Cavity-enhanced optical frequency comb spectroscopy of human breath (Thorpe et al., 2008) demonstrated a near-infrared (1.5–1.7 µm) spectrometer with 200 nm of instantaneous spectral coverage, 800 MHz resolution, a minimum detectable absorption of 8×10−10 cm−18\times10^{-10}\ \mathrm{cm}^{-1}, and parts-per-billion detection of CO, CO2_2 isotopologues, CH4_4 and NH3_3, while acknowledging a five-part development program: extension to the mid-infrared (near 3 µm) where absorption strengths grow by factors of 10–10310^{3}; high-finesse cavity architectures covering several microns instead of ~200 nm; management of the water-vapor background that limits NH3_3 detection; reduction of the residual noise floor; and large-scale clinical validation of breath-biomarker diagnostics. Since then, the same lineage has substantially realized the instrumental directions: mid-infrared cavity-enhanced direct frequency comb spectroscopy (CE-DFCS) reached parts-per-trillion multi-species breath detection (2021), a pilot clinical study of SARS-CoV-2 detection with 170 subjects and an area under the receiver-operating-characteristic curve (ROC AUC, a standard aggregate measure of diagnostic discrimination) of 0.849 was completed (2023), and modulated ringdown comb interferometry achieved 1010 cm−1^{-1} of mid-infrared coverage at cavity finesse 23,000 with simultaneous quantification of 20 breath species at parts-per-trillion sensitivity, robust to the strong dispersion of exhaled breath (2025). The precise surviving open core is clinical translation: large-scale, standardized validation of diagnostic breath-biomarker correlations and a deployable (point-of-care) comb-based breath instrument, together with the residual instrumental goals of multi-micron single-instrument coverage and miniaturization.

Background

Human breath contains over a thousand identified volatile compounds whose concentrations encode physiological and pathological processes, and several of these 'breath biomarkers' are correlated with specific diseases or metabolic states. Because breath sampling is non-invasive and fast, optical detection of exhaled biomarkers has long been pursued as an alternative to blood testing, but a practical instrument must combine broad spectral coverage (many molecules absorb in disjoint spectral windows), high spectral resolution (breath spectra are congested, with concentration contrasts of 10310^{3}–10910^{9} between abundant H2_2O/CO2_2 and trace species), high sensitivity at the parts-per-billion (ppb) to parts-per-trillion (ppt) level, and measurement times compatible with clinical throughput. An optical frequency comb is a phase-stabilized mode-locked laser whose spectrum is a grid of evenly spaced, narrow lines (teeth) that act as hundreds of thousands of simultaneous precision probes of molecular absorption; coupling a comb into a high-finesse optical cavity (two ultra-reflective mirrors, with the cavity mode spacing matched to the comb repetition rate; the cavity 'finesse', the ratio of the cavity line spacing to the resonance linewidth, sets the enhancement) multiplies the effective path length by thousands of round trips and yields cavity-enhanced direct frequency comb spectroscopy (CE-DFCS). Thorpe, Balslev-Clausen, Kirchner and Ye (2008) built the first such system applied to human breath: a near-infrared (1.5–1.7 µm) comb from an erbium-doped-fiber mode-locked laser in a cavity whose mirrors supported only ~200 nm of instantaneous coverage, giving a minimum detectable absorption of 8×10−10 cm−18\times10^{-10}\ \mathrm{cm}^{-1}, 800 MHz resolution, ppb-level detection of CO, CH4_4 and NH3_3, and sub-minute measurement times. The paper's discussion explicitly framed the follow-on program: comb technology was 'currently being extended' to the 3 µm region, where fundamental (rather than overtone) vibrational bands raise molecular absorption strengths by factors of 10–10310^{3} and should improve detection limits; new high-finesse prism-cavity designs were expected to expand cavity spectral coverage 'from 200 nm to several microns', greatly increasing the number of simultaneously accessible biomarkers; water-vapor absorption overlap was the practical limit for NH3_3 (18 ppb minimum detectable concentration in lab air at 25% relative humidity, higher in breath, with cold-trap drying suggested but its analyte-loss trade-offs unstudied); residual laser intensity noise, cavity–laser coupling noise and interference fringes set the absorption noise floor; and translation into a clinical screening tool would require large-scale trials to gather the biomarker statistics needed to institute non-invasive health-screening tests. Reviews of the field (Adler et al., 2010; Henderson et al., 2018) later confirmed that comb-based cavity spectroscopy advanced rapidly while clinical breath analysis as a whole remained stuck before standardized, validated deployment. In the years since, the same lineage has closed much of the instrumental program: mid-infrared CE-DFCS breath spectroscopy reached parts-per-trillion detection (Liang et al., 2021), a 170-subject pilot study of SARS-CoV-2 breath detection reported a ROC AUC of 0.849 (Liang et al., 2023), and modulated ringdown comb interferometry — a variant that measures cavity ringdown decays (the exponential decay of light circulating in the cavity after the laser is interrupted) on many comb lines at once, and is therefore insensitive to the strong dispersion that water and CO2_2 impose on cavity transmission — demonstrated 1010 cm−1^{-1} of mid-infrared coverage directly on exhaled breath (Liang et al., 2025). The clinical closure has not been demonstrated: a large-scale, standardized validation of comb-measured breath biomarkers and a deployable instrument.

Problem Statement

The problem is the development program posed by Thorpe et al. (2008) for frequency-comb-based cavity-enhanced breath analysis: (i) extend the technique from the near-infrared (1.5–1.7 µm, 8×10−10 cm−18\times10^{-10}\ \mathrm{cm}^{-1} minimum detectable absorption, ~200 nm coverage) into the mid-infrared near 3 µm, where molecular absorption strengths are 10–10310^{3} times larger, by integrating a robust broadband comb source with a compatible high-finesse cavity; (ii) expand single-instrument spectral coverage from ~200 nm toward several microns so that many more breath biomarkers are accessible at once; (iii) remove the water-vapor background limitation on trace-species detection (notably NH3_3) without uncontrolled analyte loss from sample pretreatment; (iv) reduce the absorption noise floor below the intensity-noise, coupling-noise and etalon-fringe limits (etalon fringes being interference fringes from parallel optical surfaces); and (v) validate the instrument clinically through large-scale trials that establish standardized breath-sampling protocols and statistically confirmed biomarker–disease correlations. Since then, components (i), (iii) and (iv) have been substantially achieved — mid-infrared CE-DFCS breath spectroscopy now reaches parts-per-trillion detection with near-shot-noise-limited readout, and modulated ringdown comb interferometry operates directly on exhaled breath with 1010 cm−1^{-1} of coverage at finesse 23,000 — so the precise surviving open core is: complete the clinical translation of comb-based breath analysis, namely execute the large-scale, standardized clinical validation, on independent, prospectively enrolled cohorts, that establishes which comb-measured breath biomarker patterns carry diagnostic value (and at what sensitivity and specificity); define the reference ranges and sampling protocols needed for reproducible inter-study comparison across independent cohorts and sites; and realize the remaining instrumental steps (instantaneous multi-micron cavity coverage and a deployable, ideally point-of-care, instrument) that clinical deployment requires.

References

  1. Michael J. Thorpe, David Balslev-Clausen, Matthew S. Kirchner, Jun Ye (2008). Cavity-enhanced optical frequency comb spectroscopy: application to human breath analysis. Optics Express 16(4), 2387–2397. DOI: 10.1364/oe.16.002387, https://doi.org/10.1364/oe.16.002387 ; arXiv:0708.3205, https://arxiv.org/abs/0708.3205.
  2. F. Adler, M. J. Thorpe, K. C. Cossel, Jun Ye (2010). Cavity-enhanced direct frequency comb spectroscopy: technology and applications. Annual Review of Analytical Chemistry 3, 175–205. DOI: 10.1146/annurev-anchem-060908-155248, https://doi.org/10.1146/annurev-anchem-060908-155248 (PMID 20636039).
  3. B. Henderson, A. Khodabakhsh, M. Metsälä, I. Ventrillard, F. M. Schmidt, D. Romanini, G. A. D. Ritchie, S. Te Lintel Hekkert, R. Briot, T. Risby, N. Marczin, F. J. M. Harren, S. M. Cristescu (2018). Laser spectroscopy for breath analysis: towards clinical implementation. Applied Physics B 124, 161. DOI: 10.1007/s00340-018-7030-x, https://doi.org/10.1007/s00340-018-7030-x (PMID 30956412).
  4. Qizhong Liang, Ya-Chu Chan, P. Bryan Changala, David J. Nesbitt, Jun Ye, Jutta Toscano (2021). Ultrasensitive multispecies spectroscopic breath analysis for real-time health monitoring and diagnostics. Proceedings of the National Academy of Sciences 118(40), e2105063118. DOI: 10.1073/pnas.2105063118, https://doi.org/10.1073/pnas.2105063118 ; author preprint arXiv:2103.09221, https://arxiv.org/abs/2103.09221 (the preprint bears the variant title 'Ultra-sensitive multi-species spectroscopic breath analysis for real-time health monitoring and diagnostics').
  5. Qizhong Liang, Ya-Chu Chan, Jutta Toscano, Kristen K. Bjorkman, Leslie A. Leinwand, Roy Parker, Eva S. Nozik, David J. Nesbitt, Jun Ye (2023). Breath analysis by ultra-sensitive broadband laser spectroscopy detects SARS-CoV-2 infection. Journal of Breath Research 17(3), 036001. DOI: 10.1088/1752-7163/acc6e4, https://doi.org/10.1088/1752-7163/acc6e4 ; arXiv:2202.02321, https://arxiv.org/abs/2202.02321 (PMID 37016829).
  6. David Nesbitt (2024). Ultrasensitive cavity-enhanced frequency comb breath spectroscopy for point-of-care medical diagnostics. Proceedings of SPIE, conference 'Quantum Effects and Measurement Techniques in Biology and Biophotonics'. DOI: 10.1117/12.3000101, https://doi.org/10.1117/12.3000101.
  7. Qizhong Liang, Apoorva Bisht, Andrew Scheck, Peter G. Schunemann, Jun Ye (2025). Modulated ringdown comb interferometry for sensing of highly complex gases. Nature 638(8052), 941–948. DOI: 10.1038/s41586-024-08534-2, https://doi.org/10.1038/s41586-024-08534-2 ; author preprint arXiv:2406.03609, https://arxiv.org/abs/2406.03609 (the preprint bears the variant title 'Modulated Ringdown Comb Interferometry for next-generation high complexity trace gas sensing').
  8. Anna Paleczek, Artur Rydosz (2025). Current clinical trials of exhaled breath analysis. In: Exhaled Breath Analysis (Elsevier). DOI: 10.1016/b978-0-443-33796-3.00004-0, https://doi.org/10.1016/b978-0-443-33796-3.00004-0.