Virtual bronchoscopic navigation and endobronchial ultrasound with a guide sheath without fluoroscopy for diagnosing peripheral pulmonary lesions with a bronchus leading to or adjacent to the lesion: A randomized non-inferiority trial.

BACKGROUND AND OBJECTIVE: Transbronchial sampling of peripheral pulmonary lesions (PPLs) is routinely performed under fluoroscopy. However, advanced ancillary techniques have become available, such as virtual bronchoscopic navigation (VBN) and radial endobronchial ultrasound with a guide sheath (rEBUS-GS). This study was performed to determine whether the diagnostic utility of VBN and rEBUS with a GS is similar with or without fluoroscopy. METHODS: This multicenter non-inferiority trial randomized patients to a VBN-rEBUS-GS with or without fluoroscopy group at three centres. The primary endpoint was the diagnostic yield. The secondary endpoints were the time for rEBUS, GS, and the total operation. Complications were also recorded. RESULTS: Four hundred and ninety-six subjects were assessed and 426 subjects were included in the analysis (212 in non-fluoroscopy-guided-group and 214 in fluoroscopy-guided-group). The diagnostic yield in the non-fluoroscopy-guided-group (84.0%) was not inferior to that in the fluoroscopy-guided-group (84.6%), with a diagnostic difference of -0.6% (95% CI: -6.4%, 5.2%). Multivariable analysis confirmed that bronchus sign and lesion nature were valuable diagnostic predictors in non-fluoroscopy-guided-group. The non-fluoroscopy-guided-group had shorter rEBUS, GS, and total operation time. No severe complications occurred in either group. CONCLUSION: Transbronchial diagnosis of PPLs suspicious of malignancy and presence of a bronchus leading to or adjacent to lesions using VBN-rEBUS-GS without fluoroscopy is a safe and effective method that is non-inferior to VBN-rEBUS-GS with fluoroscopy. Bronchus leading to lesions and malignant nature are associated with high diagnostic yield in VBN-rEBUS-GS without fluoroscopy for the diagnosis of PPLs.

The Novel Application of Robotic-Assisted Bronchoscopy Combined with Photodynamic Therapy for Adenoid Cystic Carcinoma of the Trachea.

Robotic platforms have been widely used in the various fields of clinical diagnosis and therapy of diseases in the past decade. Robotic-assisted bronchoscopy (RAB) demonstrates its advantages of visibility, flexibility, and stability in comparison to conventional bronchoscopic techniques. Improving diagnostic yield and navigation yield for peripheral pulmonary lesions has been defined; however, RAB platform of treatment was not reported. In this article, we report a case of a 52-year-old woman who was diagnosed with the tracheal adenoid cystic carcinoma and recurred in the second postoperative year, leading to the involvement of the entire tracheal wall and lumen obstruction. Since the lesion was inoperable, we combined RAB and photodynamic therapy (PDT) for the patient. The potential advantages of using RAB for PDT delivery include precise light irradiation of target lesions and stable intra-operative control over the long term. This is a novel application of RAB combined with PDT for airway diseases. The case report may provide a new insight into the diagnosis and treatment of pulmonary diseases. In addition to improving the diagnostic rates, the RAB platform may also play an important role in the treatment of airway and lung disease in the future.

5G-Based Remote Virtual Bronchoscopic Navigation-Guided Transbronchial Lung Biopsy for Diagnosis of Lung Cancer: Description of 2 Cases.

INTRODUCTION: Bronchoscopic navigation combined with endobronchial ultrasound-guided transbronchial lung biopsy (EBUS-TBLB) is an important approach for the diagnosis of peripheral pulmonary lesions (PPL). The fifth-generation (5G) network, characterized by low latency and high stability, has shown promising possibilities in telemedicine for remote areas. METHODS: We present two cases of PPL in primary hospitals without navigation equipment. The EBUS-TBLB was performed with the guidance of remote augmented reality virtual bronchoscopic navigation (VBN) based on a 5G network. In practice, the 5G network could enable the matching of actual/virtual bronchoscopic images and navigation paths in real time, as well as high-speed transmission at long distances (>20 km), without any visual delay (<500 ms). CONCLUSION: Both patients were successfully diagnosed with lung cancer after accurate positioning and obtaining biopsies of target lesions. This 5G-based remote VBN-guided EBUS-TBLB appears to be safe with reliable connections in both cases and shows potential for cost-effectiveness. It would be an optimal resource for undeveloped regions and/or regions lacking endoscopists with extensive experience in navigation-related procedures.

Deep Learning-Based Segmentation of Airway Morphology from Endobronchial Optical Coherence Tomography.

BACKGROUND: Manual measurement of endobronchial optical coherence tomography (EB-OCT) images means a heavy workload in the clinical practice, which can also introduce bias if the subjective opinions of doctors are involved. OBJECTIVE: We aim to develop a convolutional neural network (CNN)-based EB-OCT image analysis algorithm to automatically identify and measure EB-OCT parameters of airway morphology. METHODS: The ResUNet, MultiResUNet, and Siamese network were used for analyzing airway inner area (Ai), airway wall area (Aw), airway wall area percentage (Aw%), and airway bifurcate segmentation obtained from EB-OCT imaging, respectively. The accuracy of the automatic segmentations was verified by comparing with manual measurements. RESULTS: Thirty-three patients who were diagnosed with asthma (n = 13), chronic obstructive pulmonary disease (COPD, n = 13), and normal airway (n = 7) were enrolled. EB-OCT was performed in RB9 segment (lateral basal segment of the right lower lobe), and a total of 17,820 OCT images were collected for CNN training, validation, and testing. After training, the Ai, Aw, and airway bifurcate were readily identified in both normal airway and airways of asthma and COPD. The ResUNet and the MultiResUNet resulted in a mean dice similarity coefficient of 0.97 and 0.95 for Ai and Aw segmentation. The accuracy Siamese network in identifying airway bifurcate was 96.6%. Bland-Altman analysis indicated there was a negligible bias between manual and CNN measurements for Ai (bias = -0.02 to 0.01, 95% CI = -0.12 to 0.14) and Aw% (bias = -0.06 to 0.12, 95% CI = -1.98 to 2.14). CONCLUSION: EB-OCT imaging in conjunction with ResUNet, MultiResUNet, and Siamese network could automatically measure normal and diseased airway structure with an accurate performance.

Clinical application of endobronchial ultrasonography-guided transbronchial needle aspiration biopsy-a single center, large sample, real-world study.

BACKGROUND: Endobronchial ultrasonography-guided transbronchial needle aspiration biopsy (EBUS-TBNA) has been used for more than 10 years in China. Its clinical application and diagnostic value in different diseases with large sample was lack of report. METHODS: A retrospective analysis was performed about the application and diagnostic value of EBUS-TBNA in different disease of patients in Respiratory Intervention Center of Guangzhou Institute of Respiratory Health from January 2012 to July 2020. RESULTS: A total 5758 patients were included with 182 patients excluded for lack of information. Finally, data of 5576 patients (3798 males and 1778 females) were analyzed. For anesthetize, most patients were undergoing general anesthesia of intravenous with spontaneous breathing (69.4%), followed by general anesthesia of intravenous and inhalation with tracheal intubation and mechanical ventilation (17.9%) and conscious sedation and analgesia (12.8%). Lymph nodes were the main sites of biopsy obtained (76.4%). Tumors accounted for the highest proportion of disease (66.4%), followed by infection diseases (9.9%), sarcoidosis (3.9%), lymphoma (1.1%), and others (18.7%). The sensitivity of EBUS-TBNA for diagnosis of tumor was 89.7%, and 40.8% for infection diseases. There were significant differences in the puncture site and proportions of diseases between male and females (both p < 0.05). Higher diagnostic value was found in male patients (p < 0.05). CONCLUSION: EBUS-TBNA has good diagnostic value for different mediastinal and central pulmonary space-occupying lesions diseases, with highest sensitivity for tumors. Higher diagnostic value was found in male patients.

Hierarchical clock-scale hand-drawn mapping as a simple method for bronchoscopic navigation in peripheral pulmonary nodule.

BACKGROUND: A feasible and economical bronchoscopic navigation method in guiding peripheral pulmonary nodule biopsy is lacking. OBJECTIVE: To investigate the utility of hierarchical clock-scale hand-drawn mapping for bronchoscopic navigation in peripheral pulmonary nodules. METHODS: We developed a hierarchical clock-scale hand-drawn mapping for bronchoscopic navigation in peripheral pulmonary nodules. Patients with peripheral pulmonary nodules were recruited and assigned to two groups in this retrospective study, subjects in VBN group received conventional bronchoscopy in conjunction with virtual bronchoscopic navigation (VBN) and radial probe endobronchial ultrasound (RP-EBUS) for biopsy (VBN group), while HBN group underwent ultrathin bronchoscopy and RP-EBUS under the guidance of hand-drawn bronchoscopic navigation (HBN). The demographic characteristics, procedural time, operating cost and diagnostic yield were compared between these two groups. RESULTS: Forty-eight patients with peripheral pulmonary nodule were enrolled in HBN group, while 42 in VBN group. There were no significant differences between VBN and HBN groups in terms of age, gender, lesion size, location and radiographic type. The time of planning pathway (1.32 vs. 9.79 min, P < 0.001) and total operation (23.63 vs. 28.02 min, P = 0.002), as well as operating cost (758.31 ± 125.21 vs.1327.70 ± 116.25 USD, P < 0.001) were markedly less in HBN group, compared with those in VBN group. The pathological diagnostic efficiency of benign and malignant disease in HBN group appeared similar with those in VBN group, irrespective of the size of pulmonary lesion (larger or smaller than 20 mm). The total diagnostic yield of HBN had no marked difference from that of VBN (75.00% vs. 61.90%, P = 0.25). CONCLUSIONS: Hierarchical clock-scale hand-drawn mapping for bronchoscopic navigation could serve as a feasible and economical method for guiding peripheral pulmonary nodule biopsy, providing a comparable diagnostic yield in comparison with virtual bronchoscopic navigation.

A clinical and canine experimental study in small-airway response to bronchial thermoplasty: Role of the neuronal effect.

BACKGROUND: The effects of bronchial thermoplasty (BT) on smooth muscle (SM) and nerves in small airways are unclear. METHODS: We recruited 15 patients with severe refractory asthma, who received BT treatment. Endobronchial optical-coherence tomography (EB-OCT) was performed at baseline, 3 weeks' follow-up and 2 years' follow-up to evaluate the effect of BT on airway structure. In addition, we divided 12 healthy beagles into a sham group and a BT group, the latter receiving BT on large airways (inner diameter >3 mm) of the lower lobe. The dogs' lung lobes were resected to evaluate histological and neuronal changes of the treated large airways and untreated small airways 12 weeks after BT. RESULTS: Patients receiving BT treatment had significant improvement in Asthma Control Questionnaire (ACQ) scores and significant reduction in asthma exacerbations. EB-OCT results demonstrated a notable increase in inner-airway area (Ai) and decrease in airway wall area percentage (Aw%) in both large (3rd-to 6th-generation) and small (7th-to 9th-generation) airways. Furthermore, the animal study showed a significant reduction in the amount of SM in BT-treated large airways but not in untreated small airways. Protein gene product 9.5 (PGP9.5)-positive nerves and muscarinic receptor 3 (M3 receptor) expression in large and small airways were both markedly decreased throughout the airway wall 12 weeks after BT treatment. CONCLUSIONS: BT significantly reduced nerves, but not SM, in small airways, which might shed light on the mechanism of lung denervation by BT.

Feasibility and Safety of Radiofrequency Ablation Guided by Bronchoscopic Transparenchymal Nodule Access in Canines.

BACKGROUND: The treatment of pulmonary malignancies remains a challenge. The efficacy and safety of bronchoscopic radiofrequency ablation (RFA) for the treatment of lung cancer are not well elucidated. OBJECTIVE: This study aimed to evaluate the feasibility and safety of RFA guided by bronchoscopic transparenchymal nodule access (BTPNA) in vivo. METHODS: In an attempt to determine the parameters of RFA, we first performed RFA in conjunction with automatic saline microperfusion in the lung in vitro with various ablation energy (10, 15, 20, 25, and 30 W) and ablation times (3, 5, 8, and 10 min). The correlation between ablated area and RFA parameter was recorded and analyzed. Further, we conducted a canine study with RFA by BTPNA in vivo, observing the ablation effect and morphological changes in the lung assessed by chest CT and histopathologic examination at various follow-up time points (1 day, n = 3; 30 days, n = 4; 90 days, n = 4). The related complications were also observed and recorded. RESULTS: More ablation energy, but not ablation time, induced a greater range of ablation area in the lung. Ablation energy applied with 15 W for 3 min served as the appropriate setting for pulmonary lesions ≤1 cm. RFA guided by BTPNA was performed in 11 canines with 100% success rate. Inflammation, congestion, and coagulation necrosis were observed after ablation, which could be repaired within 7 days; subsequently, granulation and fibrotic scar tissue developed after 30 days. No procedure-related complication occurred during the operation or in the follow-up periods. CONCLUSION: The novel RFA system and catheter in conjunction with automatic saline microperfusion present a safe and feasible modality in pulmonary parenchyma. RFA guided by BTPNA appears to be well established with an acceptable tolerance; it might further provide therapeutic benefit in pulmonary malignancies.

Postintubation Tracheal Stenosis Evaluated by Endobronchial Optical Coherence Tomography: A Canine Model Study.

BACKGROUND: The predictors and airway morphological changes during the development of postintubation tracheal stenosis (PITS) have not been well elucidated. OBJECTIVES: To elucidate the validation of endobronchial optical coherence tomography (EB-OCT) in assessing the airway morphological changes in PITS. METHODS: We performed oral endotracheal intubation in 12 beagles to establish the PITS model. EB-OCT was performed respectively before modeling and on the 1st, 7th, and 12th day after extubation in 9 canines, and was conducted consecutively in 3 canines during the development of PITS. Histological findings and the thickness and gray-scale value of the tracheal wall assessed by EB-OCT measurements were analyzed and compared. RESULTS: The tracheal wall edema, granulation tissue proliferation, cartilage destruction in PITS, and airway wall thickening detected by EB-OCT were in concordance with the histopathological measurements. The consecutive EB-OCT observation of the airway structure demonstrated the tracheal wall thickness significantly increased from 344.41 ± 44.19 µm before modeling to 796.67 ± 49.75 µm on the 9th day after modeling (p < 0.05). The airway wall gray-scale values assessed by EB-OCT decreased from 111.19 ± 14.71 before modeling to 74.96 ± 4.08 on the 9th day after modeling (p < 0.05). The gray-scale value was negatively correlated with the airway wall thickness (r = -0.945, p = 0.001). CONCLUSION: The EB-OCT imaging, in concordance with the histopathological finding, was validated for assessing the airway morphological changes during the development of PITS. The EB-OCT evaluation of cartilage damage and gray-scale value measurement might help predict the development and prognosis of PITS.

Breathing Hydrogen-Oxygen Mixture Decreases Inspiratory Effort in Patients with Tracheal Stenosis.

BACKGROUND: Hydrogen-oxygen mixture (H2-O2) may reduce airway resistance in patients with acute severe tracheal stenosis, yet data supporting the clinical use of H2-O2 are insufficient. OBJECTIVES: To evaluate the efficacy and safety of breathing H2-O2 in acute severe tracheal stenosis. METHODS: Thirty-five consecutive patients with severe acute tracheal stenosis were recruited in this prospective self-control study. Air, H2-O2 and O2 inhalation was given in 4 consecutive breathing steps: air for 15 min, H2-O2 (6 L per min, H2:O2 = 2: 1) for 15 min, oxygen (3 L per min) for 15 min, and H2-O2 for 120 min. The primary endpoint was inspiratory effort as assessed by diaphragm electromyography (EMGdi); the secondary endpoints were transdiaphragmatic pressure (Pdi), Borg score, vital signs, and impulse oscillometry (IOS). The concentration of H2 in the ambient environment was obtained with 12 monitors. Adverse reactions during the inhalation were recorded. RESULTS: The mean reduction in the EMGdi under H2-O2 was 10.53 ± 6.83%. The EMGdi significantly decreased during 2 H2-O2 inhalation steps (Steps 2 and 4) compared with air (Step 1) and O2 (Step 3) (52.95 ± 15.00 vs. 42.46 ± 13.90 vs. 53.20 ± 14.74 vs. 42.50 ± 14.12% for Steps 1 through 4, p < 0.05). The mean reduction in the Pdi under H2-O2 was 4.77 ± 3.51 cmH2O. Breathing H2-O2 significantly improved the Borg score and resistance parameters of IOS but not vital signs. No adverse reactions occurred. H2 was undetectable in the environment throughout the procedure. CONCLUSIONS: Breathing H2-O2 may reduce the inspiratory effort in patients with acute severe tracheal stenosis and can be used for this purpose safely.

[The diagnostic utility of virtual bronchoscopic navigation combined with endobronchial ultrasound guided transbronchial lung biopsy for peripheral pulmonary lesions].

OBJECTIVE: To evaluate the diagnostic utility of endobronchial ultrasound combined with virtual bronchoscopic navigation guided transbronchial lung biopsy for solitary pulmonary nodules. METHODS: A total of 105 patients with suspected peripheral pulmonary lesions who underwent transbronchial lung biopsy in the First Affiliated Hospital of Guangzhou Medical University between January and December 2013 were prospectively evaluated. The patients were divided into a conventional group, an endobronchial ultrasound group(EBUS group)and a virtual bronchoscopic navigation combined with endobronchial ultrasound(VB+ EBUS) group. The diagnostic yield and operation time were compared. RESULTS: The lesion size of the conventional group, the EBUS group and the EBUS+ VB group were (23±8), (20±8)and(18±7)mm, respectively, and there was no significant difference in diagnostic yields by the lesion size (F=0.52, P=0.60). The EBUS+ VB group had the highest diagnostic yield(22 of 29, 76%), which was higher than that of the conventional group(17 of 36, 47%, χ(2)=7.47, P=0.024), but not that of the EBUS group(29 of 40, 72%, χ(2)=0.10, P=0.75). The EBUS group and the EBUS+ VB group did not differ in lesion location by pulmonary segments or histologic findings. The procedure time was significantly longer in the EBUS group than the EBUS+ VB group [(365±221)s verses (256±205)s, t=2.08, P=0.042]. CONCLUSIONS: EBUS guided TBLB improves the diagnostic yield in solitary pulmonary lesions, but it should be combined with virtual bronchoscopic navigation for the optimal yield.

[Refractory pulmonary alveolar proteinosis treated with inhaled granulocyte macrophage colony stimulating factor: a case report and review of the literatures].

OBJECTIVE: To observe the efficacy of granulocyte macrophage colony stimulating factor inhalation therapy for pulmonary alveolar proteinosis (PAP) patients and to improve the understanding of the therapy. METHODS: The clinical data of a patient with autoimmune PAP treated with inhaled granulocyte macrophage colony stimulating factor were described and the literatures were reviewed. RESULTS: This 70-year-old female patient was diagnosed with autoimmune PAP and had been treated with whole lung lavage for 10 times, but only achieved short-term symptomatic relief after each operation. Inhalation of granulocyte macrophage colony stimulating factor was then used after the recurrence of PAP, and a significant improvement in oxygenation and clinical symptoms were abserved. The patient remained stable 4 month after the therapy. CONCLUSION: Treatment with inhalation of granulocyte macrophage colony stimulating factor is safe, effective and economic for patients with PAP, and would be the first choice for the recurrence of autoimmune PAP after whole lung lavage.

[Electromagnetic navigation bronchoscopy real-time guidance lung biopsy for the diagnosis of small peripheral pulmonary lesions].

OBJECTIVE: To evaluate the diagnostic value and safety of electromagnetic navigation bronchoscopy real-time guidance lung biopsy (ENB-guided TBLB) of small peripheral lung lesions (diameter < 3 cm). METHODS: Patients were selected by small peripheral pulmonary lesions (diameter < 3 cm) on thoracic CT in our hospital between September 1, 2011 and May 30, 2012. ENB-guided TBLB was performed by flexible bronchoscope, while fluoroscopy-guided lung biopsy was performed as a control. Then the diagnostic yield of ENB-guided and fluoroscopy-guided lung biopsy was compared. The final diagnosis was confirmed by pathologic examination of surgically removed lesions or by 12 month follow-up until clinical cure being confirmed. The intraoperative and postoperative complications were recorded. RESULTS: Seventeen patients (11 males, 6 females) with 20 small peripheral pulmonary lesions all underwent ENB- guided TBLB. The patients' average age was (53 ± 14) yrs , and the lesions' average diameter was (1.8 ± 0.7) cm. Eleven patients finally received surgical resection, while other 6 patients received medical treatment and follow-up for more than 12 months to achieve clinical cure. Pathologic results showed that the diagnostic yield of ENB group was 80.0% (16/20) and that of the fluoroscopy group was 45.0% (9/20), the difference between the 2 groups being statistically significant (χ² = 5.227, P = 0.048). Subgroup analysis showed that, if the lesion diameter was 2-3 cm, the diagnostic yield between the 2 groups had no statistical difference (χ² = 0.000, P = 1.000), but if the lesion diameter was < 2 cm, the diagnostic yield of the 2 groups was statistically different (χ(2) = 8.224, P = 0.012). In the ENB method, the diagnostic yield between the diameter of 2-3 cm group and the < 2 cm group had no statistical difference (χ² = 0.208, P = 1.000). Analysis of the diagnostic yield of different navigation plan CT graph layer thickness showed that, the 1 cm group was higher than the 2 cm group (χ² = 6.111, P = 0.026). The diagnostic yield of lesions with "air bronchogram" was higher than that of lesions without "air bronchogram" (χ² = 6.111, P = 0.026). CONCLUSION: Electromagnetic navigation bronchoscopy real-time guidance lung biopsy for small peripheral pulmonary lesions is safe, and has higher diagnostic rate.

[The predictive value of gray scale sonographic features in malignant hilar and mediastinal lymph node].

OBJECTIVE: To explore the sonographic features of malignant hilar and mediastinal lymph nodes and the predictive value. METHODS: We performed retrospective analysis of 390 lymph nodes under the endobronchial ultrasound-guided transbronchial needle aspiration for diagnosis in the bronchoscope center of the First Affiliated Hospital of Guangzhou Medical University from November 1st, 2012 to November 31st, 2013. They were divided into malignant and benign groups, and sonographic features of lymph nodes were analyzed, including the length of short axis, shape, margin, echogenicity, homogenicity, coagulation necrosis, calcification, coalesence and posterior acoustic enhancement. RESULTS: A total of 390 lymph nodes were evaluated, including 207 malignant and 183 benign lymph nodes, respectively. The accuracy based on the length of short axis, shape, margin, echogenicity, homogenicity, coagulation necrosis, calcification, coalesence and posterior acoustic enhancement to predict the malignancy were 61.0% (238/390), 75.4% (294/390), 75.6% (295/390), 73.8% (288/390), 65.9% (257/390), 68.2% (266/390), 51.8% (202/390), 55.1% (215/390), 67.4% (263/390), respectively. Binary multivariable logistic regression analysis revealed that 5 parameters, including round, distinct margin and heterogeneous, hypoechoic, and posterior acoustic enhancement possessed a significant predictive value(P < 0.05).We found that only 35.3% (73/207) malignant lymph nodes possessed these 5 ultrasonographic characteristics, and the diagnosis rate was 93.6% (73/78).72.46% (150/207) malignant lymph nodes possessed any 4 or more positive sonographic features, and the diagnosis rate was 85.2% (150/176).85.0% (176/207) malignant lymph nodes had any 3 or more positive sonographic features, and the diagnosis rate was 84.6% (176/208). CONCLUSION: The combination of gray scale sonographic features has a predictive value in the diagnosis of malignant hilar and mediastinal lymph nodes.

[The cause and efficacy of benign tracheal stenosis].

OBJECTIVE: To analysis the causes of benign tracheal stenosis and evaluate the curative effect of intraluminal bronchoscopic treatment. METHODS: 158 patients with benign tracheal stenosis in our hospital from September 2005 to September 2012 were collected to retrospectively analysis the causes and clinic features of tracheal stenosis. Interventional treatments through bronchoscopy were used to treat the benign tracheal stenosis and the curative effects were evaluated. RESULTS: 158 cases of benign tracheal stenosis were recruited to our study, 69.6% of them were young and middle-aged. The main causes of benign tracheal stenosis were as follows: secondary to postintubation or tracheotomy in 61.4% (97/158), tuberculosis in 16% (26/158), benign tumor in 5.1% (8/158) and other 27 cases. 94.3% patients improved in symptoms with alleviation immediately after bronchoscopic treatment, the average tracheal diameter increased form (4.22 ± 2.06) mm to (10.16 ± 2.99) mm (t = 21.48, P < 0.01), dyspnea index decreased from 2.29 ± 0.75 to 0.63 ± 0.67 (t = 19.85, P < 0.01). The recurrence rate in 1 and 3 month after interventional treatment were 38.3% and 26.8%, respectively. CONCLUSION: The cases of benign tracheal stenosis were increasing year by year. The most common cause of benign tracheal stenosis was postintubation and tracheotomy. Interventional treatments through bronchoscopy is effective in treating benign tracheal stenosis, but repeated interventional procedures may be required to maintain the favorable long-term effects.

Distinguishing bronchoscopically observed anatomical positions of airway under by convolutional neural network.

Background: Artificial intelligence (AI) technology has been used for finding lesions via gastrointestinal endoscopy. However, there were few AI-associated studies that discuss bronchoscopy. Objectives: To use convolutional neural network (CNN) to recognize the observed anatomical positions of the airway under bronchoscopy. Design: We designed the study by comparing the imaging data of patients undergoing bronchoscopy from March 2022 to October 2022 by using EfficientNet (one of the CNNs) and U-Net. Methods: Based on the inclusion and exclusion criteria, 1527 clear images of normal anatomical positions of the airways from 200 patients were used for training, and 475 clear images from 72 patients were utilized for validation. Further, 20 bronchoscopic videos of examination procedures in another 20 patients with normal airway structures were used to extract the bronchoscopic images of normal anatomical positions to evaluate the accuracy for the model. Finally, 21 respiratory doctors were enrolled for the test of recognizing corrected anatomical positions using the validating datasets. Results: In all, 1527 bronchoscopic images of 200 patients with nine anatomical positions of the airway, including carina, right main bronchus, right upper lobe bronchus, right intermediate bronchus, right middle lobe bronchus, right lower lobe bronchus, left main bronchus, left upper lobe bronchus, and left lower lobe bronchus, were used for supervised machine learning and training, and 475 clear bronchoscopic images of 72 patients were used for validation. The mean accuracy of recognizing these 9 positions was 91% (carina: 98%, right main bronchus: 98%, right intermediate bronchus: 90%, right upper lobe bronchus: 91%, right middle lobe bronchus 92%, right lower lobe bronchus: 83%, left main bronchus: 89%, left upper bronchus: 91%, left lower bronchus: 76%). The area under the curves for these nine positions were >0.98. In addition, the accuracy of extracting the images via the video by the trained model was 94.7%. We also conducted a deep learning study to segment 10 segment bronchi in right lung, and 8 segment bronchi in Left lung. Because of the problem of radial depth, only segment bronchi distributions below right upper bronchus and right middle bronchus could be correctly recognized. The accuracy of recognizing was 84.33 ± 7.52% by doctors receiving interventional pulmonology education in our hospital over 6 months. Conclusion: Our study proved that AI technology can be used to distinguish the normal anatomical positions of the airway, and the model we trained could extract the corrected images via the video to help standardize data collection and control quality.

Endobronchial optical coherence tomography or computed tomography for evaluating progression of bronchiectasis.

Background: The early radiological signs of progression in bronchiectasis remain unclear. The objective of the present study was to compare endobronchial optical coherence tomography (EB-OCT) and chest computed tomography (CT) for the evaluation of radiological progression of bronchiectasis via stratification of the presence (TW+) or absence (TW-) of thickened-walled bronchioles surrounding dilated bronchi in patients with bronchiectasis based on CT, and determine the risk factors. Methods: In this prospective cohort study, we performed both chest CT and EB-OCT at baseline and 5-year follow-up, to compare changes in airway calibre metrics. We evaluated bacterial microbiology, sputum matrix metalloproteinase-9 levels and free neutrophil elastase activity at baseline. We compared clinical characteristics and airway calibre metrics between the TW+ and TW- groups. We ascertained radiological progression at 5 years via CT and EB-OCT. Results: We recruited 75 patients between 2014 and 2017. At baseline, EB-OCT metrics (mean luminal diameter (p=0.017), inner airway area (p=0.005) and airway wall area (p=0.009) of seventh- to ninth-generation bronchioles) were significantly greater in the TW+ group than in the TW-group. Meanwhile, EB-OCT did not reveal bronchiole dilatation (compared with the same segment of normal bronchioles) surrounding nondilated bronchi on CT in the TW- group. At 5 years, 53.1% of patients in the TW+ group progressed to have bronchiectasis measured with EB-OCT, compared with only 3.3% in TW- group (p<0.05). 34 patients in the TW+ group demonstrated marked dilatation of medium-sized and small airways. Higher baseline neutrophil elastase activity and TW+ bronchioles on CT predicted progression of bronchiectasis. Conclusion: Thickened-walled bronchioles surrounding the dilated bronchi, identified with EB-OCT, indicates progression of bronchiectasis.