Endoscopic treatment of lower airway pathologies requires accurate quantification of airway dimensions. We demonstrate the application of a real-time endoscopic optical coherence tomography system that can image lower airway anatomy and quantify airway lumen dimensions intra-operatively. Results demonstrate the ability to acquire 3D scans of airway anatomy and include comparison against a pre-operative X-ray CT. The paper also illustrates the capability of the system to assess the real-time dynamic changes within the airway that occur during respiration.
© 2008 Optical Society of America
The lower airway is a complex structure composed of branching, tapering airways, and extends from below the vocal cords through to the terminal bronchioles. Its primary function is to conduct air from the upper airway to the gas exchange portions of the lungs. The airway tree can be considered in terms of large (8–25mm diameter); intermediate (2–8mm); and small (<2mm) airways. From trachea to terminal bronchioles, the airway branches up to 25 times.
Several important pathologies may affect the lower airways, including asthma (chronic inflammatory disease usually triggered by allergens); bronchiectasis (chronic inflammation and dilation of the airways and a build-up of mucus); and chronic obstructive lung disease (fixed obstruction to airflow caused predominantly by long term exposure to cigarette smoke). In addition, certain processes may lead to focal narrowing (stenosis) of the large or intermediate sized airways, resulting in obstruction to airflow and subsequent clinical conditions. The etiology of such stenoses includes both benign conditions and malignant growths such as primary lung, esophageal and metastatic cancer.
Endoscopic clinical interventions in the lower airway are performed by specialist interventional pulmonologists and by ear, nose and throat surgeons. Pre-operative assessment and planning requires accurate imaging of the airway dimensions and is critical to help guide these interventions.
There are several imaging modalities currently available. Pre-operative X-ray computed tomography (CT) allows visualization and quantification of the lower airway. However, for optimal pre-operative planning, the CT should be performed close to the time of bronchoscopy. This is often not possible for both logistical reasons and issues of radiation dosage. In addition, CT cannot provide measurements during an interventional procedure. Real-time measurement is important because airway dimensions are not static but may vary with changes in regional anatomy (such as cartilaginous rings and adjacent external organs). Also, airway size may be affected by alterations of lung volume, phase of respiration and cardiac motion. Finally, the radiation burden makes CT inappropriate for repeated imaging.
Endobronchial ultrasound, used increasingly to stage lymph node involvement in lung cancer, cannot quantify airway dimensions because poor transducer-air coupling confounds imaging. Initial work has been published on the use of magnetic resonance imaging (MRI) using hyperpolarized gases such as 3He or 129Xe [1,2], but this is still far from clinical use. Preliminary ex-vivo results have also demonstrated potential for non-coherent optical techniques , although such techniques have not yet been utilized in vivo.
In current clinical practice, interventions are guided using real-time video bronchoscopy. A persisting limitation of such guidance is the inability of bronchoscopy to provide quantitative images. This is because of the lack of depth discrimination in the single 2D image, and is compounded by the distortion introduced by the wide-angle or “fish-eye” lens at the tip of the bronchoscope . Although lens distortion can be corrected [5–7], the time intensive nature of applying complex algorithms to airway images, as well as the inability to provide real-time measurements, has meant that no user-friendly system has yet been developed or been widely adopted by bronchoscopists. Clinicians, therefore, rely on subjective estimation from bronchoscopic images in order to assess airway sizes during a procedure. There is a need for an endoscopic, real-time imaging modality which allows dynamic measurement of airway dimensions.
Anatomical optical coherence tomography (aOCT) [8,9] is an endoscopic optical modality designed to provide quantitative cross-sectional images of large internal hollow organ anatomy over extended observational periods. It has previously been used in assessment of sleep disorders in the upper airway [10,11], and preliminary work has been presented on its application in the lower airways . It acquires anatomical cross-sectional images of lumens over a range of several centimeters.
This differs from other research with endoscopic subsurface OCT, which has focused on subsurface morphological changes [13–18] or imaging the thickness of the airway wall . In such applications, the systems are characterized by extremely high resolution and an axial scanning range of a few millimeters or less. However, the determination of airway lumen size and shape does not require subsurface imaging and there is a need for a significantly longer scanning range than is typical with subsurface OCT.
In this paper, we will demonstrate the application of aOCT to quantifying the lower airway lumen, demonstrating several physiological features, including 3D structure and dynamic compliance (displacement versus applied pressure) during breathing.
A schematic diagram of the aOCT system is presented in Fig. 1. The system utilizes a frequency-domain optical delay line (FDODL) consisting of a grating, lens and galvanometer mirror in a folded, double-pass arrangement. This configuration allows the system to achieve a delay line length of 36mm, although typically only 25mm is used in the lower airway. A-scans are acquired at a frequency of 500Hz. The light source has a center wavelength of 1310nm, with a bandwidth of 32nm. Details are given in Refs. [8,9,12].
The sample arm utilizes a rotating fiber-optic endoscopic probe for deployment in the airway, encased in a transparent plastic catheter with outer diameter of 2.2mm. It is attached by 1.8m of optical fiber encased in a biplex torque-transmission stainless steel coil. The light beam is focused using a 1.0mm diameter gradient index lens, and redirected at right angles to the probe head with a right-angle prism of 0.7mm width. The probe rotates at approximately 2.5Hz, tracing out an axial image. The rotary joint and probe coupling are mounted on a movable carriage to allow the rotating probe to be translated through a section of airway to obtain a 3D data set.
The aOCT probe has a beam waist of 138µm at a distance of 9mm from the probe head expanding to 249µm at an axial distance of 25mm. Because a rotating probe is used, the transverse resolution is a function of both the distance from the probe and A-scan frequency. The sampling rate of 1.8 degrees per A-scan means the transverse resolution ranges from the 143.5µm per pixel at the probe head, to 0.8mm per pixel at a distance of 25mm.
Our portable aOCT system has been configured for use in the bronchoscopic procedures theatre, and has been used to make measurements on anesthetized subjects undergoing a variety of procedures. The catheter is passed into the lower airway via the working channel of a standard bronchoscope.
During scanning, external analogue respiratory signals are recorded by a lab data acquisition system (ADInstruments, Australia) and integrated into the aOCT data in post-processing. When using the two systems simultaneously, a set of clock and synchronization signals are generated by the aOCT system and used for subsequent synchronization. Examples of analogue physiological parameters that can be measured and synchronized in this way include lung volume, airflow, esophageal pressure and airway pressure.
We report the use of bronchoscopic aOCT to assess both 3D airway anatomy and dynamic movement during quiet breathing. Scans have been performed on over forty patients. In this paper, we illustrate with results from two patients. Informed consent was obtained and the study was approved by the Human Research Ethics Committee of Sir Charles Gairdner Hospital. Bronchoscopic procedures were performed by one of the authors (JPW).
Figure 2 (View 1) shows a typical chest-lung CT used in interventional planning, obtained using a LightSpeed 16 scanner (GE Healthcare, WI, USA) with acquisition settings of 120kV peak voltage and 406mA X-ray tube current. This data is available online (Case 1). The CT was performed at total lung capacity (a breath-hold at full inspiration) and extends from the abdomen up to above the larynx. The trachea, annotated in the image, bifurcates into the left and right main bronchi. The left main bronchus bifurcates into the left upper and left lower lobes and the right divides into three lobes: the right upper, right middle and right lower lobes.
Figure 3 (View 2) demonstrates an intra-operative aOCT scan obtained during bronchoscopy. The data may also be accessed online (Case 2). The catheter, containing the aOCT probe, was inserted through the bronchoscope’s working channel and into the patient’s left lower lobe. The catheter then remained stationary as the probe was rotated and withdrawn through the left main bronchus and into the distal section of the trachea, acquiring a 3D scan of the airway lumen. As shown in the axial view of Fig. 3, the aOCT scan enabled quantification of the lumen diameters at the time of the bronchoscopy.
A strong correlation was observed between CT and aOCT estimates of airway lumen diameters. A representative site in the proximal left main bronchus was selected for the purposes of illustration, with the same anatomical site visually identified for comparison. Using CT, the airway diameter was estimated to be 17.8mm×14.1mm (Fig. 2). In the aOCT scan, the diameter was measured as 17.3mm×13.9mm. Note that with the CT scan, we have used the oblique (not axial) view, so as to orient the measurement perpendicular to the central axis of the airway.
Note also that each axial image in the aOCT scan is centered on the rotating aOCT probe and, hence, is relative to the position of the catheter. The bending of the catheter through the airway is not accounted for in these scans. The images, therefore, accurately represent dimensions of the airway lumen and distances along the lumen, but not the relative position of anatomical features with respect to an external frame of reference. However, for endoscopic interventions, the lumen-specific measurements of aOCT are appropriate. Further, the bending of the catheter within the airway tree ensures that the probe is imaging each airway close to perpendicular thereby providing orthogonal cross-sections of each airway regardless of the orientation of that airway. This reduces the difficulties associated with obtaining an accurate oblique view in standard CT imaging software.
A scan of the proximal section of a lower airway is shown in Fig. 4 (View 3). The data may also be accessed online (Case 3). In this example, the aOCT probe was used to image from the proximal trachea, through the larynx and into the laryngeal mask. The mask is an airway maintenance device that traverses the upper airway to facilitate anesthesia and insertion of the bronchoscope. Note that lumen dimensions are substantially larger in this region, typically twice the diameter of those found in the left and right main bronchi. The aOCT probe is able to image over this large range because of the configuration of the frequency-domain optical delay line, and the long working distance of the probe lens, allowing measurements within a circle of up to 72mm in diameter.
Figures 5–7 demonstrate the ability of aOCT to acquire dynamic measurements of airway lumens. In this study, the aOCT probe was positioned at specific anatomical locations and used to image the same axial plane over several breath cycles. Simultaneous measurements were recorded of both air pressure and flow at the laryngeal mask and synchronized with the aOCT. Figure 5 shows respiratory movement of the airway in the distal trachea. Figure 6 illustrates this dynamic respiratory movement in the right main bronchus. In Fig. 7, the aOCT probe was positioned in the right middle lobe and demonstrates the ability of the probe to quantify movement in smaller airways. Three illustrative images from the breath cycle have been superimposed for each location, and the position in the breath cycle is shown relative to airflow, airway pressure and airway lumen area measured from the aOCT images.
We observed that airway caliber varies with breathing in a complex way. Intra-thoracic airways tend to increase in size with increasing lung volume, as occurs during inspiration, reaching a maximal cross-sectional area at end inspiration (when lung volume is greatest). The opposite occurs in the upper airway and the extra-thoracic portions of the lower airway (e.g. the proximal trachea). The degree of caliber variation depends on several factors including airway size, degree of respiratory effort, as well as the presence of underlying airway diseases which may affect airway compliance such as asthma and bronchiectasis.
Results presented here demonstrate the ability of aOCT to quantify airway lumen dimensions during clinical intervention and assess dynamic change due to factors such as respiratory motion. Several artifacts were observed during imaging. In some axial slices, the airway did not appear as a continuous contour. We have identified two mechanisms for this artifact.
Like ultrasound, aOCT is an imaging modality based on detection of signal reflections. Both imaging modalities are subject to shadowing effects. In ultrasound, bone will shadow the signal, obscuring imaging of any distal structures. In aOCT, the limited penetration depth of the light can cause the airway to shadow itself. This will occur in airway sections that have a complex, non-concave structure in which part of the airway obscures the line of sight from the probe. Catheter placement within the airway will have a significant impact on this. However, fine manipulation of the catheter position is often difficult in the lower airway.
Secondly, a buildup of airway secretions or blood on the surface of the airway mucosa may affect the back-reflection properties of the air-tissue interface. This can result in a loss of aOCT signal. It is a standard procedure to flush the airway wall with 0.9% saline during a procedure if a buildup of secretions or blood is noted. Our experience is that this results in improved aOCT signal. Note that our use of a stationary catheter to enclose the rotating probe avoids additional stimulation of the airway mucosa and underlying smooth muscle.
Breathing artifact is evident in the 3D image acquisition of the airway. This is most evident in the coronal and sagittal images shown in this paper. Breathing artifact is manifested as regular, jagged edges in the transverse images of the airway wall. As the probe is slowly translated during a 3D scan, the airway cross-sectional-area varies with the phase of respiration. We have found that the airway wall does not move evenly over the entire breath cycle, but rather the majority of expansion or contraction occurs over a small part of the overall breath. This results in rapid changes in the location of the air-tissue interface, giving rise to the jagged features. As demonstrated in Figs. 5–7, detailed quantification of this movement is possible by acquiring images in the same axial plane over several breath cycles.
Through standard video bronchoscopy, we also observed that longitudinal movement may occur in the airway walls during breathing. Unless the aOCT catheter moves with the airway, the cross-sectional nature of the aOCT scanning probe does not currently allow for correction of images produced in the presence of this movement. However, because of the gradual nature of airway tapering, this small movement artifact does not appear to have a significant impact on clinical measurements.
This paper has described the application of aOCT as a tool to image the human lower airway and quantify its internal dimensions. Unlike standard video bronchoscopy, aOCT allows accurate quantification in three dimensions and, in contrast to other structural imaging modalities, we have demonstrated that it may be performed during clinical interventions. Results have shown both the application of aOCT to acquire 3D images as well as dynamically assess respiratory motion. This tool is currently being utilized in an ongoing study to characterize lower airway compliance in a number of airway diseases, including asthma, bronchiectasis and chronic obstructive pulmonary disease. An improved understanding of lower airway characteristics in the presence of such diseases presents new possibilities in the assessment of disease and its changes with time and treatment.
References and links
1. D. A. Yablonskiy, A. L. Sukstanskii, J. C. Leawoods, D. S. Gierada, G. L. Bretthorst, S. S. Lefrak, J. D. Cooper, and M. S. Conradi, “Quantitative in vivo assessment of lung microstructure at the alveolar level with hyperpolarized He-3 diffusion MRI,” Proc. Natl. Acad. Sci. USA 99, 3111–3116 (2002). [CrossRef] [PubMed]
3. N. Jowett, R. A. Weersink, K. Zhang, P. Campisi, and V. Forte, “Airway luminal diameter and shape measurement by means of an intraluminal fiberoptic probe: a bench model,” Arch. Otolaryngol. 134, 637–642 (2008). [CrossRef]
4. I. B. Masters, M. M. Eastburn, P. W. Francis, R. Wootton, P. V. Zimmerman, R. S. Ware, and A. B. Chang, “Quantification of the magnification and distortion effects of a pediatric flexible video-bronchoscope,” Respir. Res. 6, 14–16 (2005). [CrossRef]
5. P. Czaja, J. Soja, P. Grzanka, A. Cmiel, A. Szczeklik, and K. Sladek, “Assessment of airway caliber in quantitative videobronchoscopy,” Respiration 74, 432–438 (2007). [CrossRef]
6. W. V. Dorffel, I. Fietze, D. Hentschel, J. Liebetruth, Y. Ruckert, P. Rogalla, K. D. Wernecke, G. Baumann, and C. Witt, “A new bronchoscopic method to measure airway size,” Eur. Respir. J. 14, 783–788 (1999). [CrossRef] [PubMed]
7. P. K. McFawn, L. Forkert, and J. T. Fisher, “A new method to perform quantitative measurement of bronchoscopic images,” Eur. Respir. J. 18, 817–826 (2001). [CrossRef]
8. J. J. Armstrong, M. S. Leigh, I. D. Walton, A. V. Zvyagin, S. A. Alexandrov, S. Schwer, D. D. Sampson, D. R. Hillman, and P. R. Eastwood, “In vivo size and shape measurement of the human upper airway using endoscopic long-range optical coherence tomography,” Opt. Express 11, 1817–1826 (2003). [CrossRef] [PubMed]
9. M. S. Leigh, J. J. Armstrong, A. Paduch, J. H. Walsh, D. R. Hillman, P. R. Eastwood, and D. D. Sampson, “Anatomical optical coherence tomography for long-term, portable, quantitative endoscopy,” IEEE Trans Biomed. Eng. 55, 1438–1446 (2008). [CrossRef] [PubMed]
10. J. J. Armstrong, M. S. Leigh, D. D. Sampson, J. H. Walsh, D. R. Hillman, and P. R. Eastwood, “Quantitative upper airway imaging with anatomic optical coherence tomography,” Am. J. Respir. Crit. Care 173, 226–323 (2006). [CrossRef]
11. J. H. Walsh, M. S. Leigh, A. Paduch, K. J. Maddison, D. L. Philippe, J. J. Armstrong, D. D. Sampson, D. R. Hillman, and P. R. Eastwood, “Evaluation of pharyngeal shape and size using anatomical optical coherence tomography in individuals with and without obstructive sleep apnoea,” J. Sleep Res. 17, 230–238 (2008). [CrossRef] [PubMed]
12. J. J. Armstrong, S. Becker, R. A. McLaughlin, M. S. Leigh, J. H. Walsh, D. R. Hillman, P. R. Eastwood, and D. D. Sampson, “Anatomical optical coherence tomography - a safe and effective tool for quantitative long-term monitoring of upper airway size and shape,” Proc. SPIE 6842, 68421N (2008).
13. 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, 1640–1644 (1998). [PubMed]
14. F. I. Feldchtein, G. V. Gelikonov, V. M. Gelikonov, R. V. Kuranov, A. M. Sergeev, N. D. Gladkova, A. V. Shakhov, N. M. Shakhov, L. B. Shakhova, A. B. Snopova, E. V. Terent’eva, Yu. P. Zagainova, I. A. Chumakov, and Kuznetzova, “Endoscopic applications of optical coherence tomography,” Opt. Express 3, 257–270 (1998). [CrossRef] [PubMed]
15. S. Han, N. H. El-Abbadi, N. Hanna, U. Mahmood, R. Mina-Araghi, W. Jung, Z. Chen, H. Colt, and M. Brenner, “Evaluation of tracheal imaging by optical coherence tomography,” Respiration 72, 537–541 (2005). [CrossRef] [PubMed]
16. N. Hanna, D. Saltzman, D. Mukai, Z. Chen, S. Sasse, J. Milliken, S. Guo, W. Jung, H. Colt, and M. Brenner, “Two-dimensional and 3-dimensional optical coherence tomographic imaging of the airway, lung, and pleura,” J. Thorac. Cardiovasc. Surg. 129, 615–622 (2005). [CrossRef] [PubMed]
17. S. C. Whiteman, Y. Yang, D. G. van Pittius, M. Stephens, J. Parmer, and M. A. Spiteri, “Optical coherence tomography: real-time imaging of bronchial airways microstructure and detection of inflammatory/neoplastic morphologic changes,” Clin. Cancer Res. 12, 813–818 (2006). [CrossRef] [PubMed]
18. 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, 030506 (2008). [CrossRef] [PubMed]
19. 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 177, 1201–1206 (2008). [CrossRef]