A Review of Breath Sensing for Disease Monitoring and Diagnosis: Breathomics, Artificial Intelligence, and Nanomaterials for the Future of Medicine

ABSTRACT

Breathomics has advanced into a non‐invasive diagnostic strategy capable of detecting disease‐specific volatile organic compounds (VOCs) linked to altered metabolic and inflammatory pathways. Since the first identification of over 200 VOCs in human breath, research has shown strong diagnostic potential across diabetes, lung cancer, chronic obstructive pulmonary disease, asthma, pulmonary arterial hypertension, tuberculosis, renal disease, liver cirrhosis, gastric cancer, and Helicobacter pylori infection (all covered herein). Key biomarkers such as acetone (elevated to >2 ppm in diabetes), 1‐octene (lung cancer), ammonia (>1500 ppb in chronic kidney disease), and nitric oxide (reduced in pulmonary arterial hypertension, PAH) demonstrate clinically relevant discrimination; for example, Au‐nanoparticle sensors have been shown to achieve a 92% accuracy distinguishing PAH patients from controls. Progress is driven by nanomaterials including graphene, carbon nanotubes, noble‐metal nanoparticles, metal‐organic frameworks, MoS ₂ , and metal oxides such as SnO ₂ and ZnO, which offer high sensitivity and response. Emerging fabrication techniques such as screen printing, inkjet printing, and aerosol jet printing enable scalable production of chemiresistive, colorimetric, electrochemical, and SERS‐based sensors. Combined with AI‐driven pattern recognition, these technologies position breath‐based diagnostics as a realistic pathway toward portable, real‐time, and affordable disease monitoring.