Miniaturized olfactory sensor chip-based AI-wearable biometric ring for human body metabolic odor analysis

The human body emanated metabolic odor encompass biometric information used for tracking metabolic conditions of individuals. The challenge of decoding these long-unappreciated physiological signals of metabolic states impedes the development of technologies for real-time monitoring and next-generation personal health digitization. Here we introduce a monolithic AI-wearable biometric ring designed to non-invasively and reliably monitor skin volatile-organic-compounds associated with diet-related metabolism. The ring utilizes the attention-based feature fusion algorithm on a miniaturized olfactory sensor chip, incorporating three-dimensional vertical hetero-interfaces within a small footprint of 0.0081 mm². It discriminates six categories of diet-induced emanated volatile-organic-compound patterns, three physical activities, and quantifies food intake with a 0.991 determination coefficient. Through gas chromatography-mass spectrometry validation and the personalized AI health agent, our biometric ring demonstrates broad efficacy in human body odor monitoring with evident diet-induced metabolic adherence.