A Portable Real-Time Ringdown Breath Acetone Analyzer: Toward Potential Diabetic Screening and Management

Chenyu Jiang, Meixiu Sun, Zhennan Wang, Zhuying Chen, Xiaomeng Zhao, Yuan Yuan, Yingxin Li, Chuji Wang2016 · DOI

TL;DR

This paper introduces a novel, portable, real-time breath acetone analyzer based on cavity ringdown spectroscopy (CRDS). The device was validated against GC-MS and used to measure breath acetone in a large cohort of diabetic and healthy individuals, suggesting its potential for non-invasive diabetes screening and management.

What Problem Does This Paper Address?

Diabetes management and screening currently rely heavily on blood glucose meters (BGMs), which can be invasive, costly, and prone to measurement errors. Breath analysis offers a promising non-invasive alternative, with breath acetone identified as a potential biomarker for diabetes. However, previous studies have yielded inconsistent results due to a lack of standardized measurement methods and limited sample sizes. This paper aims to address this by developing and validating a high-throughput, portable breath acetone analyzer to enable more robust clinical studies and pave the way for point-of-care diabetes diagnostics.

Methods

The core of this work is the development of a LaserBreath-001, a portable breath acetone analyzer utilizing Cavity Ringdown Spectroscopy (CRDS). CRDS is a laser-based technique known for its high sensitivity and accuracy in trace gas analysis. The instrument uses a 266 nm pulsed Nd:YAG laser coupled into a gas cell. Breath samples are introduced into the cell, and the decay time of the laser light within the cell is measured. Acetone concentration is determined by analyzing changes in this decay time, employing a background subtraction method to account for the laboratory atmosphere. The instrument was validated against a certified Gas Chromatography-Mass Spectrometry (GC-MS) system, a standard analytical technique for volatile organic compounds.

Datasets

The study collected and analyzed 1257 breath samples from 386 human subjects. This cohort included:

  • 22 Type 1 Diabetic (T1D) patients
  • 312 Type 2 Diabetic (T2D) patients (noted as one of the largest T2D cohorts in a single study for this type of analysis)
  • 52 non-diabetic healthy subjects

Samples were collected under various conditions, including fasting, 2 hours post-breakfast, 2 hours post-lunch, and 2 hours post-dinner. Blood glucose (BG) levels were measured simultaneously using a standard diabetic management BG meter.

Key Insights and Ideas

  • CRDS for Breath Analysis is Viable: The paper demonstrates that CRDS, a sensitive laser spectroscopy technique, can be effectively miniaturized and adapted for portable breath analysis, offering a promising alternative to traditional methods.
  • High-Throughput Measurement is Crucial: The authors highlight that previous inconsistencies in breath acetone studies are partly due to limited sample sizes. Their high-throughput analyzer allows for the collection of a significantly larger dataset, leading to more reliable conclusions.
  • Breath Acetone is Elevated in Diabetics: The study provides strong evidence that diabetic individuals (both T1D and T2D) generally exhibit elevated mean breath acetone concentrations compared to healthy individuals.
  • Humidity is a Key Interferent: The research identifies humidity in breath samples as a significant factor affecting measurement stability and reproducibility, necessitating the use of a membrane filter to mitigate this issue.
  • Correlation with BG is Complex: While a general trend of elevated acetone in diabetics is observed, a direct, individual-level correlation between breath acetone and blood glucose was not consistently found across all groups. However, a linear correlation was observed in T1D subjects without ketoacidosis when analyzing group means.

Key Results

  • GC-MS Validation: The LaserBreath-001 analyzer showed good agreement with GC-MS, with a linear fitting coefficient of 0.98 and a slope of 0.99, indicating consistent acetone concentration measurements.
  • Mean Acetone Concentrations:
    • T1D subjects: 4.9 ± 16 ppm
    • T2D subjects: 1.5 ± 1.3 ppm
    • Healthy subjects: 1.1 ± 0.5 ppm
    • Diabetic subjects (T1D and T2D combined) showed mean breath acetone concentrations approximately 4.5 and 1.4 times higher than healthy subjects, respectively.
  • Correlation in T1D (No Ketoacidosis): A preliminary quantitative correlation (R = 0.56, p < 0.05) was found between the mean breath acetone concentration and mean BG levels in 20 T1D subjects without ketoacidosis.
  • Screening Accuracy: For T1D subjects under fasting conditions, the screening accuracy of an individual breath acetone measurement was estimated to be around 74%.
  • Instrument Stability: The instrument demonstrated good stability, with a measured stability of 0.17% and a theoretical detection limit of 30 ppb.

Limitations and Future Work

  • Interference from Physiological Factors: The authors acknowledge that many physiological parameters (fasting, exercise, diet, etc.) affect breath acetone levels, requiring specific controlled conditions for accurate diabetes-specific measurements.
  • Individual vs. Group Correlation: While group means showed correlations, direct individual-level correlations between breath acetone and BG were not consistently observed, suggesting that a single breath acetone measurement might not be sufficient for precise diabetes management without further context.
  • Screening Accuracy: The 74% screening accuracy for T1D suggests room for improvement, particularly in reducing false positives.
  • Future Work: The authors propose further development to:
  • Reduce instrument size and weight for a true point-of-care device.
  • Develop algorithms to minimize interferences from other physiological parameters.
  • Optimize measurement conditions to specifically detect diabetes-related acetone.
  • Conduct larger clinical studies to further validate the device and explore its utility in diabetes screening and management, especially for monitoring ketone bodies and evaluating exercise effects.

Summary updated

BibTeX
@misc{jiang2016portable,
  author = {Chenyu Jiang and Meixiu Sun and Zhennan Wang and Zhuying Chen and Xiaomeng Zhao and Yuan Yuan and Yingxin Li and Chuji Wang},
  title = {{A Portable Real-Time Ringdown Breath Acetone Analyzer: Toward Potential Diabetic Screening and Management}},
  year = {2016},
  doi = {10.3390/s16081199},
}
Figure