NephroSense: Developing a Chronic Kidney Disease Screening Model for AI Breath Biomarker Testing

Overview

Proactive disease screening is vital for preventive care, yet common biomarker tests are invasive, wasteful, and indefinitely slow. Conversely, NephroSense proposes generalized breath analyses as a viable alternative to current methods, developing a proof-of-concept model for the detection of chronic kidney disease (CKD) through exhaled ammonia levels. An electrochemical sensor, which reacts with ammonia to output proportional voltage increases, was implemented into a user-friendly device; its optimum reaction times were tested and integrated into the interface for standardized ammonia measurements. Meanwhile, sensor readings were calibrated to user ammonia concentrations, which were then input into a newly-formulated logistic regression model trained with data correlating ammonia and CKD presence. If the predicted disease risk was elevated, the results were further analyzed by a custom-trained neural network for CKD stage evaluations. The device met the engineering objectives, as its assessment accuracies were over 99% without compensating for speed, user comfort, or environmental sustainability.

Awards (2)

• Bronze Medal
• Selected for CWSF 2026

Competition history

• CWSF 2026 Disease & Illness

Related projects

Closest projects by meaning, across every fair and year in the corpus.