Positive matrix factorization: A data preprocessing strategy for direct mass spectrometry-based breath analysis.

Area of Science:

• Analytical Chemistry
• Environmental Health
• Biomedical Engineering

Background:

• Breath analysis using direct online mass spectrometry (MS) offers non-invasive disease diagnosis and monitoring.
• Complex chemical compositions and large datasets in breath analysis hinder source identification of volatile organic compounds (VOCs).
• Positive Matrix Factorization (PMF), a receptor model, is effective for source apportionment in atmospheric studies.

Purpose of the Study:

• To explore the application of PMF for pretreatment and source apportionment of direct breath measurement data.
• To evaluate PMF's utility in analyzing exhaled VOCs from real-time MS-based breath analysis.
• To demonstrate PMF's potential for both data processing and data mining in breath analysis.

Main Methods:

• Applied the PMF receptor model to direct breath measurement data.
• Analyzed data from replicate measurements of human exhaled breath at single time points.
• Analyzed online measurement data collected over longer periods (1.5 hours).

Main Results:

• PMF successfully distinguished between endogenous compounds and background contaminants in single-time-point breath measurements.
• PMF enabled the isolation of exhaled VOC evolution over longer periods, aiding pharmacokinetic studies (e.g., ketamine).
• PMF demonstrates promise for processing and mining ambient MS-based breath analysis data.

Conclusions:

• PMF is a valuable tool for analyzing complex exhaled VOC data from real-time MS.
• The method aids in distinguishing breath components and understanding VOC temporal dynamics.
• PMF facilitates exhalome investigations, biomarker identification, and pharmacokinetic studies.