Cutting-edge potential of Artificial Intelligence: A paradigm shift in canine clinical practice

Introduction

Artificial intelligence (AI) is a branch of computing systems that has broadly converged from various scientific and technological fields, including data science, applied mathematics, psychometrics, statistical genetics, genomics, cybernetics, econometrics and computer engineering. AI was first defined in 1955 by John McCarthy, since then the discipline has relentlessly expanded its application in wide spectrum of technologies, including automobile, financial, e-commerce, cyber security, manufacturing, education and health care industries [13] (Figure 1A). Veterinary medicine evolves from a diverse field encompassing vast knowledge of animal health and disease research, zoonotic disease monitoring, epidemiology and disease surveillance, animal production, and small and large animal clinical practice [4][5][6] . Therefore, the convergence of these two discrete disciplines warrants remarkable technological possibilities to influence each other and advance the current landscape of clinical knowledge in animal health care industries. The global animal health care market employing AI technologies is expected to grow from 1.41 billion in 2024 to 8.23 billion in 2034 with an incremental compound annual growth rate (CAGR) of 19.3% in the next 10 years [5] . The current understanding of AI technology has broken our pervasive myths that a computer can only performs programmable solution to a problem as instructed by human inputs. However, with the qualitative and quantitative data inputs, it can overcome diagnostic challenges, automate clinical workflow and support rapid decision-making in a cost-effective manner.

The discipline of AI is broadly classified into three fundamental genre of computing systems: Natural language processing (NLP), computer vision (CV), and machine learning (ML) [7][8][9] (Figure 1B). The NLP tasks attempts to recognize relevant clinical information from unstructured documents (referred as SOAP notes creation; Subjective, Objective, Assessment Plan) of veterinary practitioner's notes or electronic health report (biochemical, radiological, or pathological), thereby ensures higher degree of accuracy and consistency in medical record keeping simultaneously transforming time management and efficiency in animal care [10], [11], [9] . In clinical practice NLP tasks have numerous potential applications that could leverage from two critical foundational components: The Unified Medical Language System (UMLS) that is used to minimize the ambiguity in the biomedical language text, and Semantic Medline Database (SemMedDB)-a semantic prediction extracted from PubMed citations revealing subject-predicate-object relationships [12][13][14] . The second subfield of AI is the computer vision that serves to analyse large number of relevant biomedical images using deep learning (DL)-based connecting nodes of neural networks [Artificial Neural Networks (ANN); Convoluted Neural Networks (CNN)] architecture that automatically extract complex features from raw data. DL utilizes ML methodologies for decision-making and is extensively recommended for predictive analysis in large imaging data set used in veterinary field, including clinical practice in dogs [15][16][17] . Application of DL through ANN and several CV algorithms have exhibited superior accuracy and sensitivity in veterinary diagnostic imaging [18][19][20] . The overgrowing availability of commercial CNN-based trained models along with lack of appropriate number of certified clinicians in this sub-speciality to determine and interpret 'blind spot' in a lesion from large number of radiograph, CT or MRI images are currently the driving force for adopting this cutting-edge technology in veterinary clinics and research [21][22][23] . ML is another subdivision of AI that relies on manual feature extraction of complex raw data and builds on four main models of supervised learning, semi-supervised learning, unsupervised learning and reinforcement learning. The supervised learning model is used extensively in the medical field for training larger set of labelled data with prior known features associated to the task or output. This is often applied in classification and regression analysis for disease diagnosis in dogs as well as to predict efficacy of drugs (ADMET). Whereas, unsupervised models of ML algorithm are employed to predict and validate the outcomes of unlabelled large datasets, based on unknown features and defined criteria for clustering and is potentially applicable to discovery of disease subtypes and drug targets. Growing evidence of AI applications in the healthcare industry suggests that implementation of ML algorithms and programmes have the potential to minimize cognitive biases, automation of responses, recognition of specific patterns (for example to identify and distinguish specific pattern from larger scale variable inputs that can be employed with high-speed and precision by ML algorithm systems), generate superior qualitative inferences, and surpassing human capability of decision-making [24,25] (Table 1). These phenomenal features of AI/ML have set forth an unprecedented trend in global healthcare market, including pet care and animal industry. In this review we focused on available current literature pertinent to canine disease diagnosis, image analysis and discuss the application of AI in disease prediction and personalized treatment. We also discussed the potential risk factors, ethical and societal issues associated with implementation of AI/ ML models in pet care industry. 'Ethics declaration: not applicable.

Benefits of AI in disease diagnosis and investigations

Increased use of AI tools to amplify the impact in canine disease diagnosis is an evolving area amongst veterinary clinicians and pet owners. AI systems not only improve the diagnostic accuracy of large complex dataset harnessed from electronic health records, imaging results, and biomarker profiling, but also analyze the data with unprecedented accuracy and precision. Processing and clinical validity of these dataset in real-time with enhanced efficiency to identify patterns are critical for early prediction and disease diagnosis.

Influence of AI-powered tools in therapeutic management

The pet wearable market is forecasted to grow at a striking rate of > 15% CAGR between 2024 and 2034 with an enhance capability to monitor real-time health-associated biometrics and numerous vital sign data points (Table 2). Proposition has been made to commercially develop AI / ML-based DNN algorithms as cost-effective wearable pet collars for optimizing real-time monitoring and classification of disease severity in osteoarthritis (OA) (DAC. Digital Inc., Poland). These essential AI-driven gadgets are promising tool to improve the evidence-based diagnostic accuracy and early interventions of a particular diseased behavior, which are often difficult to interpret, biased and missed in a shorter consultation time at the clinics. The efficiency of advance computer system and a high-frequency multidimensional analyzer have been evaluated for pruritic behavior of scratching and headshaking from 361 dogs with > 99% overall accuracy [26] . Continuous monitoring of intracranial electroencephalographic (iEEG) data from epileptic seizure in dogs has been evaluated using ML-based training method, which provides invaluable forecasting of the disease behavior [27][28][29] . Predictive epidemiological risk with clinical prognostic outcomes of canine mammary gland tumors have been reported in a multicenter study using supervised ML tools with a significant level of diagnostic accuracy [30] . The power of data-driven insights lies in AI's ability to aggregate and analyze large datasets, uncovering trends in disease prevalence and treatment efficacy. Studies like SYMPLIFY demonstrates how AI can identify high-risk cohorts for targeted cancer screening, a strategy applicable to canine populations [31] . Moreover, AI fosters standardization of protocols by reducing inter-clinician variability in cancer detection or behavioral disorder investigations. In particular, objective assessment of canine behavior employing ML algorithms has been reported to clinically validate pain-induced stress, anxiety, or neurological dysfunction using video images or wearable sensors, thereby addressing subjective interpretation over operator-based limitations with improved patient outcome [32, 33] . Collectively, these emerging applications of AIpowered algorithms foster an immense capability to lead early disease detections with précised accuracy and optimize the therapeutic interventions in-clinic.

Applications of AI-enabled automation in routine clinical task

The scope and applications of AI tools further falls under ubiquitous workflow of routine blood counts, blood smear examination, histopathological slide evaluation and assessments of parasitic disease burdening the clinics. Enhanced efficiency is achieved through automation of routine tasks and interpretation of imaging studies or lab results. For example, VETSCAN Imagyst system can be employed for automated fecal egg and oocyst counting in an unbiased manner, which minimizes different observerintended traditional limitations without further requirements of specialized training [34][35][36] . Moreover, employing MLbased tools with logistical regression and tree algorithms coupled to blood smear hematology analyzer (ADVIA) data, superior and rapid screening of Babesia merozoites in dogs could be achieved with 95.7% specificity and 100% sensitivity [37] . Utilizing CNN algorithms and CV systems, rapid detection and classification of skin lesions have been used as a tool for feature extraction of images [38] . Moreover, clinical presentations of chronic, non-infectious type of inflammatory and neoplastic diseases are often challenging because of worse prognosis and shortened survival time, thereby requires early diagnosis for better patient outcome. Based on clinic pathological data from routine blood counts and serum chemistry panels, ML algorithms have been robustly applied for differential diagnosis of chronic hypoadrenocortism (CHA) conditions in dogs from closely related other disorders, such as glucocorticoid deficiency, primary GI tract and chronic kidney diseases (CKD) or hepatic insufficiency [39] . Early diagnosis of CKD in canines using routine clinical data from electronic health records (EHR) have been developed and validated employing DL strategies and recurrent neural network (RNN) algorithms with 91.4% sensitivity and 97.2% specificity [40] . Additionally, DL-based network architecture has been extended with enhanced diagnostic capabilities for ultrasonography pathological grading of advanced CKD stages and sustained remissions. Therefore, careful and pertinent implementation of AI algorithms in canine clinical practice has substantial possibilities in advancing disease detection and assessment; with a better understanding of a patient predisposition to certain disease condition and personalizes therapy for better outcomes.

AI in disease prediction

Faster and accurate diagnoses are hallmarks of AI application, simultaneously processing the real-time data with rapid results. Clinical prediction engages AI with eight domains of medical practices, including diagnosis, assessment of risks, response to treatment, disease progression, risks associated with recurrence, and complication, prognosis of the disease, and mortality prediction [41] . AI enables personalized treatment plans by tailoring therapies and has potential to accurately predict disease remissions and prognosis. By understanding an individual patient's disease trajectory, early intervention and treatments can be better tailored to their specific needs for improving efficacy and minimising adverse effects. Furthermore, AI-powered wearable sensors have potential to refine the treatment plans by monitoring outcome responses in real-time, which allows scalable adjustments of drug dosages and complete protocol remissions.

Influence of AI in canine oncology and personalized treatment

Based on known therapeutic outcomes from the standardized multi-agent chemotherapy as first-line of treatment (combination of Cyclophosphamide, Hydroxyduanomycin, Oncovin (Vincristine sulfate), and Prednisone-commonly referred as CHOP therapy), AI-based algorithms have been successfully employed for personalized prediction profiling of canine lymphoma and leukemia cases [42,43] . A commercially available service from Imprimed Company provides AI-based therapeutic efficacy and accurate remission prediction of 13 anticancer drugs on live canine cancer cells [42,44] . Multi-cancer early detection (MCED) tests adapted from human patients have been clinically applied and validated for dogs by liquid biopsy company PetDx to detect cancer at an early stage and predict the tumor aggressiveness [45][46][47] . Interestingly, the applications of AI-powered cytology methods demonstrate enhanced capacity to discriminate between fine-needle aspiration and conventional histopathology techniques for diagnosis of cutaneous masses [48,49] .

Potential application of AI in point-of-care facilities

Based on thermal properties, perfusion rate and metabolic activities of tumor tissues, ML-based decision-making algorithms of thermal imaging signals have been employed to distinguish between cutaneous and subcutaneous tumors noninvasively [50] . Therefore, development of ML program enables high degree of certainty in detecting subtle patterns to classify malignancies by thermal imaging or volatile organic compounds (VOCs) assessmentand minimizes the risk of false-positive cases [50], [51] . Moreover, ML-based application has been suggested for discrimination between diseased and sustainable remissions based on osteosarcomaderived exosomal gene signatures as prognostic marker for canines [52] . This approach enables refined applications of AI tools with muti-omics data in canine oncology research and disease intervention. Collectively, such tools outperform the traditional diagnostic and risk-predictive methods minimizing false positive cases, thereby unleashing the power of AI in early cancer detection and persistent remissions towards canine health improvement. Continuing interdisciplinary research and collaboration between AI experts and clinicians will entails to fine-tune and amplify the full potential of AI applications in personalized canine treatments.

Influence of AI in canine diagnostic imaging

Transformative impact of AI-powered algorithms have improved our understanding to extract, quantify and analyze large amount of medical images evolving from Radiomics research studies and apply them for disease phenotyping and diagnosis. The clinical use of digitalized data curated from radiographs, CT scans, MRI, PET, and ultrasound images in veterinary diagnostic have significant promise to further analyze, detect artifacts, apply segmentation, and assign the associated image reports for disease prediction and accurate stage-classification (Table 3). For example, several canine radiograph images have been analyzed so far by employing CNN-based algorithm and DL method extraction features, such as-age determination from bones, components of joints and abnormality identification, fracture categorization followed by comparing the results with expert radiologist for validation [53][54][55] . These novel algorithms reveal high accuracy, sensitivity, and specificity for all the assigned tasks and performs > 80% match with radiologistdetermined reports and classification. The commercially available AI-based software (Vetology Innovations Inc, San Diego, CA, USA) have successfully been employed in determining normal and abnormal canine thoracic and extrathoracic radiograph images with higher positive predictive values [56][57][58] . Implementing CNN-based AI tools, canine corneal ulcerative images have been optimized, validated, and categorized based on severity of the disease with superior accuracy and sensitivity [59] .

Application of AI on digital images in-clinic

Furthermore, to pin-point the diagnostic accuracy and binary classification of degenerative (vs. non-degenerative) liver diseases in dogs, ML tools have been developed from pretrained deep neural networks (DNN) using ultrasound images. The study reveals superior resultant DNN data accuracy over conventional cytological and biochemical determinants with 100% sensitivity and > 82% specificity (AUC 0.91) for detection of degenerative canine liver diseases [60] . Similar ML application and DL methodology have been opted for better signal analysis and enhancing diagnostic capabilities of computer tomography (CT) scan images [22,61,62] . Using ML-based feature engineering of CT image modalities, AI can narrow the decision making window of canine diagnostic imaging in different disease classification, lesion detection, and tumor segmentation [63], [64], (65,66). Implementation of sophisticated ML pipelines on MRI images exhibited précised diagnostic accuracy indistinguishing different stages ofcanine inflammatory or neoplastic brain diseases [67], [68] . Using texture analysis feature of MRI images non-infectious inflammatory meningioencephalitis could be clinically distinguished from the overlapping images of canine glial cell neoplasia. Although the subjective radiological differentiation seems arduous, random forest algorithm enabled featured engineering application provides plausible solution to distinguish between gliomas grades and inflammatory subtypes (Granulomatous meningioencephalitis; GME, necrotizing meningioencephalitis; NME, and necrotizing leukoencephalitis; NLE) [67] . Difficult to diagnose MRI images fromcanine skull and cranial cervical vertebrae disorder (Syringomyelia and pain associated Chiari-like malformation) in Cavalier King Charles Spaniel (CKCS) breed have been successfully adopted for neuromorphological characterization using ML tools [69] . In the latest study, CNN-based algorithms have been shown to distinguish between intervertebral disc protrusions from extraction (IVDP vs. IVDE) with higher degree of accuracy (> 95% specificity, and > 90% sensitivity) in canine spinal cord disease detection [70] . Moreover, DL-based novel algorithms have been developed to minimize the timeframe with increased quality of compressed-sensing associated MRI image acquisition and reconstruction [71,72] . Collectively, these findings demonstrate that integration of AI tools have unmet potential to eliminate observer-based decision making errors, there by facilitating the diagnostic imaging accuracy and treatment outcomes in canine clinical practice.

Ethical, legal and societal implications of AI in canine practice

The integration of artificial intelligence (AI) and big data analytics into veterinary medicine heralds a transformative era in animal healthcare, marked by innovations in diagnostics, personalized treatment, and operational efficiency. However, these advancements are inextricably linked to ethical, legal, and societal challenges that demand interdisciplinary collaboration to ensure equitable and responsible adoption by the professionals and pet owners.

Resource inequity and access barriers

The promise of AI-driven veterinary care is tempered by disparities in overall usage of systemic resources, particularly in rural and underserved regions. Urbanized clinical practice increasingly deploys AI tools, including DL algorithms and thermal imaging systems for cancer detection and enhanced diagnostic precision [51] . In contrast, canine practice in rural regions often lacks the financial and infrastructural capacity to adopt these technologies, including high-speed internet and advanced computing hardware [5] . For example, there are > 40% inaccessibility to AI-based model adoption and computer utilization for appendicular bone neoplasia detection and screening in lowincome regions. Although technological access drives better animal care and quality, the demographical disparities in clinical practice entrench risks of two-tiered healthcare system. Addressing the division demands policy interventions, such as subsidized AI platforms, infrastructure grants, and public-private partnerships to democratize access for all [31] .

Ethical and legal challenges in big data Governance

The ethical integration of AI and big data in veterinary medicine hinges on equitable access to transparent governance and preservation of clinical judgment. Policymakers must prioritize infrastructure investments for underserved regions, enforce algorithmic accountability, and modernize data protection laws. The proliferation of big data in veterinary medicine raises unresolved ethical dilemmas, particularly regarding privacy, ownership, and algorithmic bias. In accordance with the current medical health care sector, the US Food and Drug Administration (FDA), the Medicines and Health care products Regulatory Agency (MHRA) of United Kingdom and Health Canada have jointly identified 10 guiding principles for development of Good Machine Learning Practice (GMLP) [73, 74] . These guiding principles includes, inclusion of multidisciplinary expertise throughout the AI/ML methods and systems, implementing good software engineering and security practices, applications of software in clinical scenario should represents the data set of the intended population. In addition, the datasets in veterinary medicine initiate various sensitive information from genomic profiles, treatment histories, and owner demographics, thereby requires regulatory framework and compliance of General Data Protection Regulation (GDPR) and Health Insurance Portability and Accountability Act (HIPAA) to address specific issues [75] . Moreover, pet owners frequently lack awareness of how their data is monetized and shared with third parties, underscoring the need for informed and transparent consent management system [76] . Regulatory mandates for algorithmic audits and equity assessments are critical to ensure AI outputs align with diverse clinical practice [77] .

AI-driven decision-making and autonomy in clinics

The ascendancy of AI in clinical workflows challenges clinical autonomy and trust. While AI reduces diagnostic errors by 15-20% in thoracic lesion evaluations, there are chances of over-reliance on algorithms that compromises clinical judgment and animal welfare [78] . In practice, expertise in contextualizing patient history, behavior, and owner preferences remains irreplaceable, particularly in complex cases where empathy over nuanced decisionmaking is required [79] . Considerable amount of ambiguity exist amongst stakeholders, clinicians and academic institutions over the usage and interpretation of black box (or heuristic algorithm) ML models that affects safety, ethics, and regulations which compounded distrust in healthcare applications [80,81] . To reconcile these ambiguities, AI must function as an augmentative tool with veterinarians retaining their ultimate clinical authority. Concurrently, educational reforms must cultivate the digital literacy and training to empower codes of ethical practice for AI output evaluation.

Societal trust and companion relationship

The AI's pivotal role in accelerating veterinary science and animal health care industry are forthcoming. The goal of model design should be tailored on available data or training data sets that are independent of test sets and provides transparent clinically relevant information to the users. However, the full spectrum of AI/ML adoptability and its clinical application in veterinary practice possess higher risk of algorithm transparency, data privacy and sharing with ethical, legal and societal complications and acceptance. These consistent developments and challenges of cuttingedge technologies will continue to shape the future of animal care industry in twenty first century. The societal implications of AI adoption extend beyond clinical settings, impacting public trust and the human-animal relationship. Pet owners increasingly expect transparency in AI-driven diagnoses; however, the existing opaque systems risk eroding confidence in animal care and clinical practice [82] . There are possibilities of misdiagnoses attributed to AI tools without clear explanations, which deter owners from pursuing critical treatments [80] . Therefore, preserving trust is entitled and requires explainable AI (XAI) frameworks to demystify algorithmic decisions ensuring informed consent policies for data security [75] .

Collaboration and future perspectives

By making complete and intelligent use of the technology to subjugate and transform vast clinical information-petpractitioners, owners, and researchers along with national and international authorities should be tasked with delivering a quantum leap forward for development and use of this new knowledge in animal welfare. Simultaneously, the veterinary profession must champion the irreplaceable role of human expertise, ensuring technology amplifies-rather than supplants-the human-animal bond. By fostering collaboration among clinicians, technologists, and policymakers, the field can harness AI's full potential while safeguarding ethical integrity and societal trust.

Conclusion

The transformative impact of AI in the field of veterinary medicine surpasses the limitation of clinical documentation, record keeping, rapid point-of-care testing, differential disease diagnosis, and tailoring personalized medical plans, thereby promise a new dimension of technological integration that encompasses every facet of canine practice (Figure 2). As AI technology rapidly evolves, AI-assisted algorithms and novel tools will be invaluable to assist veterinarians for a better understanding of the disease process, providing much better service and more promising outcomes. The unification of data-driven AI application across the diverse fields of veterinary clinical practice reflects an innovative shift towards greater operational efficiency revitalizing the work force culture with accurate diagnostic tools and enhanced animal welfare. However, despite this emerging inquisitiveness and research efforts, successful implementation and adoption of AI tools in pet clinics and hospitals warrants interdisciplinary collaboration, clinical validity, large infrastructure to manage and employ big data sets and deployment of trained personnel and technicians.

Conflict of interest

All authors have read and approved of the manuscript and do not have any conflicts of interest to declare. The authors declare that the review materials in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author contributions

All the authors contributed to the study conception and design. D.C., N.M, S.N., R.B., and A.M. collected the literature and drafted the manuscript; A.M generated the images and supervised the writing of the manuscript. A.M. and N.S. conceived the idea and edited the manuscript to its current version.

Funding and acknowledgement

Open access funding provided by Siksha 'O' Anusandhan (Deemed to be University). We thank Siksha-O-Anusandhan University (Deemed to be University) for institutional support and advice. The content is solely the responsibility of the authors and does not necessarily represent the official views of the University.

Availability of data

Data sharing does not apply to this article as no new data were generated or analysed in this study.

12.

Split the learning data into classes and sort out hierarchical classification and regression problems

In predictive analysis and early disease detection and classification. [83]

Bayesian networks (BN)

Probability distribution with graphic model illustrating a set of arbitrary variable Used in veterinary epidemiological data for risk factor analysis. [84] K-nearest neighbors (kNN)

Classifies a point depending on known classification of other points

Used in disease classification and analysis. [85] Relevance Vector Machine (RVM)

Probability-based algorithm includes Bayesian extension and SVM Real-time prediction of variables in animal research [86] Decision Tree (DT)

Rule-based graphical model predicting end-nodes and leaves

Used in disease prediction and diagnostic analysis. [87] Unsupervised Learning Fuzzy C Means (FCM)

Partitioning-based algorithm that measure the distance between cluster center and each data point Analysis and clustering of big data in disease progression and prognosis. [88] K-Means Type of clustering algorithm Used in clustering analysis of canine behavioral research studies [87] Multilayer Perception (MLP)

Basic unit (perception) comprising multiple layers that allows backward propagation with learning algorithm Analysis and determination of various disease variables for automation. [89] Convolutional Neural Networks (CNNs)

MLP applies to a 2D matrices by a convolution operation

Used for image analysis and Computer Vision (CV) task. [16] Recurrent Neural Networks (RNNs)

Time-lapse or sequential data linked to loops between layers Automation of medical text narratives and voice recognition used in Natural Language Processing (NLP) [90] Long short-term Memory (LSTM)

Extended version of RNNs that stored information for longer time Used for improving detection, performance for different biomedical devices, images and signals. [91] Thorax Radiographs cardiogenic pulmonary edema Vetology software > 90% accuracy, sensitivity, and specificity [58] Thorax Radiographs Pleural effusion Vetology software >90% sensitivity < 85% accuracy and specificity [57] Cardio-thoracic Radiographs Left atrial enlargement DL > 80% accuracy and specificity [93] Abdominal CT scans Classification of focal lesions in spleen ML 67% overall accuracy [22] Pulmonary CT scans Thromboembolism detection SVM-based supervised ML > 94% sensitivity and specificity [63] Abdominal CT scans Heterogeneity in hepatic carcinoma ML > 73% accuracy [94] Brain MRI Inflammatory vs. neoplastic DL > 85% accuracy [67] International Journal of Advanced Biochemistry Research https://www.biochemjournal.com lesions Brain MRI Neuromorphological characterization of Chiari-like malformation DL >90%sensitivity69% specificity [69] Intracranial MRI Distinguishing meningiomas from gliomas DL 80% accuracy [95] Spinal cord MRI Spinal cord injury DL > 95% specificity [70] Abdominal Ultrasound Degenerative liver disease DNN based transfer learning method > 80% specificity and 100% sensitivity [60]

References

1. Kaul V, Enslin S, Gross SA. History of artificial intelligence in medicine. Gastrointest Endosc. 2020;92(4):807-812.
2. Haug CJ, Drazen JM. Artificial Intelligence and Machine Learning in Clinical Medicine. N Engl J Med. 2023 Mar 30;388(13):1201-1208.
3. Yasnitsky LN. Artificial Intelligence and Medicine: History, Current State, and Forecasts for the Future. Curr Hypertens Rev. 2020;16(3):210-215.
4. Lustgarten JL, Zehnder A, Shipman W, Gancher E, Webb TL. Veterinary informatics: Forging the future between veterinary medicine, human medicine, and One Health initiatives-a joint paper by the Association for Veterinary Informatics (AVI) and the CTSA One Health Alliance (COHA). JAMIA Open. 2020;3(3):306-312.
5. Appleby RB, Basran PS. Artificial intelligence in veterinary medicine. J Am Vet Med Assoc. 2022;260(8):819-824.
6. Min PK, Mito K, Kim TH. The Evolving Landscape of Artificial Intelligence Applications in Animal Health. Indian J Anim Res. 2024;58(10):1793-1798.
7. Briganti G, Le Moine O. Artificial Intelligence in Medicine: Today and Tomorrow. Front Med (Lausanne). 2020 Feb 5;7:27.
8. Basran PS, Appleby RB. The unmet potential of artificial intelligence in veterinary medicine. Am J Vet Res. 2022 Mar 30;83(5):385-392.
9. Karalis VD. The Integration of Artificial Intelligence into Clinical Practice. Appl Biosci. 2024;3(1):14-44.
10. Nie A, Zehnder A, Page RL, Zhang Y, Pineda AL, Rivas MA, et al. DeepTag: Inferring diagnoses from veterinary clinical notes. NPJ Digit Med. 2018;1:60.
11. Harrison CJ, Sidey-Gibbons CJ. Machine learning in medicine: A practical introduction to natural language processing. BMC Med Res Methodol. 2021;21(1):158.
12. McGreevy P, Thomson P, Dhand NK, Raubenheimer D, Masters S, Mansfield CS, et al. VetCompass Australia: A National Big Data Collection System for Veterinary Science. Animals (Basel). 2017;7(10):74.
13. Summers JF, O'Neill DG, Church D, Collins L, Sargan D, Brodbelt DC. Health-related welfare prioritisation of canine disorders using electronic health records in primary care practice in the UK. BMC Vet Res. 2019;15(1):163.
14. Kilicoglu H, Rosemblat G, Fiszman M, Shin D. Broadcoverage biomedical relation extraction with SemRep. BMC Bioinformatics. 2020;21(1):188.
15. Tahghighi P, Appleby RB, Norena N, Ukwatta E, Komeili A. Machine learning can appropriately classify the collimation of ventrodorsal and dorsoventral thoracic radiographic images of dogs and cats. Am J Vet Res. 2023;84(3):ajvr.22.10.0180.
16. Smith A, Carroll PW, Aravamuthan S, Walleser E, Lin H, Anklam K, et al. Computer vision model for the detection of canine pododermatitis and neoplasia of the paw. Vet Dermatol. 2024;35(2):138-147.
17. Hennessey E, DiFazio M, Hennessey R, Cassel N. Artificial intelligence in veterinary diagnostic imaging: A literature review. Vet Radiol Ultrasound. 2022;63(S1):851-870.
18. Banzato T, Wodzinski M, Burti S, Vettore E, Muller H, Zotti A. An AI-based algorithm for the automatic evaluation of image quality in canine thoracic radiographs. Sci Rep. 2023;13(1):17024.
19. Bouhali O, Bensmail H, Sheharyar A, David F, Johnson JP. A Review of Radiomics and Artificial Intelligence and Their Application in Veterinary Diagnostic Imaging. Vet Sci. 2022;9(8):410.
20. Hespel AM, Zhang Y, Basran PS. Artificial intelligence 101 for veterinary diagnostic imaging. Vet Radiol Ultrasound. 2022;63(S1):833-850.
21. Kang H, Noh D, Lee SK, Choi S, Lee K. Deep learning-based reconstruction can improve canine thoracolumbar magnetic resonance image quality and reduce slice thickness. Vet Radiol Ultrasound. 2023;64(5):882-892.
22. Burti S, Zotti A, Bonsembiante F, Contiero B, Banzato T. A Machine Learning-Based Approach for Classification of Focal Splenic Lesions Based on Their CT Features. Front Vet Sci. 2022;9:856443.
23. Schmid D, Scholz VB, Kircher PR, Lautenschlaeger IE. Employing deep convolutional neural networks for segmenting the medial retropharyngeal lymph nodes in CT studies of dogs. Vet Radiol Ultrasound. 2022;63(4):437-446.
24. Vayena E, Blasimme A, Cohen IG. Machine learning in medicine: Addressing ethical challenges. PLoS Med. 2018;15(11):e1002689.
25. Warnat-Herresthal S, Schultze H, Shastry KL, Manamohan S, Mukherjee S, Garg V, et al. Swarm Learning for decentralized and confidential clinical machine learning. Nature. 2021;594(7862):265-270.
26. Griffies JD, Zutty J, Sarzen M, Soorholtz S. Wearable sensor shown to specifically quantify pruritic behaviors in dogs. BMC Vet Res. 2018;14(1):124.
27. Varatharajah Y, Iyer RK, Berry BM, Worrell GA, Brinkmann BH. Seizure Forecasting and the Preictal State in Canine Epilepsy. Int J Neural Syst. 2017;27(1):1650046.
28. Nejedly P, Kremen V, Sladky V, Nasseri M, Guragain H, Klimes P, et al. Deep-learning for seizure forecasting in canines with epilepsy. J Neural Eng. 2019;16(3):036031.
29. Nasseri M, Kremen V, Nejedly P, Kim I, Chang SY, Jo HJ, et al. Semi-supervised training data selection improves seizure forecasting in canines with epilepsy. Biomed Signal Process Control. 2020;57:101734. International Journal of Advanced Biochemistry Research https://www.biochemjournal.com
30. Burrai GP, Gabrieli A, Moccia V, Zappulli V, Porcellato I, Brachelente C, et al. A Statistical Analysis of Risk Factors and Biological Behavior in Canine Mammary Tumors: A Multicenter Study. Animals (Basel). 2020;10(9):1687.
31. Nicholson BD, Oke J, Virdee PS, Harris DA, O'Doherty C, Park JE, et al. Multi-cancer early detection test in symptomatic patients referred for cancer investigation in England and Wales (SYMPLIFY): a large-scale, observational cohort study. Lancet Oncol. 2023;24(7):733-743.
32. Ezanno P, Picault S, Beaunée G, Bailly X, Muñoz F, Duboz R, et al. Research perspectives on animal health in the era of artificial intelligence. Vet Res. 2021;52(1):40.
33. Monedeiro F, Monedeiro-Milanowski M, Zmysłowski H, De Martinis BS, Buszewski B. Evaluation of salivary VOC profile composition directed towards oral cancer and oral lesion assessment. Clin Oral Investig. 2021;25(7):4513-4523.
34. Nagamori Y, Scimeca R, Hall-Sedlak R, Blagburn B, Starkey LA, Bowman DD, et al. Multicenter evaluation of the Vetscan Imagyst system using Ocus 40 and EasyScan One scanners to detect gastrointestinal parasites in feces of dogs and cats. J Vet Diagn Invest. 2024;36(3):362-371.
35. Nagamori Y, Sedlak RH, DeRosa A, Pullins A, Cree T, Loenser M, et al. Further evaluation and validation of the VETSCAN IMAGYST in-clinic feline and canine fecal parasite detection system integrated with a deep learning algorithm. Parasit Vectors. 2021;14(1):133.
36. Nagamori Y, Hall Sedlak R, DeRosa A, Pullins A, Cree T, Loenser M, et al. Evaluation of the VETSCAN IMAGYST: An in-clinic canine and feline fecal parasite detection system integrated with a deep learning algorithm. Parasit Vectors. 2020;13(1):593.
37. Pijnacker T, Bartels R, van Leeuwen M, Teske E. Identification of parameters and formulation of a statistical and machine learning model to identify Babesia canis infections in dogs using available ADVIA hematology analyzer data. Parasit Vectors. 2022;15(1):475.
38. Upadhyay A, Singh G, Mhatre S, Nadar P. Dog Skin Diseases Detection and Identification Using Convolutional Neural Networks. SN Comput Sci. 2023;4(3):284.
39. Reagan KL, Pires J, Quach N, Gilor C. Evaluation of a machine learning tool to screen for hypoadrenocorticism in dogs presenting to a teaching hospital. J Vet Intern Med. 2022;36(6):1949-1956.
40. Kokkinos Y, Morrison JA, Bradley R, Panagiotakos T, Ogeer J, Chew D, et al. An early prediction model for canine chronic kidney disease based on routine clinical laboratory tests. Sci Rep. 2022;12(1):11130.
41. Khalifa M, Albadawy M. Artificial Intelligence for Clinical Prediction: Exploring Key Domains and Essential Functions. Comput Methods Programs Biomed Update. 2024;5:100143.
42. Callegari AJ, Tsang J, Park S, Swartzfager D, Kapoor S, Choy K, et al. Multimodal machine learning models identify chemotherapy drugs with prospective clinical efficacy in dogs with relapsed B-cell lymphoma. Front Oncol. 2024;14:1367740.
43. Koo J, Choi K, Lee P, Polley A, Pudupakam RS, Tsang J, et al. Predicting Dynamic Clinical Outcomes of the Chemotherapy for Canine Lymphoma Patients Using a Machine Learning Model. Vet Sci. 2021;8(12):312.
44. Bohannan Z, Pudupakam RS, Koo J, Horwitz H, Tsang J, Polley A, et al. Predicting likelihood of in vivo chemotherapy response in canine lymphoma using ex vivo drug sensitivity and immunophenotyping data in a machine learning model. Vet Comp Oncol. 2021;19(1):160-171.
45. Flory A, Ruiz-Perez CA, Clavere-Graciette AG, Rafalko JM, OKell AL, Flesner BK, et al. Clinical validation of a blood-based liquid biopsy test integrating cell-free DNA quantification and nextgeneration sequencing for cancer screening in dogs. J Am Vet Med Assoc. 2024;262(5):630-640.
46. Schrag D, Beer TM, McDonnell CH, Nadauld L, Dilaveri CA, Reid R, et al. Blood-based tests for multicancer early detection (PATHFINDER): A prospective cohort study. Lancet. 2023;402(10409):1251-1260.
47. Klein EA, Richards D, Cohn A, Tummala M, Lapham R, Cosgrove D, et al. Clinical validation of a targeted methylation-based multi-cancer early detection test using an independent validation set. Ann Oncol. 2021;32(9):1167-1177.
48. Simeonov RS. The accuracy of fine-needle aspiration cytology in the diagnosis of canine skin and subcutaneous masses. Comp Clin Path. 2012;21(2):163166.
49. Ghisleni G, Roccabianca P, Ceruti R, Stefanello D, Bertazzolo W, Bonfanti U, et al. Correlation between fine-needle aspiration cytology and histopathology in the evaluation of cutaneous and subcutaneous masses from dogs and cats. Vet Clin Pathol. 2006;35(1):24-30.
50. Dank G, Buber T, Polliack G, Aviram G, Rice A, Yehudayoff A, et al. A pilot study for a non-invasive system for detection of malignancy in canine subcutaneous and cutaneous masses using machine learning. Front Vet Sci. 2023;10:1187746.
51. Sung J, Loughin C, Marino D, Leyva F, Dewey C, Umbaugh S, et al. Medical infrared thermal imaging of canine appendicular bone neoplasia. BMC Vet Res. 2019;15(1):234.
52. Makielski KM, Donnelly AJ, Khammanivong A, Scott MC, Ortiz AR, Galvan DC, et al. Development of an exosomal gene signature to detect residual disease in dogs with osteosarcoma using a novel xenograft platform and machine learning. Lab Invest. 2021;101(12):1597-1607.
53. McEvoy FJ, Proschowsky HF, Müller AV, Moorman L, Bender-Koch J, Svalastoga EL, et al. Deep transfer learning can be used for the detection of hip joints in pelvis radiographs and the classification of their hip dysplasia status. Vet Radiol Ultrasound. 2021;62(3):281-288.
54. Ergun GB, Guney S. Classification of Canine Maturity and Bone Fracture Time Based on X-Ray Images of Long Bones. IEEE Access. 2021;9:119642-119652.
55. Shim H, Lee J, Choi S, Kim J, Jeong J, Cho C, et al. Deep learning-based diagnosis of stifle joint diseases in dogs. Vet Radiol Ultrasound. 2023;64(3):553-562.
56. Pomerantz LK, Solano M, Kalosa-Kenyon E. Performance of a commercially available artificial International Journal of Advanced Biochemistry Research https://www.biochemjournal.com intelligence software for the detection of confirmed pulmonary nodules and masses in canine thoracic radiography. Vet Radiol Ultrasound. 2023;64(2):331339.
57. Müller TR, Solano M, Tsunemi MH. Accuracy of artificial intelligence software for the detection of confirmed pleural effusion in thoracic radiographs in dogs. Vet Radiol Ultrasound. 2022;63(4):427-436.
58. Kim E, Fischetti AJ, Sreetharan P, Weltman JG, Fox PR. Comparison of artificial intelligence to the veterinary radiologist's diagnosis of canine cardiogenic pulmonary edema. Vet Radiol Ultrasound. 2022;63(4):420-426.
59. Kim JY, Lee HE, Choi YH, Lee SJ, Jeon JS. CNNbased diagnosis models for canine ulcerative keratitis. Sci Rep. 2019;9(1):13511.
60. Banzato T, Bonsembiante F, Aresu L, Gelain ME, Burti S, Zotti A. Use of transfer learning to detect diffuse degenerative hepatic diseases from ultrasound images in dogs: A methodological study. Vet J. 2018;233:3540.
61. Burti S, Zotti A, Bonsembiante F, Contiero B, Banzato T. Diagnostic Accuracy of Delayed Phase Post Contrast Computed Tomographic Images in the Diagnosis of Focal Liver Lesions in Dogs: 69 Cases. Front Vet Sci. 2021;8:659424.
62. Pey P, Specchi S, Rossi F, Diana A, Drudi I, Zwingenberger AL, et al. Prediction of vascular invasion using a 7-point scale computed tomography grading system in adrenal tumors in dogs. J Vet Intern Med. 2022;36(1):119-127.
63. Marschner CB, Kokla M, Amigo JM, Rozanski EA, Wiinberg B, McEvoy FJ. Texture analysis of pulmonary parenchymateous changes related to pulmonary thromboembolism in dogs-a novel approach using quantitative methods. BMC Vet Res. 2017;13(1):206.
64. Choi BK, Park S, Lee G, Chang D, Jeon S, Choi J. Can CT texture analysis parameters be used as imaging biomarkers for prediction of malignancy in canine splenic tumors? Vet Radiol Ultrasound. 2023;64(1):6474.
65. Groendahl AR, Huynh BN, Tomic O, Søvik Å, Dale E, Malinen E, et al. Automatic gross tumor segmentation of canine head and neck cancer using deep learning and cross-species transfer learning. Front Vet Sci. 2023;10:1115867.
66. Jardim SVB. The Electronic Health Record and its Contribution to Healthcare Information Systems Interoperability. Procedia Technol. 2013;9:940-948.
67. Wanamaker MW, Vernau KM, Taylor SL, Cissell DD, Abdelhafez YG, Zwingenberger AL. Classification of neoplastic and inflammatory brain disease using MRI texture analysis in 119 dogs. Vet Radiol Ultrasound. 2021;62(5):545-554.
68. Gilanie G, Bajwa UI, Waraich MM, Habib Z. Computer aided diagnosis of brain abnormalities using texture analysis of MRI images. Int J Imaging Syst Technol. 2019;29(2):165-172.
69. Spiteri M, Knowler SP, Rusbridge C, Wells K. Using machine learning to understand neuromorphological change and image-based biomarker identification in Cavalier King Charles Spaniels with Chiari-like malformation-associated pain and syringomyelia. J Vet Intern Med. 2019;33(6):2613-2624.
70. Biercher A, Meller S, Wendt J, Caspari N, SchmidtMosig J, De Decker S, et al. Using Deep Learning to Detect Spinal Cord Diseases on Thoracolumbar Magnetic Resonance Images of Dogs. Front Vet Sci. 2021;8:753683.
71. Chen Y, Schonlieb CB, Lio P, Leiner T, Dragotti PL, Wang G, et al. AI-Based Reconstruction for Fast MRI. A Systematic Review and Meta-Analysis. Proc IEEE. 2022;110(2):224-245.
72. Choi H, Lee SK, Choi H, Lee Y, Lee K. Deep learningbased reconstruction for canine brain magnetic resonance imaging could improve image quality while reducing scan time. Vet Radiol Ultrasound. 2023;64(5):871-881.
73. U.S. Food and Drug Administration. Good Machine Learning Practice for Medical Device Development: Guiding Principles. Silver Spring (MD): FDA; 2021.
74. Ebers M, Hoch VRS, Rosenkranz F, Ruschemeier H, Steinrötter B. The European Commission's Proposal for an Artificial Intelligence Act. A Critical Assessment by Members of the Robotics and AI Law Society (RAILS). J. 2021;4(4):589-603.
75. Astromskė K, Peičius E, Astromskis P. Ethical and legal challenges of informed consent applying artificial intelligence in medical diagnostic consultations. AI Soc. 2021;36(2):509-520.
76. Beauchamp TL. Informed consent: Its history, meaning, and present challenges. Camb Q Healthc Ethics. 2011;20(4):515-523.
77. Medical Device Coordination Group. MDCG 2019-11: Guidance on Qualification and Classification of Software in Regulation (EU) 2017/745 and Regulation (EU) 2017/746. Brussels: European Commission; 2019.
78. Boissady E, de La Comble A, Zhu X, Hespel AM. Artificial intelligence evaluating primary thoracic lesions has an overall lower error rate compared to veterinarians or veterinarians in conjunction with the artificial intelligence. Vet Radiol Ultrasound. 2020;61(6):619-627.
79. Coghlan S, Parker C. Harm to Nonhuman Animals from AI: A Systematic Account and Framework. Philos Technol. 2023;36(2):33.
80. Rudin C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat Mach Intell. 2019;1(5):206-215.
81. Brady AP, Neri E. Artificial Intelligence in Radiology Ethical Considerations. Diagnostics (Basel). 2020;10(4):231.
82. Geis JR, Brady AP, Wu CC, Spencer J, Ranschaert E, Jaremko JL, et al. Ethics of artificial intelligence in radiology: Summary of the joint European and North American multisociety statement. Insights Imaging. 2019;10(1):101.
83. Kumar A, Singh SK, Saxena S, Lakshmanan K, Sangaiah AK, Chauhan H, et al. Deep feature learning for histopathological image classification of canine mammary tumors and human breast cancer. Inf Sci. 2020;508:405-421.
84. Kratzer G, Lewis FI, Willi B, Meli ML, Boretti FS, Hofmann-Lehmann R, et al. Bayesian Network Modeling Applied to Feline Calicivirus Infection Among Cats in Switzerland. Front Vet Sci. 2020;7:73.
85. Chadha A, Dara R, Pearl DL, Gillis D, Rosendal T, Poljak Z. Classification of porcine reproductive and respiratory syndrome clinical impact in Ontario sow herds using machine learning approaches. Front Vet Sci. 2023;10:1120427.
86. Huang Y, Zang H, Cheng X, Wu H, Li J. Data-driven soft sensor for animal cell suspension culture process based on DRVM. Appl Soft Comput. 2019;77:502-511.
87. Amirhosseini MH, Yadav V, Serpell JA, Pettigrew P, Kain P. An artificial intelligence approach to predicting personality types in dogs. Sci Rep. 2024;14(1):4755.
88. Ferraz PFP, Ferraz GAES, Leso L, Klopčič M, Rossi G, Barbari M. Evaluation of the physical properties of bedding materials for dairy cattle using fuzzy clustering analysis. Animals (Basel). 2020;10(2):315.
89. Fidya F, Priyambadha B. Automation of gender determination in human canines using artificial intelligence. Dent J (Majalah Kedokt Gigi). 2018;50(3):121-125.
90. Venkataraman GR, Pineda AL, Bear Don't Walk OJ, Zehnder AM, Ayyar S, Page RL, et al. FasTag: Automatic text classification of unstructured medical narratives. PLoS One. 2020;15(6):e0234647.
91. Fernando T, Ghaemmaghami H, Denman S, Sridharan S, Hussain N, Fookes C. Heart Sound Segmentation Using Bidirectional LSTMs with Attention. IEEE J Biomed Health Inform. 2020;24(6):1601-1609.
92. Pérez-Espinosa H, Reyes-Meza V, Aguilar-Benitez E, Sanzón-Rosas YM. Automatic individual dog recognition based on the acoustic properties of its barks. J Intell Fuzzy Syst. 2018;34(5):3343-3353.
93. Li S, Wang Z, Visser LC, Wisner ER, Cheng H. Pilot Application of artificial intelligence for detecting left atrial enlargement on canine thoracic radiographs. Vet Radiol Ultrasound. 2020;61(6):611-618.
94. Shaker R, Wilke C, Ober C, Lawrence J. Machine learning model development for quantitative analysis of CT heterogeneity in canine hepatic masses may predict histologic malignancy. Vet Radiol Ultrasound. 2021;62(6):651-660.
95. Banzato T, Bernardini M, Cherubini GB, Zotti A. A methodological approach for deep learning to distinguish between meningiomas and gliomas on canine MR-images. BMC Vet Res. 2018;14(1):317.