Abstract
Canine osteoarthritis (OA) is a chronic, progressive disease that impacts mobility and welfare, often with subtle clinical signs that fluctuate over time. This exploratory case series evaluated the potential of a wearable sensor system (Maven Pet AI System) to detect real-time deviations in activity and rest patterns in dogs with OA under home-based conditions. Five client-owned dogs were monitored over periods ranging from 56 to 126 days, generating longitudinal data on activity and rest patterns. Nine clinically relevant events were identified: seven OA-related flare-ups and two non-orthopedic health issues. In eight of these events, deviations in activity profiles were temporally aligned with symptom onset, therapeutic response, or recovery. Statistically significant changes were observed in six out of nine events, particularly in the Active and Excited categories, while visual trend analysis revealed clinically relevant deviations even in the absence of statistical significance. In one case, decreased activity preceded owner recognition, suggesting potential for early detection. Sensor data also contextualized episodes of overexertion and non-orthopedic conditions, such as pruritus and gastroenteritis. Owner and clinician feedback indicated high usability and perceived clinical value. Despite the small sample, these findings suggest that continuous sensor-based monitoring may complement conventional evaluations and support earlier, more individualized OA management in real-world settings. Further studies are needed to validate and expand these preliminary observations.