Real-Time Detection of Trace Breath Isoprene Based on Circular Domain Spectral Reconstruction Filtering Combined with...

Area of Science:

• Analytical Chemistry
• Biomedical Engineering
• Spectroscopy

Background:

• Trace isoprene in breath is a potential noninvasive biomarker for lung cancer diagnosis.
• Detection challenges include low concentrations (parts per billion) and interfering compounds in breath.
• Existing methods require complex sample handling and lack real-time capabilities.

Purpose of the Study:

• To develop and validate a novel optical sensor for real-time, noninvasive detection of trace isoprene in breath.
• To address challenges posed by interfering components and low analyte concentrations.
• To advance breath analysis using ultraviolet differential optical absorption spectroscopy (UV-DOAS).

Main Methods:

• Utilized ultraviolet differential optical absorption spectroscopy (UV-DOAS) to obtain isoprene differential absorption spectra.
• Developed a circular domain reconstruction filtering method to mitigate noise and remove interference from water vapor, ammonia, and nitric oxide.
• Constructed a convolutional neural network (CNN) model for accurate isoprene concentration inversion from filtered spectra.

Main Results:

• Achieved a detection limit of 3.98 ppb·m for the optical sensor.
• Demonstrated accurate and real-time breath isoprene sensing across a concentration range of 21.32 to 1254.20 ppb.
• Validated the sensor's effectiveness using human breath samples.

Conclusions:

• The proposed optical sensor enables the first real-time detection of breath isoprene using UV-DOAS.
• Circular domain reconstruction filtering and CNN effectively overcome detection challenges in complex breath matrices.
• The sensor shows significant potential for noninvasive lung cancer diagnosis and broadens the application of broadband spectroscopy in breath analysis.