An optical sensor for ppb-level exhaled acetone detection based on UV-DOAS and spectral upscaling

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Biomedical Engineering

Background:

• Acetone in breath is a promising biomarker for diabetes mellitus (DM) monitoring.
• Detecting exhaled acetone using ultraviolet differential optical absorption spectroscopy (UV-DOAS) is challenging due to complex gas mixtures and baseline drift.
• Existing UV-DOAS methods are rarely applied to exhaled acetone detection.

Purpose of the Study:

• To develop and validate an improved UV-DOAS optical sensor for detecting exhaled acetone.
• To enable the detection of acetone in the sub-200 nm wavelength band for the first time.
• To provide a non-invasive tool for early diabetes diagnosis.

Main Methods:

• Improved UV-DOAS with segmental fitting to mitigate baseline drift and obtain a standard acetone differential absorption spectrum.
• Spectral upscaling concentration inversion method using a wavelet coefficient matrix to handle spectral overlaps (O2, NH3, acetone).
• Utilized time-frequency information from spectral upscaling for enhanced spectral analysis.

Main Results:

• Achieved a laboratory detection limit of 14.97 ppb*m, demonstrating exceptional performance.
• Successfully detected acetone at ppb levels in human exhaled breath samples.
• Observed increased acetone concentrations correlating with elevated lipid metabolism.

Conclusions:

• The developed optical sensor accurately and stably detects exhaled acetone at ppb levels.
• The improved UV-DOAS technique effectively overcomes challenges in analyzing complex breath samples.
• This sensor shows significant potential for non-invasive early diagnosis of diabetes mellitus.