Abstract
Urinary Tract Infections (UTIs) are a significant health concern, particularly among infants, elderly individuals, and bedridden patients, where delayed detection can lead to severe complications. This paper presents a colorimetry-based smart diaper system for the early, noninvasive screening of UTIs using Internet of Medical Things (IoMT) principles. The system integrates TCS34725 color sensors with urinalysis test strips to capture RGB data corresponding to key biomarkers such as leukocytes, nitrites, and pH. An ESP32 microcontroller processes the data using a Euclidean distance-based classification algorithm, and the results are transmitted to a cloud-based backend for periodic monitoring and alert generation. The system successfully classified 14 out of 15 simulated test cases, indicating the feasibility of the proposed screening approach under controlled testing conditions. The results demonstrate the feasibility of automated, periodic screening with minimal human intervention. Compared to conventional urinalysis methods, the proposed system offers advantages in automation, accessibility, and scalability. This work establishes a foundation for low-cost, wearable screening systems and highlights the potential for further enhancement through machine learning and clinical validation.