Abstract

Bio-aerosols containing Bacillus subtilis var. niger (BG) were detected at a distance of 3 km with a passive Fourier Transform InfraRed (FTIR) spectrometer in an open-air environment where the thermal contrast was low (~ 1 K). The measurements were analyzed with a new hyperspectral detection, identification and estimation algorithm based on radiative transfer theory and advanced signal processing techniques that statistically subtract the undesired background spectra. The results are encouraging as they suggest for the first time the feasibility of detecting biological aerosols with passive FTIR sensors. The number of detection events was small but statistically significant. We estimate the false alarm rate for this experiment to be 0.0095 and the probability of detection to be 0.61 when a threshold of detection that minimizes the sum of the probabilities of false alarm and of missed detection is chosen.

© 2003 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |

  1. P. L. Hanst and S. T. Hanst, �??Gas measurement in the fundamental infrared region,�?? in Air monitoring by spectroscopic techniques, M. W. Sigrist, ed. (Wiley, New-York, NY, 1994).
  2. D. W. T. Griffith and I. M. Jamie, �??Fourier transform infrared spectrometry in atmospheric and trace gas analysis,�?? in Encyclopedia of analytical chemistry, R. A. Meyers, ed. (Wiley, Chichester, England, 2000).
  3. D. A. Ligon, A. E. Wetmore, and P. S. Gillespie, �??Simulation of the passive infrared spectral signatures of bio-aerosol and natural fog clouds immersed in the background atmosphere,�?? Opt. Express, 10, 909-919, (2002), <a href="http://www.opticsexpress.org/abstract.cfm?URI=OPEX-10-18-909">http://www.opticsexpress.org/abstract.cfm?URI=OPEX-10-18-909</a>
    [CrossRef] [PubMed]
  4. E. R. Schildkraut, R. Connors and A. Ben-David, �??Initial test results from ultra-high sensitivity passive FTIR instrumentation (HISPEC)�??, Proceedings of the International Symposium on Spectral Sensing Research, (ISSSR) 2001, (Science and Technology Corp. Hampton, VA, 2001), pp. 365-374.
  5. A. Ben-David , and A. Ifarraguerri, �??Computation of a spectrum from a single-beam Fourier-transform infrared interferogram,�?? Appl. Opt. 41, 1181-1189 (2002).
    [CrossRef] [PubMed]
  6. A. Ben-David and H. Ren, �??Detection, identification and estimation of aerosols and vapors with Fourier transform infrared spectrometer,�?? to be published in Appl. Opt. (2003).
    [CrossRef] [PubMed]
  7. C. M. Bishop, Neural Networks for Pattern Recognition, (Ch. 2, Oxford University Press, Oxford, New York, 1995).

Appl. Opt. (2)

A. Ben-David and H. Ren, �??Detection, identification and estimation of aerosols and vapors with Fourier transform infrared spectrometer,�?? to be published in Appl. Opt. (2003).
[CrossRef] [PubMed]

A. Ben-David , and A. Ifarraguerri, �??Computation of a spectrum from a single-beam Fourier-transform infrared interferogram,�?? Appl. Opt. 41, 1181-1189 (2002).
[CrossRef] [PubMed]

Opt. Expres (1)

D. A. Ligon, A. E. Wetmore, and P. S. Gillespie, �??Simulation of the passive infrared spectral signatures of bio-aerosol and natural fog clouds immersed in the background atmosphere,�?? Opt. Express, 10, 909-919, (2002), <a href="http://www.opticsexpress.org/abstract.cfm?URI=OPEX-10-18-909">http://www.opticsexpress.org/abstract.cfm?URI=OPEX-10-18-909</a>
[CrossRef] [PubMed]

Other (4)

E. R. Schildkraut, R. Connors and A. Ben-David, �??Initial test results from ultra-high sensitivity passive FTIR instrumentation (HISPEC)�??, Proceedings of the International Symposium on Spectral Sensing Research, (ISSSR) 2001, (Science and Technology Corp. Hampton, VA, 2001), pp. 365-374.

P. L. Hanst and S. T. Hanst, �??Gas measurement in the fundamental infrared region,�?? in Air monitoring by spectroscopic techniques, M. W. Sigrist, ed. (Wiley, New-York, NY, 1994).

D. W. T. Griffith and I. M. Jamie, �??Fourier transform infrared spectrometry in atmospheric and trace gas analysis,�?? in Encyclopedia of analytical chemistry, R. A. Meyers, ed. (Wiley, Chichester, England, 2000).

C. M. Bishop, Neural Networks for Pattern Recognition, (Ch. 2, Oxford University Press, Oxford, New York, 1995).

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (4)

Fig. 1.
Fig. 1.

The mean correlation coefficient of the detection events (circles) as a function of time and measurement number. Six time regions (a to f) and the time of the start of the BG release are noted in the figure. The cluster in region (b) contains 25 detection events that are believed to be valid detection events. The detections in all other regions are believed to be due to false alarm. The algorithm did not produce any detection events in regions (d) and (f). The total number of detection events in each region is given in Table 1.

Fig. 2.
Fig. 2.

The averaged deduced spectra (blue) and the library reference spectrum of BG (green) for the detection events in the regions (a), (b), (c) and (e) of Fig. 1. No detection events were found in regions (d) and (f).

Fig. 3.
Fig. 3.

Estimated mass-column density for measurements in region (a) and (b) of Fig. 1. The location of the detection events (Fig. 1) is marked as red circles. The start of the BG release is at measurement 1250.The sharp contrast between the estimated mass-column density at the location of the cluster of detection events (around measurements 1550 to 1750) and the remainder of the measurements suggests that the cluster of detection events consists of valid (real) detection events.

Fig. 4.
Fig. 4.

The probability model (blue) for mass-column concentration ρ (Fig. 3), the data histogram (green) and the two Gaussians pdf (red) for the two hypothesis; H 0 (background + noise) and H 1 (background + noise + BG cloud within the sensor’s FOV). The threshold of separation, γ=322 mg m -2, between the two hypotheses is marked. Mass-column density ρ > γ belongs to the hypothesis H 1.

Tables (2)

Tables Icon

Table 1. Number of detection events for each of the time regions (a to f) in Fig. 1.

Tables Icon

Table 2. XM94 lidar data of the BG cloud mapping. The azimuthal width of the cloud is ~100 m and the depth (radial distance) of the cloud is ~200 m (XM94 video display).

Equations (19)

Equations on this page are rendered with MathJax. Learn more.

M ( λ ) = M 0 ( λ ) [ M 0 ( λ ) B λ T ] α ( λ ) ρ
OSP ( M ) = OSP ( [ M 0 B ] α ρ )
pdf ( ρ ) = w 0 pdf ( ρ H 0 ) + ( 1 w 0 ) pdf ( ρ H 1 )
N ( x ; μ , σ 2 ) = 1 2 π σ 2 exp ( ( x μ ) 2 2 σ 2 )
pdf ( ρ H 0 ) = N ρ μ 0 σ 0 2
pdf ( ρ H 1 ) = N ρ μ 1 σ 1 2
pdf ( H 0 ρ ) = pdf H 0 ρ pdf ( ρ ) = w 0 pdf ( ρ H 0 ) pdf ( ρ )
pdf ( H 1 ρ ) = pdf H 1 ρ pdf ( ρ ) = ( 1 w 0 ) pdf ( ρ H 1 ) pdf ( ρ )
w 0 = pdf ( H 0 ρ ) d ρ
μ 0 = w 0 1 pdf ( H 0 ρ ) ρ d ρ
μ 1 = ( 1 w 0 ) 1 pdf ( H 1 ρ ) ρ d ρ
σ 0 2 = w 0 1 pdf ( H 0 ρ ) ( ρ μ 0 ) 2 d ρ
σ 1 2 = ( 1 w 0 ) 1 pdf ( H 1 ρ ) ( ρ μ 1 ) 2 d ρ
w 0 N ( γ ; μ 0 , σ 0 2 ) = ( 1 w 0 ) N ( γ ; μ 1 , σ 1 2 )
n det ections = ( n n 1 ) P FA + n 1 P D
P D ( γ ) = γ pdf ( x H 1 ) dx = 2 1 [ 1 erf ( γ μ 1 2 1 2 σ 1 ) ] = 0.61
P FA ( γ ) = γ pdf ( x H 0 ) dx = 2 1 [ 1 erf ( γ μ 0 2 1 2 σ 0 ) ] = 0.0095
erf ( x ) = 1 2 π 0 x exp ( 0.5 t 2 ) dt
n 1 = ( n det ections n P FA ) ( P D P FA ) 1 = 240

Metrics