AI-driven Telemedicine for Small Animal Veterinary Care Optimization

Thesis Abstract

The rapid expansion of digital health technologies has transformed veterinary service delivery, yet small animal care faces persistent access barriers, delays in diagnosis, and fragmented data transfer between clinics and owners. This study addresses the problem of limited timeliness and quality of care in small animal practice by evaluating an AI-driven telemedicine platform designed to triage cases, support remote diagnostics, and enhance follow-up management, thereby optimizing care pathways and clinical decision-making. The aim is to determine the effectiveness of AI-enabled telemedicine in reducing wait times, improving diagnostic accuracy, and enhancing owner satisfaction for common small animal conditions. The specific objectives are (1) to assess the platform’s impact on triage accuracy and recommended management plans compared with standard teleconsultation; (2) to evaluate changes in consultation turnaround times and patient follow-up rates; (3) to analyze owner satisfaction, perceived ease of use, and adherence to remote care recommendations; (4) to identify AI model performance across species (canine, feline) and presenting problems; and (5) to examine data governance, ethical considerations, and practitioner acceptance using a theoretical lens anchored in Technology Acceptance Model (TAM) and Diffusion of Innovations (DOI). The study adopts a mixed-methods design combining retrospective and prospective data over 18 months from 12 veterinary clinics that implement the telemedicine platform across a metropolitan area. The population includes licensed veterinarians, veterinary technicians, and client owners engaged in small animal care. A stratified random sample of 360 client cases (120 canine, 120 feline, 60; mixed presentations) will be analyzed for diagnostic concordance and triage outcomes, while 200 clients will participate in post-consultation surveys and 25 veterinarians in semi-structured interviews. Data collection instruments comprise platform-generated clinical decision support logs, standardized triage checklists, validated owner satisfaction surveys (CSQ-8 adapted for veterinary telemedicine), and interview guides. The AI component comprises convolutional neural networks for image-based inputs (skin, ear, ophthalmic), natural language processing of client-reported histories, and a rule-based clinical reasoning layer integrated with veterinary guidelines. Validity and reliability will be ensured through cross-validation of AI predictions with in-person diagnoses, inter-rater reliability checks on triage categorization, and pilot testing of survey instruments. Quantitative data will be analyzed using descriptive statistics, chi-square tests for categorical outcomes, t-tests or Mann-Whitney U tests for continuous outcomes, and multivariable logistic regression to identify predictors of diagnostic concordance and successful remote management. Time-to-treatment metrics will be analyzed via Kaplan-Meier estimates and Cox proportional hazards models to compare remote versus in-person pathways. AI performance will be evaluated with accuracy, sensitivity, specificity, precision, and F1 scores, with subgroup analyses by species and presenting problem. Qualitative data from veterinarian interviews and owner feedback will be analyzed using thematic analysis guided by Braun and Clarke, with coding triangulation to ensure credibility. The study is anchored in TAM to interpret technology acceptance and DOI to explain adoption dynamics among clinics and clients, with supplementary reference to the Technology-Organization-Environment (TOE) framework to contextualize institutional factors. Expected findings include (i) higher triage accuracy and conformance with evidence-based guidelines in AI-assisted cases, (ii) reduced mean consultation turnaround time by 28–40%, and (iii) improved owner satisfaction and adherence to remote care plans, without compromising patient safety. The contribution to knowledge lies in demonstrating how integrated AI-driven telemedicine can harmonize remote assessment with in-clinic standards, establish scalable care pathways for common conditions (dermatology, dermatologic symptoms, otic disorders, minor orthopedic concerns), and provide a robust evidence base for implementing AI-enabled veterinary telemedicine in diverse practice settings. The study will also illuminate ethical and governance considerations, including data privacy, consent, and clinician accountability within AI-supported decision-making. The main conclusion anticipated is that AI-assisted telemedicine enhances clinical efficiency, diagnostic reliability, and owner engagement while maintaining safety and professional oversight. Recommendations include developing standardized remote examination protocols, investing in clinician training for AI interpretation, expanding data-sharing agreements to support continuous model improvement, and creating policy guidelines for remote prescribing and follow-up scheduling to sustain high-quality small animal care in digital ecosystems.