Breath-Based Liver Disease Prediction: Advances, Challenges, and Future Prospect

Abstract

Liver disease remains a significant global health challenge, often diagnosed at advanced stages due to the lack of early detection methods. Exhaled breath analysis has emerged as a promising non-invasive diagnostic approach, leveraging volatile organic compounds (VOCs) as biomarkers of liver dysfunction. Recent advancements in artificial intelligence (AI), particularly deep learning models such as LSTM, BiLSTM, 1D CNN, and GRU, have enhanced the accuracy of VOC pattern recognition, improving dis- ease prediction capabilities. Additionally, the development of sensor technologies, including gas chromatography- mass spectrometry (GC-MS), electronic noses (E-noses), and spectroscopy-based methods, has further strengthened the feasibility of breath-based diagnostics. Despite these advancements, challenges such as biological variability, environmental influences, standardization of VOC detection, and regulatory hurdles persist. The integration of AI- driven models with portable and cost- effective breath analyzers holds promise for real-time screening and continuous monitoring. Future research should focus on large- scale clinical validation, interdisciplinary collaboration, and multi-disease detection potential to establish exhaled breath analysis as a reliable diagnostic tool. This review highlights recent progress, existing challenges, and future directions in the field, emphasizing the role of breath analysis in revolutionizing liver disease diagnosis and management.