In situ expansion capacity evaluation of terminal alveoli in chronic obstructive pulmonary disease using an ultrathin optical coherence tomography catheter

Abstract

Morphological structural changes in the peripheral alveoli serve as critical indicators for the early pathological features of chronic obstructive pulmonary disease (COPD). Endoscopic optical coherence tomography (OCT) has emerged as a valuable diagnostic tool for COPD, enabling the assessment of airway mucosa, smooth muscle thickness, lumen diameter, lumen area, and airway wall area. However, conventional transbronchial OCT imaging catheters are limited in size, restricting their access to terminal bronchioles and thereby impeding the morphological and structural analysis of peripheral alveoli. In this study, we present a bronchial ultrathin OCT (Bronchial-uOCT) imaging catheter with an outer diameter of 380 µm, an axial resolution of 8.4 µm, and a lateral resolution of 9.3 µm. Using New Zealand rabbit COPD models, we achieved real-time in situ 3D volumetric imaging of terminal alveoli. Statistical analysis revealed a difference of 5 times in the alveolar area between normal and COPD pathological groups (P = 0.00018). Quantitative analysis of expansion compliance revealed significant differences in the growth rate and trend of the alveolar area between the normal group and the COPD pathology group. Furthermore, the opening direction of the curve obtained through power law fitting may help diagnose COPD. These findings highlight bronchial-uOCT as a clinically significant approach capable of in situ visualization of alveolar structures in the terminal bronchioles, offering a promising approach for the early diagnosis and assessment of COPD.

1. Introduction

Small airway illnesses, such as chronic obstructive pulmonary disease (COPD), pose considerable public health issues worldwide. A prominent pathological feature of these early-stage diseases is the morphological and structural alteration of alveoli [1]. Chronic pulmonary obstruction is characterized by the disruption of airway and alveolar structures resulting from decreased elastin and increased collagen, which causes airway narrowing and thickening of alveolar walls [2,3]. Current diagnostic instruments for COPD encompass multiple modalities: Transbronchial ultrasound (EBUS) enhances the physician's visualization of peripheral airways and facilitates real-time imaging of peripheral lymph nodes. However, the resolution of EBUS is insufficient for identifying airway wall structures or alveoli [4]. Bronchoscopy allows direct visualization and targeted sampling of airway lesions in COPD, aiding in phenotyping and pathogen detection. However, its invasiveness, requirement for sedation or general anesthesia, and limited access to distal airways restrict routine use [5,6]. MicroCT and MRI: These modalities evaluate COPD by analyzing alveolar density and structural abnormalities of the airways. MicroCT, although effective, entails radiation exposure and poses possible risks, whereas MRI, despite being noninvasive, is hindered by poor resolution, rendering it insufficient for comprehensive investigation of alveolar and airway mucosa [7–11]. Fluorescence endomicroscopy: this technique can be used to evaluate airway and alveolar structures.Although fluorescence imaging provides high contrast, its limited imaging depth and restricted forward-viewing field hinder accurate assessment of peripheral airway and alveolar structures [12].

These limits highlight the need for advanced imaging technologies capable of providing detailed, real-time visualization of alveolar and airway structures to improve the diagnosis and treatment of small airway illnesses. Endoscopic optical coherence tomography (EOCT) is a non-invasive imaging technique that utilizes a fiber-optic catheter to examine luminal organs in vivo [11–15]. Employing helical scanning of reflected low-coherence light, EOCT enables real-time visualization of tissue microstructures [16]. This technology offers advantages over other in vivo imaging methods, including high sensitivity, dynamic range, resolution, and millimeter-scale imaging depth [17,18]. EOCT has been widely utilized in numerous medical domains, including endovascular [19–22], gastrointestinal [23,24], respiratory [25–29], and reproductive systems [30–32].

In the diagnosis and therapy of chronic obstructive pulmonary disease (COPD), endobronchial optical coherence tomography (EB-OCT) largely concentrates on assessing airway morphology and structure. This encompasses assessments of airway mucosa and cartilage thickness, luminal diameter, inner luminal area, and airway wall area [25–29]. Optical coherence tomography (OCT) and confocal laser endomicroscopy (CLE, also referred to as fibered confocal laser scanning microscopy, CLSM, in certain studies [33]) have emerged as promising tools for visualizing alveolar microstructures in interstitial and acute lung diseases. Previous studies using needle-based or surface-contact OCT techniques enabled static or dynamic imaging of alveoli but were limited by invasiveness or inability to reach deep distal airways. CLE provides cellular-level insights but lacks sufficient structural depth. Recent bronchoscopic approaches have introduced real-time OCT/CLE imaging in vivo; however, these still face challenges in resolution, accessibility, and comprehensive 3D visualization of peripheral alveoli through the terminal bronchioles [33–36]. This constraint occurs due to the majority of alveoli being located at the periphery of terminal bronchioles, which possess an average lumen diameter of roughly 424 µm [3]. Traditional endobronchial diagnostic methods and OCT catheters are constrained by size, preventing access to these regions for real-time optical biopsy of peripheral alveoli. Thus, this obstructs the evaluation of the correlation between alveolar architecture and COPD.

To address these challenges, a bronchial ultrathin optical coherence tomography (Bronchial-uOCT) imaging catheter was developed, featuring a maximum outer diameter of 380 µm. To the best of our knowledge, this represents the thinnest endoscopic OCT catheter currently available for three-dimensional helical scanning imaging applications [37]. Performance testing within the swept-source OCT system revealed an axial resolution of 8.4µm in air and a focused spot size of 9.3µm. A subsequent structural similarity analysis of adjacent frames was conducted to assess the stability of real-time imaging with the Bronchial-uOCT catheter inside the airway [38–49]. Furthermore, the quantitative and qualitative evaluation capabilities of the Bronchial-uOCT for COPD were validated by imaging alveoli in both the COPD and healthy New Zealand rabbit groups. Meanwhile, we investigated the differences in alveolar expansion compliance between the COPD and control groups. This advanced imaging catheter technology, with its ability to perform in situ alveolar imaging in terminal bronchioles, has the potential to advance the assessment of COPD. By enabling quantitative evaluation of alveolar area and expansion compliance, it offers a novel approach to evaluating chronic airway diseases. This innovation is expected to refine diagnostic accuracy, facilitate early intervention, and improve treatment planning and prognosis, positioning it as a promising tool for lung disease research and clinical practice.

2. Materials and methods

2.1. System and catheter

The bronchial OCT system consists of two fundamental components: an optical channel and a Bronchial-uOCT imaging catheter. The optical route includes a SS-OCT laser (1310 nm, AXP50124-8, Axsun, USA) featuring a bandwidth of 100 nm and a repetition rate of 50 kHz, linked with a balanced detector (Optowaves, Inc., USA). The basic component of the Bronchial-uOCT imaging catheter is a fiber-optic ball lens produced by ShenZhen Photostream Limited Co., located in Shenzhen, China. Supplementary components consist of a torsion guide wire (Changzhou Neorays Medical Technology Co., Ltd., Changzhou, 213000, China) and an optically clear protective sheath (inner and outer diameters of 300/380 µm, PA12 material, Haiwinmed Medical Technology Co., Ltd., Suzhou, China).

2.2. Animal sources, modeling, and ethical approval

New Zealand white rabbits, aged 12 to 17 weeks and weighing between 2.0 and 2.5 kg, were acquired from the Guangdong Animal Experiment Center. The enzyme papain (Shanghai Macklin Biochemical Technology Co., Ltd., Shanghai, China; enzyme activity > 200 U/kg; lot number C16575609) was utilized to induce COPD. The experimental group received a nebulized 5% papain solution (5 ml/kg body weight) weekly for six weeks. We maintained the control group under standard conditions. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of South China Normal University (approval number: SCNU-BIP-2024-049).

2.3. Optical path system for performance test

A custom-built setup was used to measure the lateral resolution and beam profile of the beam emitted by the Bronchial-uOCT imaging probe. The light from the SS-OCT laser was introduced into the developed Bronchial-uOCT imaging catheter. The focused beam, which passed through the imaging probe on the distal end of the imaging catheter, was sequentially passed through a 40x flat-field achromatic objective lens (PLN40X) and a sleeve lens (BTL180-A) with a focal length of 180 mm. It was then received by a spot analyzer (CinCam CMOS, Cinogy), which captured the image of the output beam and extracted the beam patterns at different axial distances by axially shifting the optical path.

2.4. Alveolar segmentation and quantification based on OCT image calibration

The OCT system employed in this study has a spatial calibration of 5 µm per pixel in the axial direction. Based on this calibration, the original OCT images were imported into FIJI ImageJ, where image scaling was applied accordingly. Alveolar segmentation was performed using the Trainable Weka Segmentation plug-in, a machine learning-based tool, enabling automatic classification of image regions into alveolar (positive) and non-alveolar (negative) categories.

The trained model selectively identified alveolar structures, allowing for quantitative analysis of their area across multiple inflation stages. Since optical coherence tomography measures optical path length rather than geometric dimensions, and no refractive index correction was applied during image reconstruction, the measured alveolar areas were expressed in arbitrary units (a.u.), rather than in absolute physical units (e.g., µm²). This is because accurate refractive index correction in heterogeneous media such as air–tissue interfaces remain technically challenging. Nevertheless, this approach enables reliable assessment of relative changes in alveolar expansion under different inflation volumes.

2.5. Imaging catheter consistency test

The Bronchial-uOCT imaging catheter must be navigated by an electronic bronchoscope via airways with diverse bending radius to access various regions of the lung bronchial tree. This entails integrating the distal outer diameter of the electronic bronchoscope (2.8 mm) with the diminutive airways in the lungs (diameter < 2 mm). The minimum bending radius between the proximal and distal tubes is primarily above 10 mm. A minimum bending radius of 10 mm was established to facilitate the active bending of the bronchoscope. Several clinical electronic bronchoscopes are presently available, including the Olympus Bronchoscope BF-XP 190 (Olympus Corporation, Japan) and the FUJIFILM Healthcare Slim Bronchoscope EB-530P (FUJIFILM, Japan). These devices possess maximum bending angles of 130°, 180°, and 210° in both upward and downward orientations, respectively. Thus, the SSIM (structural similarity index measure) of the images at the three maximum bending angles was assessed at bending radius of 20 mm, 15 mm, and 10 mm to evaluate the real-time imaging efficacy of the catheter.

2.6. Imaging principle and system

Figure 1(a) presents a schematic illustration of real-time endoscopic imaging with Bronchial-uOCT catheter. The imaging catheter progressively advances to the terminal bronchioles from the primary airway. The catheter performs in situ real-time three-dimensional volumetric imaging of the terminal bronchioles and alveoli by helical pullback 3D scanning. Figure 1(b) illustrates an optical coherence tomography (OCT) imaging system employed for Bronchial-uOCT endoscopy. The system employs a swept laser that operates at a repetition frequency of 50 kHz, featuring a central wavelength of 1310 nm and a bandwidth of 100 nm as the excitation light source. The emitted light from the source is channeled through a 10/90 beam splitter, with 10% allocated to an optical reference arm and 90% to an optical sample arm, which is the Bronchial-uOCT catheter.

Fig. 1.

Schematic diagram of the Bronchial-uOCT imaging catheter and system for terminal bronchiolar and alveolar imaging. (a) Real-time OCT imaging of the catheter in the terminal bronchioles. (b) Configuration of the SS-OCT endoscopic imaging system. (c) The overall structure of the catheter. Inset: A magnified view of the distal end of the catheter, framed by a yellow dashed line. In the figure, the scale bar is 500 µm long. TB, terminal bronchioles; APV, airway peripheral vessels; HPSL, helical pullback scanning laser; VOA, variable optical attenuator.

Ultimately, the two separate light beams from the endoscope and optical reference arm converge via the ring splitter and are transmitted to the detector, where they interact and are transformed into an electrical signal. Following signal processing and image reconstruction, the resultant OCT images can be obtained. Figure 1(c) presents a comprehensive illustration of the Bronchial-uOCT endoscopic imaging catheter. The apparatus consists of an optical interface, a rigid element, and a flexible segment. The inset presents an expanded illustration of the distal end of the Bronchial-uOCT catheter, indicated by the yellow dashed line. The overall external diameter for the three locations is 380 µm. The angularly polished fiber ball lens achieves simultaneous proximal and distal rotation through a torque coil. The outermost transparent plastic layer provides protection.

2.7. Design and fabrication of the Bronchial-uOCT imaging catheter

Figure 2(a) presents the transverse plan view (i), illustrating the optical path at the catheter end. The 1310 nm swept light is transmitted from a single-mode fiber to a coreless fiber. After the necessary beam expansion, the beam is focused and experiences total internal reflection through a fiber optic ball lens featuring a 40° reflective surface. The 40° reflection angle was chosen to balance imaging sensitivity and contrast. However, since the critical angle for total internal reflection in silica (n = 1.45) is approximately 43.6°, the 40° angle does not satisfy the TIR condition. Therefore, a thin metallic coating was applied to the side-polished surface of the ball lens to ensure efficient reflection and prevent energy loss. The axial plan view (ii) illustrates the structural dimensions of the catheter, which includes a 200 µm fiber optic ball lens situated within a torque coil. The torque coil is encased in a plastic sheath with a maximum outer diameter of 380 µm. The terminal bronchioles have dimensions under 500 µm, with a mean airway diameter of 424 µm. Consequently, a working distance of 360 µm has been established, adequately accommodating the airway size and facilitating the projection of the focused beam into the peripheral alveoli.

Fig. 2.

Design and fabrication of the Bronchial-uOCT imaging catheter. (a) Design diagram of the Bronchial-uOCT catheter: (i) Radial plane of the catheter; SMF: single-mode fiber, NCF: coreless fiber; (ii) Axial plane of the catheter; MOD: maximum outer diameter, ID: inner diameter, WD: working distance; (b) The relationship between the working distance, the focused beam spot size, and the coreless fiber length is calculated; (c) The focal depth spot size is calculated; (d) Fabrication of the catheter: (i) Fabrication of the bare fiber ball lens, (ii) angle polishing and metal coating, and (iii) assembly of the complete catheter.

Figure 2(b) presents the presence of two design parameters: the length of the coreless fiber and the diameter of the fiber ball lens. The selection of an appropriate coreless fiber length and ball lens diameter enables the determination of optimal working distance and resolution. Optical simulations are performed in Zemax OpticStudio, using SILICA as the material for the coreless fiber, consistent with the material employed in the actual production process. The outer diameter of the ball lens is fixed at 200 µm. Under these conditions, the relationship between the length of the coreless fiber and the working distance and focused spot size is investigated. When the length of the coreless fiber is in the range of 200 to 400 µm, the focusing spot size and working distance are inversely proportional to the length of the coreless fiber. The optimal length of the coreless fiber was determined to be 360 µm to achieve equilibrium between the working distance and spot size of the design.

Figure 2(c) presents the lateral resolution of the designed fiber optic ball lens within the depth of focus, measured at approximately 200 µm. The fabrication process of the Bronchial-uOCT imaging Catheter consists of three main stages, as depicted in Fig. 2(d). Figure 2(d-i) shows the complete fiber optic ball lens in its actual form. Figure 2(d-ii) presents the lens after angular polishing and coating. Figure 2(d-iii) presents the comprehensive imaging catheter with an outer diameter of 380 µm, which includes a torque guide wire and a protective sheath tube. Upon completing these three steps, the outcome will be a monolithic ultrathin imaging catheter composed entirely of fiber, enclosed within an optically transparent sheath tube.

2.8. Experimental protocol for the evaluation of alveolar expansion compliance

Alveoli, as soft tissues, undergo expansion during the process of gas exchange. In this study, the trend of alveolar area variation across different inflation stages is defined as alveolar expansion compliance, which serves as a metric for evaluating functional differences in alveoli between the pathological and normal groups. The experiment was carried out on two groups of New Zealand rabbits: a COPD pathological group and a normal group, both raised under identical conditions. Pulmonary lungs from both groups were collected, and air was introduced into the lungs via the main bronchus using a custom-designed air inflator. The inflation device consists of two distinct channels: one for air and the other for the imaging catheter. The experimental process is shown in Fig. 6(a). The inflation was performed in six stages: S1: 0 ml, S2: 0.5 ml, S3: 1.0 ml, S4: 1.5 ml, S5: 2.0 ml, and S6: 2.5 ml, with the air volume progressively increasing at each stage. Throughout the inflation process, real-time in situ imaging of the alveoli surrounding the airways was conducted using a Bronchial-uOCT imaging catheter.

Fig. 6.

Evaluation of Airway Peripheral Alveolar Expansion Compliance in COPD Pathological and Normal Groups. (a) Schematic representation of the experimental setup for assessing alveolar elasticity. Advance: Advancement of air into the lungs. Insert: Insertion of the bronchial-uOCT imaging catheter into the catheter channel. (b) Real-time in situ OCT alveolar imaging results for the COPD pathological group, corresponding to the six stages of inflation. (c) Machine learning-based alveolar segmentation results for the COPD pathological group, corresponding to the results in (b). (d) Real-time in situ OCT alveolar imaging results for the normal group, corresponding to the six stages of inflation. (e) Machine learning-based alveolar segmentation results for the normal group, corresponding to the results in (d). (f) Relationship between alveolar area and expansion compliance with inflation volume. (g) Statistical analysis of alveolar expansion compliance based on power law fitting. Scale bars are 300 µm in all figures.

3. Results

3.1. Characterization of the performance of the Bronchial-uOCT imaging catheter

The efficacy of the Bronchial-uOCT imaging catheter was assessed alongside the SS-OCT endoscopic imaging system, as shown in Fig. 1(b). Figure 3(a) presents the axial resolution measurements acquired along the imaging depth with the catheter sheath in place. The axial resolution remained consistently high, below 9.5 µm, throughout the imaging depth of 1 mm. The inset presents the point spread function at a representative axial resolution. Figure 3(b) presents the lateral resolution data within the depth of focus of the focused beam, as recorded by the CCD camera. The lateral resolution in the major and minor axis directions is comparable, the dispersion ratio is low, and the depth of focus is approximately 200 µm. A study on production scalability was undertaken to further test the reliability between design and production. Figure 3(c) presents a clear comparison between the empirically measured values and the simulated values. Given a ball lens diameter of 200 µm, two class fiber ball lenses with coreless fiber lengths of 250 µm and 360 µm (n = 3) were developed and fabricated. The working distance and focus spot size of the two fiber ball lenses were concurrently tested, with each parameter assessed three times. Data analysis reveals that the simulated values of working distance and lateral resolution closely approximate the real measured values. This part of the work fully demonstrates the reliability of the catheter design scheme.

Fig. 3.

Performance characterization of the Bronchial-uOCT imaging catheter. (a) Axial resolution measured along the imaging depth, with the inset showing a representative point spread function (PSF); (b) Measurement of the focal depth spot size; (c) Conformance testing between the design and production; (d) Actual image of the spot size within the focal depth, corresponding to the measurement in (b), with a scale bar of 10 µm; (e-g) Evaluation of the structural similarity of adjacent image frames; The inset shows an image of a multilayer coverslip, a schematic diagram of the measurement at different bending radii and bending angles. WD,working distance; LR, lateral resolution; SSIM, structural similarity index measure.

As shown in Fig. 3(e-g), we combined the minimum bending radius of the proximal and distal segments of the small airways in the human lung with the three maximum bending angles of the electronic bronchoscope (130°, 180°, and 210°). The structural similarity index (SSIM) of consecutive frame pictures was assessed at an imaging rate of 25 Fps. This was utilized to assess the imaging stability of the Bronchial-uOCT imaging catheter. Figure 3(e) presents the real-time imaging of multilayer slides exhibiting a bending radius of 20 mm and three distinct bending angles (130°, 180°, and 210°). 50 adjacent Bscans were chosen for the structural similarity assessment of OCT images, with results indicating that the SSIM values were predominantly above 0.9. Figures 3(f) and (g) present the SSIM evaluation for three angles, corresponding to bending radii of 15 mm and 10 mm, respectively. The findings indicate values exceeding 0.9. This work demonstrates that the Bronchial-uOCT imaging catheter is suitable in clinical respiratory diagnostics. It facilitates access to the narrow terminal bronchioles in clinical practice.

3.2. Ex vivo imaging evaluation of COPD pathology models

This study aimed to demonstrate the capability of the Bronchial-uOCT imaging catheter in diagnosing COPD pathology in the lungs. Consequently, we conducted further sophisticated imaging in a New Zealand rabbit model. The results underwent analysis to enable a comparison of the morphological differences in the terminal bronchioles and peripheral alveoli between the normal and COPD pathology groups. During this phase of the study, we carefully advanced the Bronchial-uOCT imaging catheter into the terminal bronchioles of ex vivo rabbit lungs, enabling real-time, in situ imaging. Figure 4(a, b) presents three-dimensional volumetric imaging of the terminal bronchioles and peripheral alveoli in both the normal and COPD pathology groups.

Fig. 4.

Bronchial-uOCT imaging of terminal bronchioles and peripheral alveoli in live New Zealand rabbits from the COPD pathological group and the normal group. (a) 3D OCT cross-sectional view in the normal group; (b) 3D OCT cross-sectional view in the COPD group; (c, d) Comparison of bronchiolar OCT imaging between the normal and COPD groups; (e, f) Comparison of alveolar OCT imaging between the normal and COPD groups; (g, h) HE-stained histological sections corresponding to (c, d); (i, j) HE-stained histological sections corresponding to (e, f). BV, blood vessel; S, smooth muscle; LP, lamina propria; EP, epithelium; AEI, airway epithelial injury. Scale bars = 300 µm.

The OCT 3D transverse section view facilitates clear visualization of the complex airway wall structure and the peripheral alveolar architecture. A comparison between the two groups indicated that the quantity of alveoli in the periphery of the terminal bronchioles was higher in the normal group compared to the COPD pathology group, which is consistent with the expected pathological outcome. The bronchioles in the COPD pathology group demonstrated significant narrowing. Pathologic verification of COPD in New Zealand rabbits was established through the identification of abnormalities in the airway wall and associated morphologic structures. The 2D OCT images corresponding to the two positions marked by red arrows (c and d) in Fig. 4(a, b) were chosen for comparison. A comparison between the normal and COPD pathology groups demonstrated significant stenosis and folds in the latter, as shown in Fig. 4(d), with white hexagonal asterisks indicating the two locations. An airway epithelial injury is observable concurrently. Two-dimensional OCT images corresponding to the positions marked by the red arrows (e and f) in Fig. 4(a, b) were selected for analysis. The images indicate that the number of peripheral alveoli in the normal group is higher than that in the COPD pathological group. In conjunction with the corresponding physiological sections in Fig. 4(g, h, i, j), our findings indicated a strong similarity between the fine microstructure of these airway peripheral alveoli and the bronchioles and the HE histologic sections.

3.3. Quantitative statistical analysis of airway peripheral alveoli

To clarify the relationship between OCT imaging of alveolar morphological structures and COPD pathology, we imaged the upper lobe terminal bronchioles and the lower lobe terminal bronchioles of New Zealand rabbits. We assessed COPD by quantifying the alveolar areas in both the normal and pathology groups for statistical comparison.

Figure 5(a, e) presents the 2D OCT image results for the upper lobe of the lungs in both the bronchial normal group and the COPD pathology group. Figure 5(b, f) presents the alveolar structures identified and extracted from the original image through processing with FIJI ImageJ. Figure 5(c, g) presents the 2D OCT imaging results for the normal and COPD pathology groups of the terminal bronchioles in the lower lobes of the lungs, respectively. Figure 5(d, h) demonstrates the identification of isolated filled alveolar structures. Ten non-adjacent B-scan images were selected, and the analysis was carried out following the approach outlined in Section 2.4 of the Materials and Methods. Figure 5(i, j) compares the area data for the bronchial peripheral alveoli in the upper and lower lobes. The blue bars represent the statistical mean of the alveolar area in the control group, while the dark blue bars represent the statistical mean of the alveolar area in the COPD pathological group. Statistical analysis revealed that the alveolar area in the upper lobe terminal bronchioles of the normal group was approximately 4.6-fold higher than that in the COPD pathology group, while the difference in the lower lobe reached 5-fold (P = 0.00018 for both comparisons, Mann–Whitney U test, n = 10 per group).

Fig. 5.

Comparison of alveolar area around terminal bronchioles between normal group and COPD pathological group. (a, e) OCT images of alveoli around terminal bronchioles in the lower lobe. (b, f) Alveolar images extracted separately from (a) and (e). (c, g) OCT images of alveoli around terminal bronchioles in the upper lobe. (d, h) Alveolar images extracted separately from (c) and (g). (i, j) Statistical comparison of alveolar area in upper and lower lobes between normal group and COPD pathological group (n = 10). Statistical analysis was performed using the Mann–Whitney U test.P = 0.00018 was considered statistically significant. Scale bars are 300 µm in all figures.

3.4. Quantitative statistical analysis of expansion compliance in peripheral airway alveoli

As shown in Fig. 6(b) i-vi, real-time in situ OCT imaging results of the peripheral alveoli in the airways across the six stages of air inflation are presented. As shown in Fig. 6(c) i-vi, to facilitate intuitive analysis, the alveolar components were separately extracted from the OCT images in Fig. 6(b) i-vi using machine learning. The qualitative experimental results of the COPD pathology group clearly reveal a trend in the change of alveolar area. At the same time, we conducted a control experiment using the normal group as the control, as shown in Figs. 6(d) and 6(e) i-vi. As shown in Figs. 6(d) and (e) i-vi, during the experiment in the normal group, the real-time in situ OCT alveolar imaging results align with the corresponding machine-learning-based alveolar segmentation results. Qualitative analysis reveals that the increase in alveolar area with inflation is more rapid in the normal group compared to the COPD pathology group. Furthermore, based on the qualitative results, we performed quantitative statistical analysis, as shown in Fig. 6(f). The results showed that during the air inflation process, alveolar expansion in the COPD group quickly reached a plateau. Although the alveolar area increased sharply during the S1-S3 stages, it showed minimal change after the S4-S6 stages. In the normal group, the alveolar area increased rapidly with the rising inflation volume. It expanded significantly from S1 (24 k a.u.) to S6 (350 k a.u.). This steady, continuous increase suggests that the alveoli in the normal group have greater expansion compliance and can effectively adapt to the increasing inflation volume. The real-time in situ imaging process is presented in Visualization 1 ^{(10.4MB, mp4)} .

As shown in Fig. 6(g), to more accurately characterize the changes in alveolar expansion compliance between the COPD pathology group and the normal group with increasing inflation volume, we performed power law fitting on the data. The fitting result for the COPD pathological group is represented by [Eq. (1)], and that for the normal group by [Eq. (2)]:

$$y=14925.66\ast x^{0.47}$$ (1)

$$y=7959.13\ast x^{2.12}$$ (2)

At the same time, the R² values for the two data sets obtained through power law fitting are 0.71 and 0.99, respectively, indicating that the fitting model has strong explanatory power. Further derivation of the two fitted functions yields the second-order derivative expressions, presented as [Eq. (3)] for the pathological group and [Eq. (4)] for the normal group:

$$\frac{d^{2}y}{d{x}^2}=-14925.66\ast 0.47\ast 0.53\ast x^{-1.53}$$ (3)

$$\frac{d^{2}y}{d{x}^2}=7959.13\ast 2.12\ast 1.12\ast x^{0.12}$$ (4)

In the COPD pathological group, [Eq. (3)] indicates that the exponent of the second-order derivative of the power law function is −1.53. According to the characteristics of power law functions, this implies that the fitted curve opens downward. Furthermore, as shown in [Eq. (1)], the exponent of the power law function is 0.47, indicating a gradually decreasing rate of alveolar area expansion within this dataset, which reflects impaired alveolar expansion compliance. In the normal group, [Eq. (4)] shows that the exponent of the second-order derivative of the power law function is 0.12. According to the properties of power law functions, this indicates that the fitted curve opens upward. Moreover, as shown in [Eq. (2)], the exponent of the power law function is 2.12, suggesting that the alveolar area expansion rate continuously accelerates within this dataset, reflecting favorable alveolar expansion compliance. Overall, there is a significant difference in the alveolar area expansion compliance between the COPD pathology group and the normal group. The opening direction of the curve obtained through power law fitting may help diagnose COPD. In the COPD group, the increase in alveolar area declined from S2 to S6, indicating a reduction in the compliance of alveolar expansion. This may be attributed to the decreased elasticity, functional deterioration, and airway obstruction of lung tissue, all of which contribute to the decline in lung function associated with the disease.

4. Discussion and conclusion

To address the limitations of current clinical imaging techniques, which are unable to access the terminal bronchioles for high-resolution and real-time in situ imaging, we propose the smallest known OCT imaging catheter globally, with an overall maximum outer diameter of 380 µm. This design was achieved through advanced optical and mechanical considerations. Comprehensive quantitative performance evaluations and stability assessments confirmed that the catheter is capable of high-speed, high-resolution, and quantitative imaging of peripheral alveoli in vivo. The Bronchial-uOCT imaging catheter enables detection of high-resolution structures in the terminal bronchioles and peri-airway (surrounding the airway) alveoli, including the fine structure of the airway wall, alveolar contours, and alveolar walls. In pathological examinations of COPD, comparative analysis of the alveolar area surrounding the airway revealed substantial differences, with the normal group exhibiting values exceeding fourfold those observed in the pathological group. The assessment of alveolar expansion compliance provides valuable insights into the lung's ability to adapt to increasing inflation volumes, revealing differences in compliance between normal and pathological groups. These studies enhance diagnostic accuracy for chronic airway diseases and offer promise for earlier intervention.

In traditional spirometry, pulmonary compliance is inferred from volume-pressure relationships during controlled breathing maneuvers. While such assessments provide global functional information, they are unable to offer region-specific morphological insights. In contrast, the present study utilizes real-time in situ imaging to assess alveolar-level expansion compliance, capturing local tissue behavior with micrometer resolution. The exponential trend observed in normal alveolar expansion mirrors the expected behavior in compliant lungs, as also reflected in pressure-volume loops. However, the bronchial-uOCT method additionally reveals regional heterogeneity in compliance, which spirometry cannot resolve.While spirometry offers a valuable global overview of lung mechanics, the method proposed here integrates high-resolution microstructural imaging with dynamic functional evaluation, filling a critical gap between imaging and function. This approach may serve as a bridge between imaging-derived morphometry and physiological assessments of compliance, especially in early-stage COPD.

While the absolute alveolar dimensions are reported in arbitrary units due to uncorrected refractive index mismatches, the relative trends and fitted models provide reproducible indicators of compliance.Importantly, the catheter is compatible with the working channels of standard bronchoscopes and manufactured using standard fiber-optic techniques, facilitating future clinical translation. Looking forward, large-scale data acquisition combined with AI-based morphological classification will further enhance diagnostic precision. Additionally, incorporating polarization-sensitive and elastic OCT techniques could provide insight into the role of collagen and elastin in alveolar mechanics, enriching our understanding of COPD progression and treatment response.

Funding

Ministry of Science and Technology of the People's Republic of China 10.13039/501100002855 ( 2022ZD0212200); National Natural Science Foundation of China 10.13039/501100001809 ( 12474430, 2022A1515011247, 2022A1515010548); Guangzhou National Laboratory ( GZNL2023A03002).

Disclosures

The authors have no conflicts to disclose.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.