Muco-ciliary transport in the human airway is a crucial defense mechanism for removing inhaled pathogens. Optical coherence tomography (OCT) is well-suited to monitor functional dynamics of cilia and mucus on the airway epithelium. Here we demonstrate several OCT-based methods upon an actively transporting in vitro bronchial epithelial model and ex vivo mouse trachea. We show quantitative flow imaging of optically turbid mucus, semi-quantitative analysis of the ciliary beat frequency, and functional imaging of the periciliary layer. These may translate to clinical methods for endoscopic monitoring of muco-ciliary transport in diseases such as cystic fibrosis and chronic obstructive pulmonary disease (COPD).
© 2012 OSA
During respiration, humans inhale thousands of airborne pathogens per hour that are deposited onto the surface of the airways. To combat this constant influx of pathogens, the respiratory tract is lined with airway surface liquid consisting of a mucus layer and a periciliary layer (PCL) . Within the PCL, cilia on the epithelium beat in a coordinated fashion to clear mucus from the airways. In diseases such as cystic fibrosis (CF)  and chronic obstructive pulmonary disease (COPD) , thickened mucus results in defective mucociliary clearance (MCC), leading to chronic lung infection and impaired pulmonary function. As such, MCC from the lung is a critical biomarker of respiratory health . Currently, in vivo measurements of MCC are typically performed by having a patient inhale technetium particles, with clearance measured by a gamma camera . This method provides the overall clearance rates across gross sections of the lung. However, this method cannot capture fine heterogeneities in MCC, or functional metrics such as the thickness of the mucus layer and the ciliary beat frequency (CBF). OCT presents the unique opportunity to accurately measure MCC, the depth of the mucus layer, and ciliary activity in the lung, providing clinicians with a more reliable measure of MCC. In this paper we describe several methods for quantifying airway functional metrics, and demonstrate them in actively transporting in vitro models and ex vivo tissues.
Since early studies using OCT to image the respiratory tract , the applications of OCT for in vivo human imaging of the airways have been growing. Some notable studies include a 44 subject study of airway wall thickness in COPD , a 148 subject study of pre-invasive bronchial lesions , a 43 subject study of airway compliance in a variety of pulmonary diseases , the development of a 2.2 mm diameter probe for real-time imaging in the bronchus , and development of an OCT probe guided by flexible bronchoscope to detect suspicious masses in the lung . OCT imaging of MCC in the human airway represents a particular challenge because of the large dynamic range of time scales involved, from slow mucus flow rates to rapid ciliary beating. Recent work has demonstrated how a variety of cellular processes can be characterized by unique speckle fluctuation spectroscopic signatures [12,13]. Here we characterize, for the first time, the speckle fluctuation spectra of human airways, and show how selecting different passbands allows us to preferentially contrast mucus flow or ciliary activity within airway epithelium.
To quantify mucus transport rates, a method for tracking cilia-driven flow using particle tracking velocimetry (PTV) in OCT was recently demonstrated , where it was emphasized that the ability for OCT to depth-resolve flow is crucial for understanding the physiology of cilia-driven flow. However, we find that PTV is not always possible in airway mucus that is turbid and produces OCT images that are non-sparse. Importantly, people with CF, COPD, and asthma suffer from thick mucus and mucus plugging which impede pulmonary function [15–17]. In CF in particular, the ability to monitor improvements in MCC during therapeutic intervention would be beneficial . To address this, here we demonstrate a cross-correlation method for quantifying flow of thick and optically turbid mucus.
Another imaging challenge is the small size of airway cilia (~7 μm in length), which would require the use of novel sub- and single-micrometer resolution OCT systems, such as those previously reported [19,20], in order to spatially resolve them. However, as we will show below, functional information on the cilia beat frequency (CBF) can still be obtained with a more moderate axial resolution of ~3 µm, which is sufficient to resolve the PCL.
Overall, this paper describes OCT-based methods for quantifying mucus flow, CBF, and imaging ciliary activity, which are demonstrated on both in vitro and ex vivo airway epithelium models.
2.1. In vitro and ex vivo airway models
Our in vitro model for airway epithelium consists of primary, normal human bronchial epithelial (hBE) cells that are cultured on a porous membrane (Fig. 1 ). This model recapitulates several important features of the human airway epithelium . As cells develop over several weeks, they become polarized and form a contiguous epithelium, acting to exclude liquid media from their apical surface so that they reside at an air-liquid interface. Also, the hBE cells secrete mucus, which accumulates over time in culture. Importantly, hBE cells grow cilia at the apical surface, which beat in a coordinated fashion and actively transport mucus. Because these cells are cultured in a circular horizontal culture dish, unlike in vivo, they tend to transport accumulated mucus in a rotational, hurricane-like, pattern.
In this study, hBE cells were cultured on 0.4 mm pore size Millicells (Millipore, Billerica, MA) coated with collagen and maintained in air-liquid interface media (UNC Tissue Core) as described previously . Cultures were examined after 6 weeks, when the hBE cells were confluent, had fully developed cilia, and were able to transport mucus. Hurricane cultures were allowed to accumulate mucus over a 48 hr period prior to imaging. Mucus hurricane imaging was performed both with OCT (described below), and, for validation, on an Olympus IX-71 Inverted Microscope operating in standard brightfield mode, with both 4 × and 10 × objectives. Microscopy image sequences were collected using a JAI CM-030GE grey-scale camera at 90 fps and sampled over 1215 × 914 μm into 656 × 494 pixels in x × y, respectively. For histology, cells were fixed with osmium tetraoxide in perfluorocarbon, Epon-embedded, and stained with Richardson’s.
For ex vivo tissues, tracheas from 3 mice (C57BL/6) were obtained from freshly sacrificed mice and kept in saline before OCT imaging. All mice were handled according to approved protocols at the Institutional Animal Care and Use Committee (IACUC) at the University of North Carolina at Chapel Hill. Two tracheas were sliced axially and opened for imaging perpendicular to the luminal surface (as shown in Fig. 1(c)), while the third trachea was kept intact for imaging of both luminal and basal surfaces (OCT beam nearly parallel to these surfaces).
Both in vitro and ex vivo models were imaged before and after isoflurane treatment, which is known to slow ciliary activity . This was performed by incubating the cells or tissues with ~20% isoflurane for 5 minutes, rinsing with saline, and immediately imaging.
2.2. OCT system hardware and data acquisition
The spectral domain OCT system used in this study has been described in detail previously . Briefly, a Ti:sapphire laser (KMLabs, Inc.) provided light centered at 810 nm with a 3 dB bandwidth of 125 nm, corresponding to a coherence length of 2 µm in tissue. The light was directed into a free space Michelson interferometer with 10 mW of optical power incident upon the sample and imaging optics providing a transverse resolution of 12 µm and a confocal parameter of 0.28 mm. The output of the interferometer was directed into a spectrometer to sample the spectral interferogram into 2048 pixels of a line scan camera (Dalsa Piranha 2). OCT images were obtained by Fourier transformation of the spectral interferogram after processing steps including reference spectrum subtraction  and digital dispersion compensation [25,26]. The signal-to-noise ratio of this system was typically >95 dB.
B-mode OCT was performed for mucus flow studies and speckle fluctuation contrast of the PCL. Images were collected in x-z (transverse × axial) over (1−3) × 1.5 mm into (150−1000) × 1024 pixels, respectively. The camera linerate was set to either 5 or 10 kHz for collection of 100 frames in a time series, at adjustable frame rates spanning from 0.82 fps to 40 fps. M-mode OCT was also performed for cilia beat rate analysis with the same parameters as in B-mode except the x dimension was not scanned, and only 10 sequential frames were recorded.
2.3. Mucus flow imaging by cross-correlation
Imaging airway mucus flow is challenging for several reasons. Mucus flow is predominantly transverse to the OCT beam axis in typical imaging geometries, but transverse flows are not possible to measure with traditional Doppler OCT techniques [27,28]. New Doppler methods for quantifying both transverse and axial flows have been developed for hemodynamic imaging , but are affected by the distribution of sizes and anisotropy of the scattering particles within the flow volume, which may be difficult to control in airway mucus. Furthermore, mucus flow velocities are typically 2 orders of magnitude smaller than blood flow, with flow velocities typically ranging from 10−60 µm/s. This longer time scale makes it more difficult to maintain phase stability needed for Doppler OCT while particles traverse the imaging beam. On the other hand, mucus flow can be effectively frozen by imaging at moderate frame rates in the 10s of Hz, suggesting the use of PTV, such as recently reported for studying cilia-driven flow . However, in that study, the fluids were optically clear and tracer particles were added. We find that mucus that has accumulated in hBE cultures contains a large amount of cellular detritus, making it optically turbid with developed speckles in the OCT image, and thus does not fit the requirement of sparse imaging needed for PTV. Importantly, people suffering from many respiratory diseases have characteristically thick mucus.
Given these challenges, we employed a normalized, 2D cross-correlation for speckle tracking. These types of methods were originally developed for motion tracking in ultrasonic imaging [30,31], and have been previously employed in OCT for compensation of motion artifacts  and for tracking elastic deformation . Given a point (x0, z0) around which we wish to compute the velocity, we first obtained a normalized cross-correlation between two images in the time series, I1 and I2, separated by Δt12 in time, as follows:
where X and Z are the window sizes and the correlation ρ(x′,z′) is computed over and , where Δxmax and Δzmax are the maximum displacement sizes. We then found the values of x′ and z′ at which ρ(x′,z′) was at a maximum, and determined the velocity components vx = (x0-x′)/Δt12 and vz = (z0-z′)/Δt12. vx and vz were then obtained at all desired values of (x0, z0) to generate a velocity map.
Balancing the added time of computation against increasing the available dynamic range in velocity, we chose window sizes of X = 30 and Z = 10 pixels, maximum displacement sizes of Δxmax = 8 and Δzmax = 5 pixels, and sampled x0 and z0 on a 5 × 5 pixel mesh within the image. We omitted any (x0, z0) points where the root-mean-squared (RMS) pixel intensities summed over the window (the denominator in Eq. (1)) was less than a threshold value. Because mucus flow is approximately in a steady-state in the in vitro models, we reduced noise by obtaining the median values of vx and vz over time intervals Δtij across the entire time series (every 5−10 frames, or 0.5−1s intervals, over 100 frames, or 10s). It was important to adjust the time interval so that the maximum displacement would be close to but less than Δxmax or Δzmax in order to maximize the dynamic range. We also rotated the reference frame to align with that of the epithelial cell layer, which was slightly tilted with respect to the incident beam. The resulting vx and vz maps were 2 × 2 mean filtered and visualized in a hue-saturation-value (HSV) color map by assigning hue to velocity and saturation and value to RMS pixel intensity.
We applied the same method to x-y microscopy images collected for flow validation, with settings Δtij = 0.55s, X = 20 and Y = 20 pixels, Δxmax = 5 and Δymax = 13 pixels, and otherwise the same settings as for OCT. Test image sets were generated from both OCT and microscopy images, where each frame was stepped a known displacement over a simulated time series. Velocities extracted by cross-correlation were confirmed to match the simulated velocities.
2.4. Ciliary beat rate quantification by Fourier analysis
With some exceptions [19,20], most OCT systems currently do not offer the resolution needed to spatially resolve individual airway cilia. However, our OCT system axial resolution of ~3 µm is, in most cases, sufficient to resolve the PCL within the airway surface liquid , while our transverse spot size of ~12 µm covers ~12 hBE cells and ~2000 individual cilia. Despite this limitation, we expect that the speckle fluctuation arising from the movement of an ensemble of sub-resolution cilia can be related to the ciliary beat frequency (CBF), for the following reasons. It is known that neighboring cilia tend to coordinate into metachronal patterns, i.e., sharing the same CBF with a constant phase difference . The ciliary beat motion has a complicated shape (Fig. 2(a) ) but is periodic with a frequency of typically 5 Hz at room temperature and 10 Hz at 37°C for healthy airway epithelium . The speckle pattern in OCT arises from interferences between sub-resolution light scatterers, such as cilia. During the cilia beat cycle we would expect these interferences to change in a highly complicated way (Fig. 2(b)), however, fundamentally this pattern should be time-periodic. As such, we would expect the Fourier spectrum of speckle fluctuations from cilia to have peaks at the CBF and its harmonics. In practice, we find that such peaks cannot be observed and are blurred together over a bandwidth several times that of the CBF. This is possibly due to overlying mucus motion changing the light beam refraction at the air-liquid interface, mucus scattering within the PCL, or creep in the system causing phase instability on this 100−200 ms time scale.
Despite the lack of a clear CBF spectral peak, we realized that, in the assumption that the beat motion pattern remains the same when the CBF increases or decreases, (i.e., the OCT amplitude is an arbitrary, time-harmonic function that does not change pattern when the cilia speed up or slow down), one would expect the Fourier spectrum to linearly expand or contract, respectively. Mathematically one might write this as
where I is the arbitrary, time-harmonic speckle amplitude arising from cilia with a characteristic CBF of ωC, S is the power spectral density for ω > 0, SDC is the DC offset (δ is the Kronecker delta function), accounts for additive white noise (such as shot noise), and negative frequencies are omitted for convenience. For noisy data, one robust method for determining the relative scale of the spectrum S is to compute its median ωm, that is, the frequency that evenly divides the area under the power spectrum curve. According to (2), we see that ωm is linearly proportional to ωC, providing a relative measure of the CBF.
This method was implemented by obtaining the power spectral density S at each depth within an M-mode OCT scan and spatially averaging spectra over 3 pixels. The white noise, , was estimated as the average over the 1.5-2.5 kHz band (2.5 kHz being the Nyquist frequency); the exact choice of the lower limit of this band between 1 and 2 kHz had little effect on our results (<5%). After subtraction of the DC term and the white noise, the median ωm was computed. Visual inspection of each image was used to determine an 8 pixel PCL region, and the maximum value of ωm within this region was recorded for each of 10 frames collected at each of 2-3 independent positions for each sample (in vitro and ex vivo) both before and after isoflurane treatment. For the purposes of display only, ωm was thresholded to omit rows below a minimum light scattering intensity.
2.5. Speckle fluctuation contrast for PCL imaging
In order to extend the 1D method above to 2D OCT imaging of the PCL, we note that high frame rates (ωm >100 Hz) would be needed to directly capture the speckle fluctuations arising from ciliary activity. However, taking cues from OCT angiography where blood flow is often too rapid to directly capture , we can semi-quantitatively image ciliary activity by computing the standard deviation (or variance) of the speckle fluctuations. In our experiments, we find that phase is not generally stable, and we instead compute the variance of the OCT amplitude, a method which was recently shown to give comparable results to phase variance methods for blood flow imaging .
Interestingly, airway epithelium dynamics occur over a wide range of time scales, from slow mucus transport to fast ciliary beating, and so it is important to define the frequency band in which variance calculations are performed. We note that, using a discrete version of Parseval’s theorem, the variance of a time series x that is sampled at intervals of Δt over a total time T = (N-1)Δt is directly proportional to the power spectral density of x, |X|2, integrated over a frequency band bounded by the spectral resolution, Δf = 1/T, and the Nyquist frequency, fNyq = 1/(2Δt), as follows:
Using this relation, one can tailor the time scale of interest to match the imaging target by the choice of Δt and T. This is also a computationally efficient method in comparison to direct Fourier analysis, which is important when collecting 2- and 3-D imaging data.
We implemented this method by analyzing B-mode OCT time series collected at variable frame rates from 0.82−40 fps. The Nyquist frequency could be further reduced from that dictated by the sampling rate by skipping frames in the series, and Δf was adjusted by choosing how many images of the series to include in the analysis. The standard deviation σ was computed for each pixel across the time series, and was normalized by the shot noise estimated as the square root of the mean value in each pixel. Mean filtering of this signal was performed in 4 × 4 pixel windows. Using this method, we processed images of the in vitro hurricane hBE model and ex vivo mouse trachea into 3 frequency bands spanning nearly 3 decades.
3. Results and discussion
3.1. Mucus flow imaging
We studied an in vitro hBE culture exhibiting mucociliary transport of thick (~100-250 µm) mucus. Using light microscopy, we found that the transverse mucus motion was in a circular, hurricane-like pattern, consistent with previous observations in these types of cultures . B-mode OCT imaging and quantitative flow analysis was performed in cross-sectional slices moving across the center of the hurricane. A representative data set obtained near the periphery of the hurricane is shown in Fig. 3 . The real-time video suggests mucus motion predominantly parallel to the epithelial surface to the right (positive x direction). Also, one can observe a more rapid speckle decorrelation rate near both edges of the image. The computed flow maps corroborate these observations, where the velocity is largely homogeneous in the positive x direction, except near the edges where high variance is observed due to the rapid decorrelation. This decorrelation is likely due to the component of mucus motion through the image plane (vy), which is expected to be greater near the edges than in the center of the hurricane. This effect dictates a tradeoff in the choice of frame rate, as sufficiently high frame rate is needed to capture speckles in regions with a high vy component, while sufficiently low frame rate is needed to allow for speckles to move a number of pixels in the x-z plane order to minimize digitization noise in the velocity measurement. In practice, we find the optimum frame rates to lie between 1−2 fps. It may be possible, in future work, to quantify vy based upon this loss of correlation using a previously described method ; in this case, careful calibration would be needed to account for the beam divergence over the large imaging depth range needed to assess thick mucus.
We note that OCT can provide new information about the axial component of velocity, vz, and the depth dependence of vx during MCC. Here we find that vx appears to be constant in z except near the air-liquid interface; the surface effect may be an artifact of more rapid decorrelation at the interface due to changes in surface height. The homogeneous depth profile of vx is consistent with the view that mucus travels as a slab. In comparison, vz is not significantly different from zero. It will be interesting in future work to learn whether mucus flow in living, breathing organisms is depth-dependent due to shear forces from air flow, which may lead to better models of muco-ciliary transport.
In order to validate flow quantification by OCT, we compared the depth-averaged vx to the components of flow vx′ and vy′ obtained by light microscopy; orientation was not maintained between the experiments so the axes may be different. The results are summarized in Fig. 4 . For OCT, the depth-averaged vx was computed within the mucus layer for B-mode images across a 2.7 mm × 3.0 mm (x-y) area of the hurricane. A clear transition from negatively-directed flow to positively-directed flow is seen as the B-mode slice is moved across the center of the hurricane, with values that ranged from −34 to + 35 µm/s. Microscopy was performed in the same culture immediately after OCT imaging within a 0.9 mm × 1.2 mm window, resulting in a similar transition in flow direction across the center of the hurricane, with values ranging from −41 to + 29 µm/s. As a point of reference, human mucus velocity has previously been measured to be 40 μm/s in the main bronchi and 92 μm/s in the trachea in healthy subjects .
While the OCT data was obtained over a larger area than the microscopy data, it is relevant to compare the velocity gradient across the hurricane. We find that the velocity gradient by microscopy was significantly asymmetric, with values of 25 and 61 μm s−1/mm for ∂vx′/∂y′ and ∂vy′/∂x′, respectively. At the same time, we noted an elongated physical appearance to the hurricane. In comparison, the velocity gradient ∂vx/∂y by OCT was 23 μm s−1/mm, which is consistent with the slow axis observed in microscopy. Previous measures of flow in these hurricane models was reported in , where a similar range of velocities and a velocity gradient of ~77 μm s−1/mm was observed. Future improvements to our OCT system may provide 3D imaging at sufficient frame rates to enable quantification of all mucus velocity components simultaneously. Alternately, rapid scanning of slightly offset B-mode planes could provide the vy component locally, as demonstrated previously in phantoms .
3.2. Ciliary beat rate
Another important parameter is the ciliary beat frequency (CBF), because it regulates the rate of mucus clearance . In order to modulate CBF in a controlled study, we used a high dose of a gaseous anesthetic, which is expected to slow the cilia [22,45,46]. As an independent measure, we first used light microscopy to monitor the effect of 20% isoflurane added to separate hBE cultures. We found that CBF dropped from an initial value of 5.2 ± 0.5 Hz to a value indistinguishable from zero in less than 5 minutes (n = 3). After extensive washing of the isoflurane, the CBF recovered to 3.7 ± 0.9 Hz, suggesting that hBE cells remain viable after isoflurane treatment. For study with OCT, we evaluated hBE cultures that had been cleared of mucus, and healthy ex vivo mouse trachea with only a thin, transparent mucus layer, in order to directly measure the speckle fluctuations from the cilia. To validate that the median frequency, ωm, of the speckle power spectrum is correlated with ciliary activity, we performed M-mode OCT on these models before and after 20% isoflurane.
Representative OCT data is displayed in Fig. 5 . Visual inspection of the M-mode images reveals rapid speckle fluctuation within the topmost light scattering layer, which is where the PCL is expected to be located. The associated Fourier spectra versus depth reveal a strong, broadband response within this layer in comparison to the cells below. It is important to note that we also expect some motion of the basal body, where the cilia are attached, and as such, the apparent layer of high fluctuation may include some of this region beneath the actual PCL. Induced motion of mucus immediately above the PCL may also exhibit rapid speckle fluctuations. As such, our following discussion on the “apparent” PCL should be understood to include these layers, in addition to the true PCL.
Computation of ωm versus depth reveals a peak within the PCL of ~100−120 Hz, which is many times larger than the actual CBF, as expected based on the discussion in 2.4 above. Interestingly, the full width at half maximum of this peak is approximately 7.4 ± 1.5 μm in depth, consistent with the known PCL thickness . As expected, we found that ωm diminished significantly after the application of isoflurane, to a value of ~10−40 Hz for both ex vivo and in vitro models, suggesting that ωm scales proportionally with ciliary activity. The statistical results of the ωm measurements are summarized in Fig. 6 . Importantly, while our OCT system axial resolution (3 μm) is sufficient to resolve the PCL as a whole, we can detect changes in the ciliary activity without spatially resolving individual cilia.
3.3. PCL imaging using speckle fluctuation contrast
The results above provide a basis for producing images that selectively contrast ciliary activity, enabling identification of the PCL. The standard deviation contrast method described in 2.5 was applied first to clean hBE cultures (with no mucus), as shown in Fig. 7 . In some places, hBE cells have grown in multiple layers, and ciliary activity on the borders of each cell can be observed as regions of high speckle fluctuation. Next, dynamic imaging of hBE cultures with thick, turbid mucus was performed (Fig. 8 ). In this case, we observe high speckle fluctuation from mucus in frequency bands up to several Hz, including particle tracks observed particularly in the middle, 0.4-5 Hz band (Fig. 8(d)). In some sense, the optimum frame rate for mucus flow imaging (section 3.1) is within this frequency band, as it allows for speckle tracking within the mucus by cross-correlation. At the highest frequencies, however, mucus transport appears frozen, as shown in Media 4, and the ciliary activity becomes more prominent. Because PCL speckle fluctuation rates exceed 100 Hz for healthy epithelium (as shown in 3.2), we might expect even higher frame rates to provide even greater contrast between mucus and PCL.
We also performed dynamic imaging of ex vivo mouse trachea. To verify that the speckle fluctuation is specific to ciliated cell surfaces, and not an artifact of a cell-liquid interface, we first imaged a tracheal tube that was on-end (Fig. 9 ), where the outer tracheal surface is not expected to be ciliated, and the inner surface is. As expected, broad regions of high speckle fluctuation are observed only on the inner tube surfaces, corresponding to ciliary activity. Next, we imaged the along the inner tracheal surface of a tube that had been cut open axially (Fig. 9(e)), where we observed that the ciliary activity is patchy, consistent with microscopy observations of patchy ciliation. In Fig. 10 , we show how imaging of the tracheal surface varies with the choice of speckle fluctuation frequency band. Because the mucus was transparent, very little interference was observed, and excellent visualization of the PCL was provided at all 3 time scales, ranging from 0.02Hz – 20Hz. Associated videos illustrate the ciliary activity within different these time scales.
In summary, we described several methods for studying MCC in an in vitro human airway model and ex vivo mouse tracheas, including direct measurements of mucus flow using cross-correlation, indirect measurements of cilia beat frequency through a proportional parameter given by the median frequency of the power spectrum, and qualitative imaging of the PCL using variance-based contrast. We also found that ciliary activity could still be visualized underneath a thick and turbid mucus layer by choice of a sufficiently high frequency band for variance contrast (>3.3 Hz).
In addition, we measured the apparent thickness of the PCL by Fourier analysis of speckle fluctuations, and found that it was well-matched to the known thickness of PCL in healthy epithelium . While future work is needed to determine whether OCT can monitor changes in PCL thickness, this ability would be clinically useful as PCL depletion is indicated in the pathogenesis of cystic fibrosis. These methods constitute a toolkit for understanding MCC in the human respiratory system, and can potentially provide new data about mucus thickness, depth-resolved flow, and ciliary activity, even in airways with thick mucus.
The ability to translate these methods to clinical imaging depends upon several factors. Patient motion and respiration will invariably cause motion artifacts on moderate (~1 Hz) time scales that may be alleviated by motion tracking methods already used in retinal imaging . For diseases where access to lower airways is desired, the development of catheters with sufficiently small diameter is technically challenging, although several groups have already made progress in this area [7,8,10,11]. In terms of resolution, in our experiments, the cilia were not spatially resolved, which relaxes requirements on beam focusing needed in endoscopic systems. Fundamentally, the ability to depth-resolve MCC provides new insight in respiratory diseases, and may lead to better methods for treatment monitoring.
We acknowledge Tyler S. Ralston at Lawrence Livermore National Laboratory for technical assistance. Support for this work came from grants from the Cystic Fibrosis Foundation CFFT BUTTON07XX0 (Button, PI), NIH 1R01HL092964 (Boucher, PI), NIH 2R01HL07546 (Superfine, PI), NFS DMS 1100281 (Hill, PI), the Cystic Fibrosis Foundation HILL0810 (Hill, PI), and startup funds from UNC Chapel Hill (Oldenburg).
References and links
1. A. Wanner, M. Salathé, and T. G. O’Riordan, “Mucociliary clearance in the
airways,” Am. J. Respir. Crit. Care Med. 154(6 Pt 1), 1868–1902
2. J. A. Regnis, M. Robinson, D. L. Bailey, P. Cook, P. Hooper, H. K. Chan, I. Gonda, G. Bautovich, and P. T. Bye, “Mucociliary clearance in patients with cystic fibrosis
and in normal subjects,” Am. J. Respir. Crit. Care
Med. 150(1), 66–71
3. G. C. Smaldone, W. M. Foster, T. G. O’Riordan, M. S. Messina, R. J. Perry, and E. G. Langenback, “Regional impairment of mucociliary clearance in chronic
obstructive pulmonary disease,” Chest 103(5), 1390–1396
4. M. R. Knowles and R. C. Boucher, “Mucus clearance as a primary innate defense mechanism
for mammalian airways,” J. Clin. Invest. 109(5), 571–577
5. M. B. Antunes and N. A. Cohen, “Mucociliary clearance--a critical upper airway host
defense mechanism and methods of assessment,” Curr. Opin.
Allergy Clin. Immunol. 7(1), 5–10
6. C. Pitris, M. E. Brezinski, B. E. Bouma, G. J. Tearney, J. F. Southern, and J. G. Fujimoto, “High resolution imaging of the upper respiratory tract
with optical coherence tomography: a feasibility study,” Am. J.
Respir. Crit. Care Med. 157(5 Pt 1), 1640–1644
7. H. O. Coxson, B. Quiney, D. D. Sin, L. Xing, A. M. McWilliams, J. R. Mayo, and S. Lam, “Airway wall thickness assessed using computed
tomography and optical coherence tomography,” Am. J. Respir.
Crit. Care Med. 177(11), 1201–1206
8. S. Lam, B. Standish, C. Baldwin, A. McWilliams, J. leRiche, A. Gazdar, A. I. Vitkin, V. Yang, N. Ikeda, and C. MacAulay, “In vivo optical coherence tomography
imaging of preinvasive bronchial lesions,” Clin. Cancer
Res. 14(7), 2006–2011
9. J. P. Williamson, R. A. McLaughlin, W. J. Noffsinger, A. L. James, V. A. Baker, A. Curatolo, J. J. Armstrong, A. Regli, K. L. Shepherd, G. B. Marks, D. D. Sampson, D. R. Hillman, and P. R. Eastwood, “Elastic properties of the central airways in
obstructive lung diseases measured using anatomical optical coherence
tomography,” Am. J. Respir. Crit. Care Med. 183(5), 612–619
10. J. Su, J. Zhang, L. Yu, H. G Colt, M. Brenner, and Z. Chen, “Real-time swept source optical coherence tomography
imaging of the human airway using a microelectromechanical system endoscope and digital signal
processor,” J. Biomed. Opt. 13(3), 030506 (2008). [CrossRef]
11. R. G. Michel, G. T. Kinasewitz, K. M. Fung, and J. I. Keddissi, “Optical coherence tomography as an adjunct to flexible
bronchoscopy in the diagnosis of lung cancer: a pilot study,”
Chest 138(4), 984–988
12. K. Jeong, J. J. Turek, and D. D. Nolte, “Speckle fluctuation spectroscopy of intracellular
motion in living tissue using coherence-domain digital holography,”
J. Biomed. Opt. 15(3), 030514 (2010). [CrossRef]
14. S. Jonas, D. Bhattacharya, M. K. Khokha, and M. A. Choma, “Microfluidic characterization of cilia-driven fluid
flow using optical coherence tomography-based particle tracking
velocimetry,” Biomed. Opt. Express 2(7), 2022–2034
15. J. C. Hogg, F. Chu, S. Utokaparch, R. Woods, W. M. Elliott, L. Buzatu, R. M. Cherniack, R. M. Rogers, F. C. Sciurba, H. O. Coxson, and P. D. Paré, “The nature of small-airway obstruction in chronic
obstructive pulmonary disease,” N. Engl. J. Med. 350(26), 2645–2653
16. T. Aikawa, S. Shimura, H. Sasaki, M. Ebina, and T. Takishima, “Marked goblet cell hyperplasia with mucus accumulation
in the airways of patients who died of severe acute asthma attack,”
Chest 101(4), 916–921
17. R. M. Shah, W. Sexauer, B. J. Ostrum, S. B. Fiel, and A. C. Friedman, “High-resolution CT in the acute exacerbation of cystic
fibrosis: evaluation of acute findings, reversibility of those findings, and clinical
correlation,” AJR Am. J. Roentgenol. 169(2), 375–380
18. S. H. Donaldson, W. D. Bennett, K. L. Zeman, M. R. Knowles, R. Tarran, and R. C. Boucher, “Mucus clearance and lung function in cystic fibrosis
with hypertonic saline,” N. Engl. J. Med. 354(3), 241–250
19. L. Liu, J. A. Gardecki, S. K. Nadkarni, J. D. Toussaint, Y. Yagi, B. E. Bouma, and G. J. Tearney, “Imaging the subcellular structure of human coronary
atherosclerosis using micro-optical coherence tomography,” Nat.
Med. 17(8), 1010–1014
20. B. Povazay, K. Bizheva, A. Unterhuber, B. Hermann, H. Sattmann, A. F. Fercher, W. Drexler, A. Apolonski, W. J. Wadsworth, J. C. Knight, P. S. J. Russell, M. Vetterlein, and E. Scherzer, “Submicrometer axial resolution optical coherence
tomography,” Opt. Lett. 27(20), 1800–1802
21. M. L. Fulcher, S. Gabriel, K. A. Burns, J. R. Yankaskas, and S. H. Randell, “Well-differentiated human airway epithelial cell
cultures,” Methods Mol. Med. 107, 183–206 (2005). [PubMed]
22. J. H. Raphael, D. A. Selwyn, S. D. Mottram, J. A. Langton, and C. O’Callaghan, “Effects of 3 MAC of halothane, enflurane and isoflurane
on cilia beat frequency of human nasal epithelium in
vitro,” Br. J. Anaesth. 76(1), 116–121
23. A. L. Oldenburg, C. M. Gallippi, F. Tsui, T. C. Nichols, K. N. Beicker, R. K. Chhetri, D. Spivak, A. Richardson, and T. H. Fischer, “Magnetic and contrast properties of labeled platelets
for magnetomotive optical coherence tomography,” Biophys.
J. 99(7), 2374–2383
24. A. L. Oldenburg, V. Crecea, S. A. Rinne, and S. A. Boppart, “Phase-resolved magnetomotive OCT for imaging nanomolar
concentrations of magnetic nanoparticles in tissues,” Opt.
Express 16(15), 11525–11539
25. A. L. Oldenburg and R. K. Chhetri, “Digital dispersion compensation for ultrabroad-bandwidth single-camera spectral-domain polarization-sensitive OCT,” Proc. SPIE 7889, 78891V (2011).
26. D. L. Marks, A. L. Oldenburg, J. J. Reynolds, and S. A. Boppart, “Digital algorithm for dispersion correction in optical
coherence tomography for homogeneous and stratified media,”
Appl. Opt. 42(2), 204–217
27. B. White, M. Pierce, N. Nassif, B. Cense, B. Park, G. Tearney, B. Bouma, T. Chen, and J. de Boer, “In vivo dynamic human retinal blood
flow imaging using ultra-high-speed spectral domain optical coherence
tomography,” Opt. Express 11(25), 3490–3497
28. J. A. Izatt, M. D. Kulkarni, S. Yazdanfar, J. K. Barton, and A. J. Welch, “In vivo bidirectional color Doppler
flow imaging of picoliter blood volumes using optical coherence
tomography,” Opt. Lett. 22(18), 1439–1441
29. V. J. Srinivasan, H. Radhakrishnan, E. H. Lo, E. T. Mandeville, J. Y. Jiang, S. Barry, and A. E. Cable, “OCT methods for capillary
velocimetry,” Biomed. Opt. Express 3(3), 612–629
31. I. A. Hein and W. R. O’Brien, “Current time-domain methods for assessing tissue motion
by analysis from reflected ultrasound echoes-a review,” IEEE
Trans. Ultrason. Ferroelectr. Freq. Control 40(2), 84–102
32. E. A. Swanson, J. A. Izatt, M. R. Hee, D. Huang, C. P. Lin, J. S. Schuman, C. A. Puliafito, and J. G. Fujimoto, “In vivo retinal imaging by optical
coherence tomography,” Opt. Lett. 18(21), 1864–1866
34. H. Matsui, B. R. Grubb, R. Tarran, S. H. Randell, J. T. Gatzy, C. W. Davis, and R. C. Boucher, “Evidence for periciliary liquid layer depletion, not
abnormal ion composition, in the pathogenesis of cystic fibrosis airways
disease,” Cell 95(7), 1005–1015
35. S. Gueron, K. Levit-Gurevich, N. Liron, and J. J. Blum, “Cilia internal mechanism and metachronal coordination
as the result of hydrodynamical coupling,” Proc. Natl. Acad.
Sci. U.S.A. 94(12), 6001–6006
36. A. B. Lansley, M. J. Sanderson, and E. R. Dirksen, “Control of the beat cycle of respiratory tract cilia by
Ca2+ and cAMP,” Am. J. Physiol. 263(2 Pt 1), L232–L242
37. D. B. Hill, V. Swaminathan, A. Estes, J. Cribb, E. T. O’Brien, C. W. Davis, and R. Superfine, “Force generation and dynamics of individual cilia under
external loading,” Biophys. J. 98(1), 57–66
38. Y. Zhao, Z. Chen, C. Saxer, Q. Shen, S. Xiang, J. F. de Boer, and J. S. Nelson, “Doppler standard deviation imaging for clinical
monitoring of in vivo human skin blood flow,”
Opt. Lett. 25(18), 1358–1360
39. G. Liu, L. Chou, W. Jia, W. Qi, B. Choi, and Z. Chen, “Intensity-based modified Doppler variance algorithm:
application to phase instable and phase stable optical coherence tomography
systems,” Opt. Express 19(12), 11429–11440
41. W. M. Foster, E. Langenback, and E. H. Bergofsky, “Measurement of tracheal and bronchial mucus velocities
in man: relation to lung clearance,” J. Appl. Physiol. 48(6), 965–971
42. H. Matsui, S. H. Randell, S. W. Peretti, C. W. Davis, and R. C. Boucher, “Coordinated clearance of periciliary liquid and mucus
from airway surfaces,” J. Clin. Invest. 102(6), 1125–1131
43. Z. Y. Shen, M. Wang, Y. H. Ji, Y. H. He, X. S. Dai, P. Li, and H. Ma, “Transverse flow velocity quantification using optical
coherence tomography with correlation,” Laser Phys.
Lett. 8(4), 318–323
44. E. Puchelle, J. M. Zahm, and D. Quemada, “Rheological properties controlling mucociliary
frequency and respiratory mucus transport,”
Biorheology 24(6), 557–563
45. A. Robertson, W. Stannard, C. Passant, C. O’Callaghan, and A. Banerjee, “What effect does isoflurane have upon ciliary beat
pattern: an in vivo study,” Clin. Otolaryngol.
Allied Sci. 29(2), 157–160