elbow flexion side region, forefoot region, hock flexion side region, and hind foot region, In one particular embodiment, the companion animal related metadata selected in said data set (A) comprises at least animal data and skin region data, wherein: - the animal data comprises at least (i) breed, (ii) species, (iii) lifestyle, (iv) age of onset, and (v) history of dermatitis; and - the skin region data comprises at least (i) inguinal region, (ii) axillary region, and (iii) non-specific atopic region.

[0016] In one particular embodiment, the method further comprises the steps of receiving a description of a pattern associated with a lesion of said companion animal, and inputting the description into the predictive model. Generating the skin lesion score is further based on an analysis of the description by the predictive model.

[0017] In one particular embodiment, the method further comprises the steps of calculating second or higher order cross-features based on cross-feature interactions between said metadata, and inputting the second or higher order cross-features into the predictive model.

[0018] Generating the skin lesion score is further based on an analysis of the second or higher order cross-features by the predictive model.

[0019] According to another aspect, the disclosure relates to a method for training an evaluation module to learn features indicative of an atopic dermatitis condition of a companion animal, using at least a plurality of previously acquired metadata related to companion animals selected in a data set (A), the data set (A) comprising at least animal data and skin area data corresponding to affected or non-affected skin areas, the animal data comprising at least two or more of the following: (i) breed, (ii) species, (iii) gender, (iv) body weight, (v) spayed or neutered status, (vi) age, (vii) age of onset, (viii) body condition, (ix) health status, (x) lifestyle, (xi) living environment, (xii) coat information, (xiii) activity level, (xiv) biological values of biological samples, (xv) age of onset and (xvi) history of dermatitis, the skin area data comprising at least two or more of the following: (i) inguinal area, (ii) axillary area, (iii) thoraco-abdominal area, (iv) perineum / genital area, (v) cervical-abdominal area, (vi) pinna area, (vii) peri-orbital area, (viii) peri-oral area, (ix) elbow flexion area, (x) forepaw area, (xi) hock flexion area, (xii) hindpaw area and (xiii) non-specific atopic area, the method comprising: - extracting at least one feature from each previously acquired metadata, - associating at least said at least one feature with an animal skin status, and - training the evaluation module to learn said association.

[0020] According to another aspect, this disclosure relates to a device for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, the device comprising an assessment module pre-trained to learn characteristics indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected in a data set (A), the data set (A) comprising at least animal data and skin region data corresponding to affected or unaffected skin areas, the animal data comprising at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age of onset, (viii) physical condition, (ix) health status, (x) lifestyle, (xi) living environment, (xii) fur information, (xiii) activity level, ( (xiv) biological values ​​of the biological sample, (xv) age of onset and (xvi) history of dermatitis, and skin region data including at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral thorax region, (iv) perineal / genital region, (v) ventral cervical region, (vi) auricular region, (vii) perioral region, (viii) perioral region, (ix) flexor elbow region, (x) forefoot region, (xi) flexor tarsal joint region, (xii) hindfoot region and (xiii) nonspecific atopic region. The trained assessment module is configured to run on metadata associated with the companion animal, the metadata including two or more animal data and two or more skin region data listed in data set (A), and generate a skin lesion score indicating at least one skin condition of the companion animal.

[0021] Therefore, this disclosure provides an efficient and reliable digital tool to assist veterinary practitioners in making diagnoses.

[0022] According to another aspect, this disclosure relates to a computer program product for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, which uses an assessment module pre-trained to learn characteristics indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected in a dataset (A), the dataset (A) having at least animal data and skin region data corresponding to affected or unaffected skin areas, the animal data including at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age at onset, (viii) physical condition, (ix) health status, (x) lifestyle, (x i) living environment, (xii) fur information, (xiii) activity level, (xiv) biological values ​​of biological samples, (xv) age of onset and (xvi) history of dermatitis, and skin region data including at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral chest region, (iv) perineal / genital region, (v) ventral neck region, (vi) auricular region, (vii) perioral region, (viii) perioral region, (ix) elbow flexion region, (x) forefoot region, (xi) tarsal joint flexion region, (xii) hindfoot region and (xiii) nonspecific atopic region, the computer program product comprising a carrier and instructions stored on the carrier that can be read by a processor, the instructions being configured to: a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Encode the metadata into a metadata vector. c) Run the trained evaluation module on the metadata vector, wherein the run includes inputting the metadata vector into the prediction model. d) Based on the analysis of the metadata vector by the prediction model associated with the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal, and e) Assess whether the companion animal has atopic dermatitis based on a skin lesion score that indicates at least one skin condition of the companion animal.

[0023] According to another aspect, this disclosure relates to a computer-readable medium having a computer program product as defined above stored thereon.

[0024] According to one exemplary embodiment, this disclosure relates to a computer-implemented method for determining a skin lesion score in a companion animal, comprising: - Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A). - Encode the metadata into a metadata vector; - Run the trained evaluation module on the metadata vector, wherein the run includes inputting the metadata vector into the prediction model; - Based on the analysis of the metadata vector by the prediction model associated with the assessment module, a skin lesion score indicating at least one skin condition of the companion animal is generated; - Assess whether the companion animal suffers from atopic dermatitis based on a skin lesion score indicating at least one skin condition; and - Transmit skin lesion scores and assessments to the device.

[0025] In one embodiment of the computer-implemented method according to the present disclosure, a skin lesion score assesses at least one other dermatitis condition, which may include at least one of the following: scabies, demodicosis, bacterial overgrowth syndrome, Malassezia dermatitis, bacterial folliculitis, contact dermatitis, and / or mucocutaneous T-cell lymphoma.

[0026] In one embodiment of the computer-implemented method according to this disclosure, at least one skin condition may include at least one atopic dermatitis condition.

[0027] In one embodiment of this disclosure, the computer-implemented method may further include: - Retrieve previously stored skin scores; - Compare previously stored skin scores with skin lesion scores; and - Evaluate the efficacy of previous treatments based on comparison.

[0028] In one embodiment of this disclosure, the computer-implemented method may further include: - Recommendations are made based on skin lesion scores; and - Transmit the suggestion to the device.

[0029] In one embodiment of the computer-implemented method according to this disclosure, the advice may be at least one health advice, at least one nutritional advice, and / or at least one medical advice.

[0030] In one embodiment of the computer-implemented method according to this disclosure, at least one health recommendation may include at least one of the following: at least one food, at least one pet service, at least one supplement, at least one ointment, at least one drug, and / or at least one pet product.

[0031] In one embodiment of the computer-implemented method according to this disclosure, at least one nutritional recommendation may include at least one instruction to feed the companion animal at least one of the following: at least one supplement and / or at least one food.

[0032] In one embodiment of the computer-implemented method according to this disclosure, at least one medical advice may include at least one of the following: at least one ointment instruction and / or at least one medication instruction.

[0033] In one embodiment of the computer-implemented method according to this disclosure, the trained evaluation module may include at least one supervised classifier machine learning model.

[0034] In one embodiment of the computer-implemented method according to this disclosure, the training of the trained evaluation module may further include: - Update the weights of at least one supervised classifier machine learning model based on the association between the animal's skin condition and at least one feature.

[0035] It should be understood that the foregoing invention content and the following description are merely exemplary and illustrative, and do not limit the claimed disclosed embodiments. Attached Figure Description

[0036] The accompanying drawings, which are included in and form part of this specification, illustrate various exemplary embodiments and, together with the description, serve to explain the principles of the disclosed embodiments.

[0037] Figure 1 A flowchart is provided illustrating an exemplary method for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, according to one or more embodiments.

[0038] Figure 2 Examples of use of the method according to one or more embodiments are shown, and an example of a user interface that allows input data (such as companion animal metadata) is depicted.

[0039] Figure 3 Examples of some of the most commonly affected areas of dogs of several breeds suffering from atopic dermatitis, according to one or more embodiments, are depicted.

[0040] Figure 4 Examples of some of the most commonly affected areas in dogs suffering from atopic dermatitis according to one or more embodiments are depicted.

[0041] Figure 5 Examples of some of the most commonly affected areas in cats with atopic dermatitis according to one or more embodiments are depicted.

[0042] Figure 6 A training example of the random forest classification algorithm used in this invention is described.

[0043] Figure 7 An example of using the evaluation module according to the present invention is described. Detailed Implementation

[0044] Definitions

[0045] In the specific implementation of this document, references to "embodiment," "an embodiment," "a non-limiting embodiment," "in various embodiments," etc., indicate that the embodiment may include specific features, structures, or characteristics, but each embodiment may not necessarily include that specific feature, structure, or characteristic. Furthermore, such phrases do not necessarily refer to the same embodiment.

[0046] Furthermore, when specific features, structures, or characteristics are described in conjunction with embodiments, it should be understood that, whether explicitly described or not, any combination of other embodiments to affect such features, structures, or characteristics falls within the knowledge scope of those skilled in the art. After reading this description, those skilled in the art will understand how to implement this disclosure in alternative embodiments.

[0047] Generally, terms can be understood at least in part based on their usage in the context. For example, terms such as “and,” “or,” or “and / or” as used herein can include a variety of meanings that can depend at least in part on the context in which such terms are used. Generally, “or,” when used in an associative list, such as A, B, or C, is intended to mean A, B, and C (used herein in an inclusive sense) and A, B, or C (used herein in an exclusive sense). Furthermore, the term “one or more” as used herein depends at least in part on the context and can be used to describe any feature, structure, or characteristic in a singular sense, or to describe a combination of features, structures, or characteristics in a plural sense.

[0048] Similarly, terms such as “a,” “an,” or “the” can be understood to express either a singular or a plural usage, depending at least in part on the context. Furthermore, the term “based on” can be understood not necessarily to express an exclusive set of factors; rather, it may allow for additional factors not explicitly described, again depending at least in part on the context.

[0049] The terms “having,” “comprising,” “containing,” and “including,” or any other variations thereof, are interchangeable, and those skilled in the art will understand that these terms are open-ended. They are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements may include not only those elements but also other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0050] As used herein, the terms “about” or “approximately” refer to a specific value within an acceptable margin of error as determined by a person skilled in the art, which will depend in part on how the value is measured or determined, i.e., the limitations of the measurement system.

[0051] As used herein, the terms “companion animal” or “pet” refer to non-human animals and may include, but are not limited to, companion mammals. For example, companion mammals may include canines, felines, dogs, cats, rabbits, hamsters, guinea pigs, rats, and / or mice. Preferred companion animals (but not limited to) are canines or felines, especially dogs and cats, particularly dogs.

[0052] As used herein, the term "mammal" may include, for example, but not limited to, humans or animals. In particular, the term "animal" may include, for example, but not limited to, ruminants, poultry, pigs, mammals, horses, mice, rats, rabbits, guinea pigs, hamsters, cattle, cats, or dogs, preferably companion animals, i.e., cats and / or dogs.

[0053] As used herein, the term "adult" can include, for example, but not limited to, animals that have passed puberty or have reached their biological maturity point, or both.

[0054] As used herein, the term "canines" can include animals, including companion animals selected from recognized breeds (some of which are further subdivided), which can include Afghan Hounds, Airedale Dogs, Akitas, Alaskan Malamutes, Basset Hounds, Beagles, Belgian Shepherds, Bloodhounds, Border Collies, Border Terriers, Borzois, Boxers, Bulldogs, French Bulldogs, Bull Terriers, Bullmastiffs, Cairn Terriers, Chihuahuas, Chow Chows, Cocker Spaniels, Collies, Corgis, Dachshunds, Dalmatians, Doberman Pinschers, Shar Peis, English Setters, Fox Terriers, German Shepherds, Golden Retrievers, and Great Danes. Greyhound, Brussels Griffon, Irish Setter, Irish Wolfhound, Cavalier King Charles Spaniel, Labrador Retriever, Lhasa Apso, Mastiff, Newfoundland, Old English Sheepdog, Papillon, Pekingese, Pointer, Pomeranian, Poodle, Pug, Rottweiler, Saint Bernard, Saluki, Samoyed, Schnauzer, Scottish Terrier, Shetland Sheepdog, Shih Tzu, Siberian Husky, Skye Terrier, Vizsla, Rhodesian Ridgeback, Staffordshire Bull Terrier, Jack Russell Terrier, Springer Spaniel, West Highland Terrier, Whippet, Yorkshire Terrier, Bichon Frise, etc.

[0055] As used herein, the term "feline" can include animals, including companion animals, selected from, but not limited to, cheetahs, cougars, jaguars, leopards, lions, lynxes, ligers, tigers, black panthers, lynxes, leopard cats, saber-toothed cats, caracals, servals, and cats. As used herein, cats include wildcats and domestic cats, with domestic cats being the preferred choice.

[0056] As used herein, a "subgroup" can include, for example, but not limited to, a collection of one or more animals of a species, but less than the entire species. For example, a "subgroup" can be defined based on genotype and / or one or more attributes of common physiological conditions among subgroup members in a subgroup with more than one member. In some embodiments, a subgroup can be defined at least partially by a specific breed. Furthermore, for example, in the case of mixed-breed animals, a subgroup can be defined at least partially by breed genetics, which can be determined by knowledge of parental breeds, phenotypic traits, genotypic assessments, or genetic markers (e.g., SNPs). In some embodiments, a subgroup can be defined at least partially by physiological condition.

[0057] As used herein, the term “metadata” or “metadata file” can include any companion animal data, such as, but not limited to, any or a combination of the animal’s attributes, including at least its breed, species, activity level, medical history, reproductive status, age, sex, weight, sterilization or castration status, biological values ​​of biological samples, physical condition, health status, lifestyle, living environment, fur information or risk factors, and / or medical data, such as age of onset of disease, presence of a history of acute hotspots, urticaria or angioedema, presence of corticosteroid-responsive pruritus, excessive shedding, scaling or dryness, gastrointestinal symptoms, indications of whether symptoms worsen after walking in grass, and / or a history of chronic and / or recurrent skin diseases or otitis media.

[0058] According to this disclosure, the term "animal data" may include at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age at onset, (viii) physical condition, (ix) health status, (x) lifestyle, (xi) living environment, (xii) fur information, (xiii) activity level, (xiv) biological values ​​of a biological sample, (xv) age at onset, and (xvi) history of dermatitis. In one particular embodiment, animal data includes at least (i) breed, (ii) species, (iii) lifestyle, (iv) age at onset, and (v) history of dermatitis. In some embodiments, animal data includes at least (i) breed, (ii) species, (iii) lifestyle, (iv) age at onset, and (v) history of dermatitis, and may also include other data, such as, but not limited to, sex, weight, sterilization or castration status, age, physical condition, health status, living environment, fur information, activity level, and biological values ​​of a biological sample; and

[0059] According to the instructions, the term "skin region data" may include at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral thoracolumbar region, (iv) perineal / genital region, (v) ventral cervical region, (vi) auricular region, (vii) periorbital region, (viii) perioral region, (ix) elbow flexion region, (x) forefoot region, (xi) tarsal joint flexion region, (xii) hindfoot and nonspecific atopic region.

[0060] - Skin region data include at least (i) the groin region, (ii) the axillary region and (iii) the nonspecific atopic region, and may also include additional data such as, but not limited to, the ventral thoracolumbar region, the perineal / genital region, the ventral cervical region, the auricular region, the periorbital region, the perioral region, the flexor elbow region, the forefoot region, the flexor tarsal joint region and the hindfoot region.

[0061] According to the instructions, the term "dermatological history" refers to the state of dermatitis over time, that is, whether the dermatitis is chronic, recurrent, or persistent.

[0062] According to the instructions, the term "nonspecific atopic region" can include all skin regions other than the following: (i) groin region, (ii) axillary region, (iii) ventral thoracolumbar region, (iv) perineal / genital region, (v) ventral cervical region, (vi) auricular region, (vii) periorbital region, (viii) perioral region, (ix) elbow flexion region, (x) forefoot region, (xi) tarsal joint flexion region, or (xii) hindfoot region.

[0063] Furthermore, for example, metadata can consist of answers to questions about (but not limited to) the aforementioned list of attributes.

[0064] As used herein, the term “evaluation module” may include, for example, but not limited to, modules associated with analytical data, learning models and learning algorithms, particularly for classification and regression analysis.

[0065] As used herein, the term "expert" can include, for example, but not limited to, individuals capable of identifying skin lesions on an animal's body surface, marking features associated with such lesions, distinguishing such lesions from and / or associating them with atopic dermatitis, other skin conditions, or other skin markings. Such an expert could, for example, be a veterinary practitioner.

[0066] As used herein, the term "biological sample" or "biological material" may include, for example, but not limited to, at least one of feces, urine, hair, blood, saliva, and tissue.

[0067] For example, the terms "biosample" or "biomaterial" can refer to tissue or fluid samples isolated from a subject, including but not limited to, blood, plasma, serum, feces, urine, bone marrow, bile, cerebrospinal fluid, lymphoid tissue and lymph, skin samples, skin, respiratory tract, intestinal and genitourinary tract secretions, tears, saliva, breast milk, blood cells, organs and / or biopsies. The terms "biosample" or "biomaterial" can also refer to samples of in vitro cell culture components, including but not limited to conditioned media produced by growing cells and tissues in culture media, such as recombinant cells and cell components. The terms "biosample" or "biomaterial" can also refer to, for example, but not limited to, polypeptides or polynucleotides, or fragmented portions of organisms or cells obtained from environmental sampling, such as airborne pathogens.

[0068] As used herein, the terms “food” or “food composition” or “diet” or “food” can refer to, for example, but not limited to, food, diet, food supplements, liquids and / or materials that may contain protein, carbohydrates and / or crude fat. For example, the term can also refer to supplements or additives such as minerals, vitamins and flavorings (see Merriam-Webster's Collegiate Dictionary, 10th edition, 1993, the entire contents of which are hereby incorporated by reference). Such food compositions or foods may or may not be nutritionally complete.

[0069] As used herein, “companion food” or “animal food” can include, for example but not limited to, products manufactured by companion food manufacturers, whether processed, partially processed or unprocessed, and / or products intended for consumption by companion animals after being placed on the market, in accordance with EU Regulation No. 767 / 2009.

[0070] As used herein, a “training dataset” or “training data” can include one or more data used to train a machine learning model. Training datasets can be collected via one or more client devices (e.g., crowdsourcing) or from other sources (e.g., databases). In some non-limiting embodiments, a training dataset for pet health assessment may include data from treatment and control groups.

[0071] Certain non-limiting embodiments are described below with reference to block diagrams and operating instructions of methods, processes, apparatuses, and devices. It should be understood that each block in the block diagram or operating instructions, and combinations of blocks in the block diagram or operating instructions, can be implemented by analog or digital hardware and computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer to modify its functionality (as described herein), a special-purpose computer, an ASIC, or other programmable data processing means, such that instructions executed by the processor of the computer or other programmable data processing means implement the function / action specified in the block diagram or one or more operating blocks. In some alternative embodiments, the functions / actions shown in the blocks may not be performed in the order shown in the operating instructions. For example, two blocks displayed consecutively may actually be executed substantially simultaneously, or sometimes in reverse order, depending on the functions / actions involved.

[0072] In some non-limiting embodiments, the term "server" should be understood as referring to a point of service that provides processing, database, and communication facilities. By way of example and not limitation, the term "server" can refer to a single physical processor with associated communication, data storage, and database facilities, or it can refer to a networked or clustered processor complex (e.g., a resilient computing cluster) and associated network and storage devices, as well as the operating software and one or more database systems and application software that support the services provided by the server. For example, a server can be a cloud-based server, a cloud computing platform, or a virtual machine. Servers can vary greatly in configuration or capabilities, but typically a server can include one or more central processing units and memory. A server can also include one or more mass storage devices, one or more power supplies, one or more wired or wireless network interfaces, one or more input / output interfaces, or one or more operating systems, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc.

[0073] For some non-limiting embodiments, "network" should be understood as a network that can couple devices to enable communication to be exchanged, for example, between server and client devices or other types of devices (including, for example, between wireless devices coupled via a wireless network). A network may also include mass storage, such as Network Attached Storage (NAS), Storage Area Network (SAN), or other forms of computer-readable or machine-readable media. A network may include the Internet, one or more local area networks (LANs), one or more wide area networks (WANs), wired connections, wireless connections, cellular networks, or any combination thereof. Similarly, subnetworks that may employ different architectures or may follow or be compatible with different protocols can interoperate within a larger network. For example, various types of devices can be provided to offer interoperability for different architectures or protocols. As an illustrative example, a router can provide links between otherwise separate and independent LANs.

[0074] Communication links or channels may include, for example, analog telephone lines (e.g., twisted-pair cables), coaxial cables, full-rate digital lines or partial-rate digital lines (including T1, T2, T3, or T4 type lines), Integrated Services Digital Network (ISDN), Digital Subscriber Line (DSL), wireless links (including satellite links), or other communication links or channels (e.g., those known to those skilled in the art). Furthermore, computing devices or other related electronic devices may be remotely coupled to the network, for example, via wired or wireless lines or links.

[0075] In some non-limiting embodiments, "wireless network" should be understood as coupling client devices to a network. Wireless networks can employ standalone self-organizing networks, mesh networks, wireless local area networks (WLANs), cellular networks, etc. A wireless network can be configured as a system comprising terminals, gateways, routers, etc., coupled via radio links, which can move freely and randomly or organize themselves arbitrarily, allowing the network topology to change, sometimes even rapidly. Wireless networks can also employ various network access technologies, including Wi-Fi, Long Term Evolution (LTE), WLAN, Wireless Router (WR) mesh networks, or second-generation (2G), third-generation (3G), fourth-generation (4G), and fifth-generation (5G) cellular technologies. For example, network access technologies can achieve wide-area coverage for devices (e.g., client devices with varying degrees of mobility). For example, a network can be configured to provide radio frequency (RF) or wireless communication via one or more network access technologies, such as Global System for Mobile Communication (GSM), Universal Mobile Telecommunications System (UMTS), General Packet Radio Service (GPRS), GSM Enhanced Data Rate Evolution (EDGE), 3GPP LTE, LTE Advanced, Wideband Code Division Multiple Access (WCDMA), Bluetooth, 802.11b / g / n, etc. A wireless network can include virtually any type of wireless communication mechanism through which signals can be transmitted between devices (e.g., client devices or computing devices), between networks, or within a network. A computing device can send or receive signals, for example, via a wired or wireless network, or can process or store signals, for example, in memory in a physical memory state. For example, a computing device can operate as a server and can include, for example, a dedicated rack server, a desktop computer, a laptop computer, a set-top box, an integrated device combining various features (e.g., two or more features of the aforementioned devices), etc. Servers can vary greatly in configuration and capabilities, but typically a server may include one or more central processing units and memory. A server may also include one or more mass storage devices, one or more power supplies, one or more wired or wireless network interfaces, one or more input / output interfaces, or one or more operating systems.

[0076] As used herein, a “machine learning model” generally encompasses a model whose instructions, data, and / or configuration are adapted to receive input and apply one or more of weights, biases, classifications, or analyses to the input to generate an output. The output may include, for example, a classification of the input, an analysis based on the input, a design, process, prediction, or recommendation associated with the input, or any other suitable type of output. Machine learning models are typically trained using training data (e.g., empirical data and / or samples of input data) that are fed into the model to establish, adjust, or modify one or more aspects of the model, such as weights, biases, criteria for forming classifications or clusters, etc. Aspects of a machine learning model may operate on the input linearly and in parallel via a network (e.g., but not limited to supervised classifier machine learning models) or via any suitable configuration.

[0077] The execution of a machine learning model may include deploying one or more classifier machine learning techniques or models, such as linear models (including, for example, logistic regression), ensemble methods (including, for example, random forest classifiers), gradient boosted machines (GBM), discriminant analysis (including, for example, partial least squares discriminant analysis), support vector classifiers, nearest neighbor (including, for example, K-nearest neighbor classifiers), Gaussian process classifiers, Naive Bayes (including, for example, Gaussian Naive Bayes classifiers), decision tree classifiers, neural network models (including, for example, multilayer perceptron classifiers), deep learning, and / or deep neural networks. Supervised training may be employed. For example, supervised learning may include providing training data and corresponding labels to the training data, such as ground truth. Any suitable type of training may be used, such as randomized training, gradient boosting training, randomized seed training, recursive training, epoch-based training, or batch training. In an exemplary use case, the machine learning model may determine a skin lesion score for a companion animal. The computer system may be configured to first receive companion animal data corresponding to the companion animal. For example, companion animal data may include at least one metadata file. The trained assessment module can then analyze or run on companion animal data, generating a skin lesion score based on the analysis. The skin lesion score can, for example, indicate at least one skin condition in the companion animal, where one of the at least one skin condition may include at least one atopic dermatitis condition. The computer system can then be configured to transmit the skin lesion score to a device that may belong to a veterinarian and / or the companion animal's guardian. In another exemplary use case, a machine learning model (e.g., the trained assessment module) can be trained to generate the skin lesion score. Training may include utilizing training data. For example, the trained assessment module may extract at least one feature from each of a plurality of previously acquired metadata. The trained assessment module can then associate the at least one feature with the companion animal's skin condition. This process can continue until associations can be made for each of the previously acquired metadata. The trained assessment module can then be further trained based on such associations.

[0078] While the examples above relate to skin lesions and atopic dermatitis conditions, it should be understood that the techniques according to this disclosure are applicable to any suitable type of companion animal health analysis. It should also be understood that the examples above are merely illustrative. The techniques and processes of this disclosure can be applied to any suitable activity.

[0079] Method for determining a skin lesion score of a companion animal

[0080] According to one aspect, this disclosure relates to a method for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, the method using an assessment module pre-trained to learn features indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected from a dataset (A) having at least animal data and skin region data, the method comprising at least the following steps:

[0081] a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Run the trained evaluation module on the metadata received in step a), and c) Based on the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal.

[0082] In one embodiment, the companion animal is a canine or a feline.

[0083] In one embodiment, the companion animal is a dog or a cat.

[0084] In one embodiment, the companion animal is a dog.

[0085] Metadata relating to a companion animal

[0086] In one embodiment, the metadata may include the animal’s breed, indications about its lifestyle (particularly whether it is primarily an indoor animal), the age of onset of atopic dermatitis, indications about whether the condition is chronic, recurrent, or persistent, and indications about at least one affected skin area (particularly the groin, armpit, or other areas).

[0087] The data set (A) may also include reproductive status, risk factors and / or indications regarding at least one affected area of ​​the companion animal’s body surface, the presence of a history of acute hotspots, urticaria or angioedema, the presence of corticosteroid-induced pruritus, excessive shedding, scaling or dryness, gastrointestinal symptoms, indications regarding whether symptoms worsen after walking in grass, or a history of chronic and / or recurrent skin diseases or otitis.

[0088] The skin region data may preferably be selected from the head, especially the ear, auricle, periorbital, perioral and / or periorbital regions; the legs, especially the forefoot, hindfoot and / or intertoe regions; the elbow flexor surface, the flexor surface of the tarsal joint and / or the extensor surface of the wrist joint; and the trunk, especially the groin, armpit, ventral chest or ventral neck, groin region, axillary region and / or perineal / genital region.

[0089] Therefore, metadata can advantageously be derived from medical history and animal characteristics. Their use in the methods according to this disclosure can improve the specificity and sensitivity of the method. Metadata can consist of answers to questions posed to the owner and / or veterinary practitioner.

[0090] The method according to this disclosure may include the step of providing a pathological profile of the companion animal based on the plurality of metadata. In some embodiments, the pathological profile may include an easily retrieved, centrally stored information record containing the plurality of metadata of the companion animal.

[0091] Evaluation module

[0092] In one particular embodiment, the evaluation module may use a predictive model.

[0093] In one particular embodiment, the prediction model may include a supervised classifier machine learning model and a metadata encoding module.

[0094] In a preferred embodiment, the model is a random forest classification algorithm, as explained below and as follows: Figure 6 Training is performed as shown. In this embodiment, the complete dataset is split into a training set and a test set, wherein: Step 1: Select random samples from the given data or training set. Step 2: Build a decision tree for each training dataset. Step 3: Vote using an average decision tree or a majority vote using a decision tree. - Step 4: Verify model performance using a test set.

[0095] Other known classification algorithms can be used.

[0096] In other embodiments, the prediction model may include one or more neural networks.

[0097] Although specific implementations use training and test sets, prediction models can also be trained using cross-validation methods, which can be configured from 5-fold to 600-fold (leave-one-out cross-validation).

[0098] The metadata encoding module can be advantageously configured to convert the metadata (in particular, the answers to the questions as defined above) into a categorical vector. The question list has a predefined set of answers, and the metadata encoding module advantageously uses a binarization function to assign a 1 or a 0 to each possible answer, thereby generating a metadata vector.

[0099] In one particular embodiment, the evaluation module may also include evaluating a supervised classifier machine learning model.

[0100] The evaluation supervised classifier machine learning model can advantageously use the metadata vector to generate a score indicating the skin condition of the companion animal.

[0101] Each vector can be associated with a companion animal and the known skin condition of the animal.

[0102] A supervised classification machine learning model can be trained and optimized to minimize error and generate scores on test data or new input data. As an example, and not a limitation, additional, unselected features can be added to the model to further optimize it.

[0103] The pattern of animal skin condition can be a breed-dependent variable. In one particular embodiment, the predictive model can use a description of the lesion pattern and a susceptible breed. The description of the lesion pattern and the susceptible breed can be two independent variables. If there is an effect where one influences the other, and if this effect has already been captured in the dataset, using a random forest that can simulate nonlinearity will help simulate this effect. Further improvements can include adding third- or fourth-order cross-feature interactions to generate new variables and use these new variables as additional inputs to the predictive model. In some embodiments, the computer-implemented method may also include providing second- or higher-order cross-features computed based on selected features. In some embodiments of the computer-implemented method, the cross-features may include third-order cross-features.

[0104] Skin lesion score

[0105] The score can be a probability or a percentage, showing the confidence level of the lesion indicative of atopic dermatitis, such as low, medium, or high.

[0106] In one embodiment, the score can be in numerical form, such as a value of 0 (no dermatitis) or 1 (dermatitis).

[0107] In another embodiment, the score can be in numerical form, such as a value between 0 and 10. In yet another embodiment, the score can be in letter form, specifically indicating that the companion animal's skin condition is considered to belong to a certain skin condition type, such as group A, group B, group C, etc.

[0108] In yet another embodiment, the rating may be a category indicating the skin condition of the companion animal, or a category indicating the dermatitis condition (atopic or other) of the companion animal, or a combination thereof, such as a category and the probability that the animal belongs to the category.

[0109] The method according to this disclosure may also include the step of providing a user interface with the rating generated in step c) that is related to the companion animal.

[0110] The one or more scores can be transmitted to the user in any suitable manner, such as by displaying on the screen of an electronic device, printing, or by speech synthesis.

[0111] The one or more scores may be used as input values ​​in another procedure, and / or may be combined with other information, such as clinical data and / or biological data.

[0112] The methods disclosed herein can be used to suggest supplementary analyses to further optimize the scores. As an example, and not a limitation, scores can be improved by adding new questions / inputs. Optimal dietary recommendations associated with each score can also be provided.

[0113] The score generated in step c) relating to the companion animal can further assess the evolution of the severity of the atopic dermatitis. To assess the evolution of the severity of the atopic dermatitis, a particular embodiment may add the size of each lesion (e.g., in millimeters), which can be normalized to the pet's body size so that the percentage of lesion coverage on the pet's body can be assessed.

[0114] In one embodiment, if the score generated in step c) relating to the companion animal assesses that the at least one skin lesion does not indicate atopic dermatitis in the companion animal, then the score assesses other dermatitis conditions, such as scabies, demodicosis, bacterial overgrowth syndrome, Malassezia dermatitis, bacterial folliculitis, contact dermatitis, or mucocutaneous T-cell lymphoma. In this embodiment, different features and / or clinical criteria than those used for atopic dermatitis may be used to best describe such other dermatitis conditions.

[0115] In one embodiment, previous scores associated with the same companion animal are stored, and the method may further include a step of comparing the previous scores with a score generated in step c) to assess the efficacy of the treatment method employed. According to another embodiment, this disclosure relates to a method for assessing the nature of skin lesions in a companion animal suspected of having atopic dermatitis, the method using at least a pre-trained assessment module to learn characteristics indicative of atopic dermatitis in the companion animal, based at least on multiple previously acquired metadata related to the companion animal selected from a data set (A) that includes at least animal data and skin region data, the method comprising at least the following steps:

[0116] a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Run the trained evaluation module on the metadata, and c) Based on the assessment module, generate a score that at least assesses whether the metadata indicates atopic dermatitis in the companion animal.

[0117] Method for training an evaluation module to learn features indicative of an atopic dermatitis condition of a companion animal

[0118] According to another aspect, this disclosure relates to a method for training an evaluation module to learn characteristics indicative of atopic dermatitis in a companion animal, the method using at least a plurality of previously acquired metadata related to the companion animal selected from a data set (A) that includes at least animal data and skin region data, the method comprising: - Extract at least one feature from each previously acquired metadata. - Associate at least one of the aforementioned features with the animal's skin condition, and - The evaluation module is trained to learn the aforementioned associations.

[0119] In one specific embodiment, for learning, not all data comes from companion animals with skin lesions; some data may come from animals unaffected by skin lesions, while some data may come from animals with lesions due to conditions other than atopic dermatitis. The global learning dataset preferably includes atopic animals, animals affected by other skin diseases, and healthy animals.

[0120] In one embodiment, the evaluation module may include at least one supervised classifier machine learning model, the method including the step of updating the weights of the supervised classifier machine learning model based on the association between the at least one feature and the animal skin state.

[0121] The evaluation module can be configured to assign weights to each input during the learning phase, with weight optimization specifically performed by a solver of the "stochastic gradient descent" type.

[0122] In one embodiment, the association between the at least one feature and the animal's skin condition can be used to modify reference atopic dermatitis features in subsequent method iterations in order to improve the accuracy and reliability of the method according to this disclosure.

[0123] The features of the methods described above for determining skin lesion scores can be applied to the methods for training the evaluation module, and vice versa.

[0124] Apparatus, computer-implemented method and medium

[0125] The provided method can be a computer-implemented method.

[0126] Therefore, in one embodiment, the provided method for determining a skin lesion score in a companion animal suspected of having atopic dermatitis and / or for training an assessment module to learn characteristics indicative of atopic dermatitis in the companion animal can be implemented offline, i.e., not controlled by a device such as a computer-aided system; or alternatively implemented online, i.e., controlled by a computer-aided system, such as a device including means adapted to determine a skin lesion score in a companion animal suspected of having atopic dermatitis and having means adapted to perform the steps of the method; or alternatively, a combination of offline and online methods.

[0127] Therefore, according to one embodiment, this disclosure relates to an apparatus for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, the apparatus comprising an assessment module pre-trained to learn features indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected from a data set (A) that includes at least animal data and skin region data, the trained assessment module being configured to run on the metadata related to the companion animal, the metadata including two or more animal data items and two or more skin region data items listed in data set (A), and to generate a skin lesion score indicative of at least one skin condition in the companion animal.

[0128] The device may include an acquisition module for acquiring the metadata.

[0129] According to another aspect, this disclosure relates to an apparatus for training an assessment module according to a method for training an assessment module to learn characteristics of atopic dermatitis symptoms in a companion animal, as described above.

[0130] According to another embodiment, this disclosure relates to a computer program product for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, which uses an evaluation module pre-trained to learn characteristics indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected from a dataset (A) that includes at least animal data and skin region data. The computer program product includes a carrier and instructions stored on the carrier that can be read by a processor, the instructions being configured to: a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Run the trained evaluation module on the metadata received in step a), and c) Based on the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal.

[0131] According to another embodiment, this disclosure relates to a computer program product for training an evaluation module according to a method for training an evaluation module to learn characteristics of atopic dermatitis in companion animals, as described above.

[0132] According to another embodiment, this disclosure relates to a computer-readable medium on which one or both of the aforementioned computer program products are stored. Such a computer-readable medium may include or be composed of a physical embodiment of a data set, the data set including one or more datasets configurable in one or more databases. Therefore, it may include or be composed of a medium on which or on which such data can be stored. Such a computer-readable medium may also include or be composed of more than one medium; however, in this case, the media may be functionally linked.

[0133] The computer-aided system disclosed herein may typically include one or more user interfaces for inputting input data. Such input data may, for example, include metadata related to the companion animal.

[0134] The embodiments of this disclosure and all functional operations described in this specification can be implemented in digital electronic circuits, or in computer software, firmware, or hardware, including the structures disclosed in this specification and their structural equivalents, or combinations thereof. Embodiments of this disclosure can be implemented as one or more computer program products, i.e., one or more computer program instruction modules encoded on a computer-readable medium for execution by a data processing apparatus or for controlling the operation of a data processing apparatus.

[0135] Computer-readable media can be machine-readable storage devices, machine-readable storage substrates, storage devices, combinations of materials that realize machine-readable propagated signals, or combinations of one or more thereof. The term "data processing apparatus" encompasses all means, devices, and machines for processing data, including, for example, programmable processors, computers, or multiple processors or computers. In addition to hardware, the apparatus may also include code that creates an execution environment for the computer program in question, such as code constituting processor firmware, protocol stacks, database management systems, operating systems, or combinations of one or more thereof. Propagated signals are artificially generated signals, such as machine-generated electrical, optical, or electromagnetic signals, which are generated to encode information for transmission to a suitable receiver device.

[0136] Computer programs (also known as programs, software, software applications, scripts, or code) can be written in any programming language, including compiled or interpreted languages, and can be deployed in any form, including as standalone programs or as modules, components, subroutines, or other units suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program can be stored as part of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple co-located files (e.g., a file storing one or more modules, subroutines, or portions of code). A computer program can be deployed to execute on a single computer or on multiple computers located at a single site or distributed across multiple sites and interconnected by a communication network.

[0137] Computers can be embedded in other devices, such as mobile phones, personal digital assistants (PDAs), mobile audio players, and Global Positioning System (GPS) receivers (to name just a few). Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and storage devices, including, for example, semiconductor storage devices such as EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM optical disks.

[0138] To provide interaction with the user, embodiments of this disclosure can be implemented on a computer or smartphone having a display device for displaying information to the user and a keyboard and pointing or tactile devices through which the user can provide input to the computer.

[0139] Embodiments of this disclosure can be implemented in a computing system that may include backend components (e.g., as a data server), middleware components (e.g., an application server), or frontend components (e.g., a client computer having a graphical user interface or web browser through which a user can interact with an implementation of this disclosure), or any combination of one or more such backend, middleware, or frontend components. Components of the system can be interconnected via digital data communication (e.g., a communication network) of any form or medium. Examples of communication networks include local area networks (“LANs”) and wide area networks (“WANs”), such as the Internet.

[0140] According to a seventh aspect, this disclosure relates to a computer-implemented method for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, the method comprising: receiving at least one piece of metadata associated with the companion animal, determining a skin lesion score of the companion animal based on the data, and determining a recommendation for a pet owner based on the score. In one particular embodiment, the recommendation is a nutritional recommendation. In another particular embodiment, the method may further include the step of transmitting the recommendation to a mobile device of the pet owner.

[0141] Based on the rating, recommendations can be determined and sent to one or more of pet owners, veterinarians, researchers, and / or any combination thereof. For example, the recommendations may include one or more health advice to prevent companion animals from developing one or more diseases, symptoms, discomforts, and / or any combination thereof. For example, the recommendations may include one or more of the following: food, pet services, supplements, ointments, medications for improving pet health, pet products, and / or any combination thereof. In other words, the recommendations may be nutritional advice. In some embodiments, nutritional advice may include instructions to feed companion animals one or more of the following: chewing substances, supplements, food, and / or any combination thereof. In some embodiments, the recommendations may be medical advice. For example, medical advice may include instructions to apply ointments to companion animals, administer one or more medications to companion animals, and / or provide one or more medications to companion animals. The term "pet product" may include, for example, but not limited to, any type of product, service, or device designed, manufactured, and / or used by companion animals. For example, pet products may be toys, chewing substances, food, clothing, collars, medications, health tracking devices, location tracking devices, and / or any combination thereof. In another example, pet products may include pet genetic or DNA testing services. The term "pet owner" can include any individual, organization, and / or group of individuals who own and / or are responsible for any aspect of the care of a companion animal.

[0142] Embodiments

[0143] According to the examples of the method for determining skin lesion scores in companion animals suspected of having atopic dermatitis as disclosed in this disclosure, each step is... Figure 1 It is described in the text and will be described in detail below.

[0144] This method uses a pre-trained evaluation module to learn features indicative of atopic dermatitis in a companion animal, based at least on multiple previously acquired metadata related to the companion animal. The metadata is selected from a dataset (A) containing at least animal data and skin region data, wherein, in this example: - Animal data should include at least (i) breed, (ii) species, (iii) lifestyle, (iv) age of onset, and (v) history of dermatitis. - Skin region data include at least (i) the groin region, (ii) the axillary region and (iii) the nonspecific atopic region.

[0145] In step 10, companion animal data corresponding to the companion animal is received, wherein the companion animal data includes multiple metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in the data group (A), as described above.

[0146] During step 11, the trained evaluation module is run on the at least one metadata of the companion animal. During step 12, a score indicating the skin condition of the companion animal is generated based on the evaluation module.

[0147] Some optional steps can lead to a more reliable assessment, such as using more data, more testing, or more questions.

[0148] Figure 2 An example of using the method according to this disclosure via a smartphone application is shown.

[0149] In this example, a user of the method according to this disclosure is asked to identify the breed of their dog from a defined list of susceptible breeds. Figure 2 A user interface is shown that allows pet owners to input some data, namely metadata related to the companion animal, such as its known breed, the duration of the so-called dermatitis flare-ups, whether it lives indoors or outdoors, and whether it has chronic or recurrent dermatitis or presents as persistent dermatitis. Therefore, four pieces of animal data are used in this example.

[0150] Pet owners are also asked to select the areas of their pet's body most characteristic of any lesions, such as the head, armpits, groin, or feet. As mentioned above, all of this input data allows for the generation of a score indicating the animal's skin condition, which can be displayed on a subsequent interface. In this example, the user is asked to select an area from a predefined list of 12 areas to form skin area data.

[0151] The table below summarizes the performance obtained in this example using four animal data points and fifteen skin region data points.

[0152]

[0153] In a preferred embodiment, the evaluation module uses a prediction model that includes a supervised classifier machine learning model and a metadata encoding module, such as... Figure 7 As shown.

[0154] To train the supervised classifier machine learning model for the evaluation module, at least 100 companion animals with skin lesions, some healthy animals, and some animals with characteristics other than atopic dermatitis were used. This selection could include different ages, behaviors, all different coat colors, co-variations with age, and all different coat densities and compositions (primary / intermediate / secondary hair). The distribution of animals needed to be as broad as possible to construct the most important database.

[0155] In the example shown, the metadata encoding module is configured to encode the metadata (i.e., question q) i Answer a i (n) is converted into a categorical vector a1(1)...a i (n), where i is the number of questions and n is the number of predefined answers.

[0156] In the example shown, the evaluation supervised classifier machine learning model uses the metadata vector from the metadata encoding module to generate a score indicating the skin condition of the companion animal, based on a pre-learned vector associated with a known animal skin condition.

[0157] Figure 3 This diagram shows some common areas affected by atopic dermatitis in dogs of several common breeds. It can be noted that breed can influence the areas affected by the condition.

[0158] Figure 4 This shows some of the most common areas in dogs suffering from atopic dermatitis.

[0159] Figure 5 This shows some of the most common areas in cats with atopic dermatitis.

[0160] Therefore, the most commonly affected areas are usually the head, especially the ears, the area around the mouth and / or eyes, the legs, especially the forefoot and / or the area between the toes, the flexor surfaces of the tarsal joints and / or the extensor surfaces of the wrist joints, and the trunk, especially the groin, armpits, ventral and / or perineal areas.

[0161] It should be understood that in the foregoing description of exemplary embodiments of this disclosure, various features of this disclosure are sometimes combined in a single embodiment, drawing, or description thereof in order to simplify this disclosure and aid in understanding one or more of the various inventive aspects. However, this approach to disclosure should not be construed as reflecting an intention to require more features than expressly listed in each claim. Rather, as reflected in the following claims, the inventive aspect does not lie in all features of a single embodiment of the foregoing disclosure. Therefore, the claims following the detailed description are hereby expressly incorporated into this detailed description, wherein each claim exists independently as a separate embodiment of this disclosure.

[0162] Furthermore, while some embodiments described herein include certain features but not others in other embodiments, combinations of features from different embodiments are intended to be within the scope of this disclosure and form different embodiments, as will be understood by those skilled in the art. For example, any claimed embodiments may be used in any combination in the following claims.

[0163] Therefore, although certain embodiments have been described, those skilled in the art will recognize that other and further modifications can be made thereto without departing from the spirit of this disclosure, and it is intended that all such changes and modifications fall within the scope of this disclosure. For example, functions can be added or removed from the block diagrams, and the operation of function blocks can be interchanged. Steps can be added or removed from the methods described within the scope of this disclosure.

[0164] The subject matter disclosed above should be considered illustrative rather than restrictive, and the appended claims are intended to cover all such modifications, enhancements, and other implementations falling within the true spirit and scope of this disclosure. Therefore, to the fullest extent permitted by law, the scope of this disclosure should be determined by the broadest permissible interpretation of the following claims and their equivalents, and should not be limited or constrained by the foregoing detailed description. Although various embodiments of this disclosure have been described, it will be apparent to those skilled in the art that further embodiments are possible within the scope of this disclosure. Therefore, this disclosure is not limited thereto, but should be limited by the appended claims and their equivalents.

Claims

1. A method for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, using an assessment module pre-trained to learn characteristics indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected in a dataset (A), said dataset (A) comprising at least animal data and skin region data corresponding to affected or unaffected skin areas, said animal data comprising at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age at onset, (viii) physical condition, (ix) health status, (x) Lifestyle, (xi) living environment, (xii) fur information, (xiii) activity level, (xiv) biological values ​​of biological samples, (xv) age of onset, and (xvi) history of dermatitis, and the skin region data includes at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral thoracolumbar region, (iv) perineal / genital region, (v) ventral cervical region, (vi) auricular region, (vii) periorbital region, (viii) perioral region, (ix) flexor elbow region, (x) forefoot region, (xi) flexor tarsal joint region, (xii) hindfoot region, and (xiii) nonspecific atopic region, the method comprising at least the following steps: a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Encode the metadata into a metadata vector. c) Run the trained evaluation module on the metadata vector, wherein the run includes inputting the metadata vector into the prediction model. d) Based on the analysis of the metadata vector by the prediction model associated with the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal, and e) Assess whether the companion animal has atopic dermatitis based on the skin lesion score, which indicates at least one skin condition of the companion animal.

2. The method according to claim 1, further comprising the following step: Receive a description of the pattern associated with the lesions in the companion animal; and Input the description into the prediction model; The generation of the skin lesion score is also based on the analysis of the description by the prediction model.

3. The method according to claim 1 or 2, further comprising the following step: Second-order or higher-order cross features are calculated based on the cross feature interactions between the metadata. and The second-order or higher-order cross features are input into the prediction model; The generation of the skin lesion score is also based on the analysis of the second-order or higher-order cross features by the prediction model.

4. The method according to any one of the preceding claims, wherein, The data set (A) also includes reproductive status, risk factors and / or indications regarding at least one affected area of ​​the companion animal's body surface, the presence of a history of acute moist dermatitis, urticaria or angioedema, the presence of corticosteroid-induced pruritus, excessive hair loss, scaling or dryness, gastrointestinal symptoms, indications regarding whether symptoms worsen after walking in grass, or a history of chronic and / or recurrent skin diseases or otitis.

5. The method according to any one of the preceding claims further includes the step of providing a pathological profile of the companion animal based on the plurality of metadata, wherein the pathological profile includes an information record that includes the plurality of metadata of the companion animal.

6. The method according to any one of the preceding claims, wherein, The prediction model includes at least a supervised classifier machine learning model.

7. The method according to any one of the preceding claims further includes the step of providing the skin lesion score related to the companion animal to a user interface.

8. The method according to any one of the preceding claims further includes the following step: Determine the corresponding size associated with each lesion in the companion animal. The corresponding dimensions associated with each lesion were normalized using the body size of the companion animal, and The severity of the atopic dermatitis is assessed based on the skin lesion score and the corresponding normalized size associated with each lesion.

9. A method for training an assessment module to learn characteristics of atopic dermatitis in companion animals, using at least a plurality of previously acquired companion animal-related metadata selected in a data set (A), said data set (A) comprising at least animal data and skin region data corresponding to affected or unaffected skin regions, said animal data comprising at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age of onset, (viii) physical condition, (ix) health status, (x) lifestyle, (xi) living conditions The method includes: (xii) environmental factors, (xii) fur information, (xiii) activity level, (xiv) biological values ​​of biological samples, (xv) age of onset, and (xvi) history of dermatitis, and the skin region includes at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral thoracolumbar region, (iv) perineal / genital region, (v) ventral cervical region, (vi) auricular region, (vii) periorbital region, (viii) perioral region, (ix) flexor elbow region, (x) forefoot region, (xi) flexor tarsal joint region, (xii) hindfoot region, and (xiii) nonspecific atopic region. - Extract at least one feature from each previously acquired metadata. - Associate at least one of the aforementioned features with the animal's skin condition, and - Train the evaluation module to learn the association.

10. The method according to the preceding claim, wherein, The evaluation module includes at least one supervised classifier machine learning model, and the method includes the step of updating the weights of the supervised classifier machine learning model based on the association between the at least one feature and the animal skin state.

11. An apparatus for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, the apparatus comprising an assessment module pre-trained to learn characteristics indicative of atopic dermatitis in the companion animal based at least on a plurality of previously acquired metadata related to the companion animal selected in a dataset (A), the dataset (A) comprising at least animal data and skin region data corresponding to affected or unaffected skin regions, the animal data comprising at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age of onset, (viii) physical condition, (ix) health status, (x) lifestyle, (xi) living environment, (xii) fur information, (xiii) activity level, (xiv) biological values ​​of biological samples, (xv) age of onset and (xvi) history of dermatitis, and skin region data including at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral thoracolumbar region, (iv) perineal / genital region, (v) ventral cervical region, (vi) auricular region, (vii) perioral region, (viii) perioral region, (ix) flexor elbow region, (x) forefoot region, (xi) flexor tarsal joint region, (xii) hindfoot region and (xiii) nonspecific atopic region, a trained assessment module is configured to run on the companion animal data and generate a skin lesion score indicating at least one skin condition of the companion animal, wherein the companion animal data includes multiple metadata related to the companion animal, including two or more animal data and two or more skin region data listed in the data set (A).

12. The device according to the preceding claim, comprising an acquisition module for acquiring data on the companion animal.

13. A computer program product for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, comprising an assessment module pre-trained to learn characteristics indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected in a dataset (A), said dataset (A) comprising at least animal data and skin region data corresponding to affected or unaffected skin areas, said animal data comprising at least two or more of the following: (i) breed, (ii) species, (iii) sex, (iv) weight, (v) sterilization or castration status, (vi) age, (vii) age of onset, (viii) physical condition, (ix) health status, (x) lifestyle, (xi) living environment. (xii) fur information, (xiii) activity level, (xiv) biological values ​​of the biological sample, (xv) age of onset, and (xvi) history of dermatitis, and the skin region data includes at least two or more of the following: (i) groin region, (ii) axillary region, (iii) ventral chest region, (iv) perineal / genital region, (v) ventral neck region, (vi) auricular region, (vii) periorbital region, (viii) perioral region, (ix) flexor elbow region, (x) forefoot region, (xi) flexor tarsal joint region, (xii) hindfoot region, and (xiii) nonspecific atopic region, the computer program product comprising a carrier and instructions stored on the carrier that can be read by a processor, the instructions being configured to: a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Encode the metadata into a metadata vector. c) Run the trained evaluation module on the metadata vector, wherein the run includes inputting the metadata vector into the prediction model. d) Based on the analysis of the metadata vector by the prediction model associated with the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal, and e) Assess whether the companion animal has atopic dermatitis based on the skin lesion score, which indicates at least one skin condition of the companion animal.

14. A method for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, using an assessment module pre-trained to learn features indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected from a dataset (A) that includes at least animal data and skin region data, the method comprising at least the following steps: a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Run the trained evaluation module on the metadata received in step a), and c) Based on the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal.

15. A method for training an evaluation module to learn characteristics indicative of atopic dermatitis in a companion animal, using at least a plurality of previously acquired metadata related to the companion animal selected from a data set (A) that includes at least animal data and skin region data, the method comprising: - Extract at least one feature from each previously acquired metadata. - Associate at least one of the aforementioned features with the animal's skin condition, and - Train the evaluation module to learn the association.

16. An apparatus for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, the apparatus comprising an assessment module pre-trained to learn features indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected from a data set (A) that includes at least animal data and skin region data, the trained assessment module being configured to run on the companion animal data, wherein the companion animal data includes multiple metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in the data set (A), and to generate a skin lesion score indicative of at least one skin condition in the companion animal.

17. A computer program product for determining a skin lesion score in a companion animal suspected of having atopic dermatitis, comprising an evaluation module pre-trained to learn features indicative of atopic dermatitis in the companion animal based at least on multiple previously acquired metadata related to the companion animal selected from a dataset (A) including at least animal data and skin region data, the computer program product comprising a carrier and instructions stored on the carrier that can be read by a processor, the instructions being configured to: a) Receive metadata related to the companion animal, including two or more animal data items and two or more skin region data items listed in data group (A), b) Run the trained evaluation module on the metadata received in step a), and c) Based on the assessment module, generate a skin lesion score indicating at least one skin condition of the companion animal.

Images