Application of hand-drawn bronchial navigation and rapid on-site evaluation combined with endobronchial ultrasound for the diagnosis of peripheral lung lesions

Abstract

Objective

To investigate the clinical value of radial endobronchial ultrasound (rEBUS) combined with rapid on-site evaluation (ROSE) under hand-drawn bronchial navigation (HBN) planning for the diagnosis of peripheral pulmonary lesions.

Methods

A total of 186 patients were enrolled in this retrospective case–control study and classified into three groups according to the intervention method: rEBUS group (n = 74), HBN-rEBUS group (n = 57), and HBN-rEBUS-ROSE group (n = 54). The final diagnostic yield, lesion detection rate, time to lesion localization, number of biopsies performed, and incidence of complications were compared across the three groups.

Results

The lesion detection rate was higher in the HBN-rEBUS group than in the rEBUS group, and the HBN-rEBUS group demonstrated fewer ultrasound probe passes and a shorter time to lesion localization (p < 0.001). The HBN-rEBUS-ROSE group required fewer biopsies and had a lower incidence of bleeding than the HBN-rEBUS group (p < 0.001). The HBN-rEBUS-ROSE group also achieved a higher diagnostic yield than both the rEBUS and HBN-rEBUS groups (p < 0.05).

Conclusion

HBN-rEBUS facilitates faster and more precise localization of peripheral pulmonary lesions, shortens procedural duration, and, when combined with ROSE, reduces bleeding and improves the diagnostic yield of peripheral pulmonary lesions.

Introduction

Global cancer statistics from 2022 indicate that lung cancer remains a malignant tumor with high incidence and mortality rates. ¹ Although low-dose computed tomography screening in high-risk populations reduces lung cancer mortality, it can yield false-positive results. This often leads to overdiagnosis and unnecessary invasive surgical procedures, thereby increasing the patient burden. ² Consequently, preoperative localization and definitive characterization of peripheral pulmonary lesions (PPLs) are critical, with histopathology serving as the gold standard for diagnosis. ³ Radial endobronchial ultrasound (rEBUS) offers precise localization with minimal invasiveness. ⁴ However, standard rEBUS lacks intrinsic navigation capabilities. ⁵ In recent years, significant progress has been made in the diagnosis of PPLs through advanced guidance modalities, including virtual navigation systems, electromagnetic navigation bronchoscopy (ENB), and robotic bronchoscopy, all of which have markedly improved diagnostic efficiency. ⁶ However, these advanced systems are expensive and technically demanding. Their limited accessibility in most hospitals—particularly primary healthcare institutions—and their high cost for patients pose significant challenges to the widespread, accurate diagnosis of PPLs. ⁷

Hand-drawn bronchial navigation (HBN) is an effective, simplified navigation method that integrates CT with bronchoscopic orientation. By correlating the spatial orientation of CT images with bronchoscopic views through manual scrolling and rotation, operators can map a schematic diagram of the bronchial branches to locate PPLs. ⁸ This method is practical, rapid, effective, and virtually cost-free. By planning the pathway for rEBUS examination, lesions can be detected quickly and efficiently.^5,6 HBN is well-suited for implementation in medical centers at all levels, especially in primary healthcare units. Currently, however, domestic and international data regarding hand-drawn intraluminal navigation remain limited.

Even after successful lesion localization using HBN planning combined with rEBUS, challenges related to complications and specimen positivity rates during the subsequent biopsy warrant further investigation. Factors such as the location, size, and nature (solid vs. ground-glass opacity) of the lesion, as well as the resistance encountered during probe advancement, may influence the diagnostic sensitivity of rEBUS-guided transbronchial lung biopsy (TBLB).^9,10 This uncertainty regarding specimen adequacy often prompts physicians to increase the number of biopsies to ensure sufficient tissue sampling, ultimately prolonging the procedure and increasing the risks of hemorrhage and pneumothorax. Extensive research has confirmed that integrating rapid on-site evaluation (ROSE) during procedures such as TBLB and transbronchial needle aspiration (TBNA) effectively enhances diagnostic accuracy, shortens procedure time, reduces the number of biopsies, and consequently lowers the incidence of complications.^4,9–11

Therefore, the objective of this study was to investigate the clinical value of combining HBN pathway planning with ROSE during rEBUS, with a particular focus on improving the diagnostic yield of PPLs and reducing the incidence of procedure-related complications.

Materials and methods

Study design and participants

This was a retrospective cohort study. The study was conducted in accordance with the Declaration of Helsinki of 1975, as revised in 2024, and was approved by the Institutional Review Board of Anshun City People's Hospital (ASSRMYYLL-20250120). All patient-related data used in this study were fully de-identified to protect patient privacy and confidentiality. The Institutional Review Board waived the requirement for individual informed consent because of the retrospective study design and the use of fully de-identified data. A total of 185 patients with PPLs who underwent rEBUS bronchoscopy at Anshun City People's Hospital between January 2024 and October 2025 were enrolled. Patients were divided into three groups based on the sequential implementation of diagnostic techniques during the study period. All interventional procedures were performed by experienced endoscopists. All operators received standardized training and strictly followed the same operating procedures to ensure procedural consistency throughout the study. The rEBUS group (n = 74) comprised patients who underwent rEBUS alone without HBN or ROSE. The HBN-rEBUS group (n = 57) consisted of patients who underwent rEBUS guided by HBN. The HBN-rEBUS-ROSE group (n = 54) included patients who underwent rEBUS with both HBN planning and ROSE. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. ¹²

Inclusion criteria were as follows: (a) age ≥18 years; (b) chest CT revealing PPLs with a “bronchus sign,” where the target bronchus was not visible within the lobar bronchus corresponding to the diameter of a 4.0 mm bronchoscope; (c) general health status suitable for general anesthesia; and (d) availability of complete clinical and imaging data. Exclusion criteria included the following: (a) patients with a prior diagnosis who were currently undergoing treatment and required repeat biopsy; (b) incomplete biopsy procedures due to factors unrelated to intraoperative complications; (c) patients requiring concurrent bronchial mucosal biopsy; and (d) presence of preoperative pneumothorax, uncontrolled hemoptysis, or active airway bleeding.

Instruments and equipment

Bronchoscopy was performed using an Olympus BF-P290 electronic bronchoscope (outer diameter, 4.2 mm; working channel, 2.0 mm) connected to an Olympus CV-290 video processor (Olympus, Tokyo, Japan). Ultrasound examination was conducted using an EU-10 endoscopic ultrasound system (Kaili, Shenzhen, China) paired with a UM-1720 radial ultrasound microprobe. Cytological evaluation was performed using an Olympus CX43 biological microscope.

HBN planning

A thin-section (1 mm) chest CT scan was selected for planning. Following the mapping methods described by Mingli ⁸ and Lei, ¹³ the spatial orientation of the bronchi (cranial-caudal, ventral-dorsal, and mediastinal-lateral) was aligned with the bronchoscopist's perspective while standing at the patient's head. The CT images were rotated to match the orientation of the bronchoscopic view, referring to the hierarchical clock-scale hand-drawn mapping method described by Zhong Changhao⁶: (a) 90° counterclockwise rotation for the right upper lobe; (b) 90° clockwise rotation for the left upper lobe; (c) 180° rotation for the lower lobes, right middle lobe, and lingula; and (d) no rotation for the dorsal segments. The pathway was planned by scrolling through the coronal CT images to identify the target lesion and tracing the bronchial pathway proximally to the segmental bronchus. If the peripheral bronchi were indistinguishable, adjacent blood vessels were used as landmarks. A navigation map was manually drawn within 5 min: bronchial orifices were represented as circles, originating from the segmental bronchus and branching distally to the target lesion. The target bronchus was marked with an arrow (→) inside the final circle. A representative case is shown in Figure 1.

Figure 1.

A representative case of hand-drawn bronchial navigation planning. A1, A2 = third level bronchus; B1, B2 = fourth-level bronchus; C1, C2 = fifth-level bronchus; and D1, D2 = sixth-level bronchus.

rEBUS procedure

All procedures were performed under general anesthesia with airway management using either a nasopharyngeal airway or a laryngeal mask. Conventional bronchoscopy was first used to identify the target lung segment. In the rEBUS group, the target segment was identified based on the operator's mental reconstruction of the thin-section chest CT images. In the HBN-rEBUS group, the operator followed the manually drawn navigation map. The radial ultrasound probe was inserted through the working channel into the target bronchus, and timing was initiated. The rEBUS group proceeded based on preoperative CT interpretation, whereas the HBN-rEBUS group followed the navigation map to advance the probe. Upon reaching the target area, the ultrasound mode was activated to visualize the lesion. The time required to detect the lesion (seconds) and the total number of probe passes required for localization were recorded.

Once the lesion was located, the probe was withdrawn, and the distance was marked on the biopsy forceps to ensure consistent insertion depth. TBLB was then performed. To balance diagnostic yield with safety, the total number of biopsies was limited to ≤10. The operator visually assessed specimen quality; if deemed adequate, biopsies were continued at the same site. Otherwise, the biopsy angle or site was adjusted. Complications were monitored rigorously. Hemorrhage was classified according to the Adult Diagnostic Flexible Bronchoscopy Guidelines, ¹⁴ and moderate or severe bleeding requiring pharmacological hemostasis or balloon tamponade was recorded as “Hemorrhage.” Pneumothorax was assessed by postoperative arterial blood gas analysis and chest imaging, with clinically significant pneumothorax defined by the need for intervention or the presence of hypoxia or dyspnea. All patients underwent synchronous segmental bronchoalveolar lavage for microbiological analysis, including smear and culture, as well as targeted next-generation sequencing (tNGS).

ROSE procedure

In the HBN-rEBUS-ROSE group, the TBLB procedure followed the HBN-rEBUS protocol. Biopsy specimens were immediately smeared onto glass slides and stained using the Diff-Quik method (15 s in Solution A, rinse; 15 s in Solution B, rinse). A respiratory physician trained in cytopathology evaluated the slides on-site (Figure 2). If the specimen demonstrated definitive benign or malignant features, sufficient tissue was retained, and the procedure was terminated. If the specimen showed suspicious or atypical cells, additional samples were obtained from multiple sites before termination. If the specimen was inadequate (e.g. only blood or mucosal cells), the biopsy angle or site was adjusted, and the biopsy was repeated. A representative case is shown in Figure 2.

Figure 2.

A representative case of rapid on-site evaluation procedure. (a) The lesion in the basal segment of the right lower lobe is imaged with left–right inversion using the hand-drawn navigation method. (b) Hand-drawn bronchial navigation map. (c) Lesion target bronchus. (d) Lesion ultrasound image. (e) Rapid on-site evaluation abnormal cells. (f) Pathological adenocarcinoma.

Diagnostic interpretation

A definitive diagnosis was established based on the following criteria: (a) neoplastic lesions were confirmed by positive histopathology; (b) infectious diseases (e.g. tuberculosis and lung abscess) were confirmed by pathology combined with microbiological findings (culture/tNGS) and clinical response to therapy; (c) nonneoplastic, noninfectious lesions (e.g. organizing pneumonia) were confirmed by excluding infection and malignancy, supported by clinical improvement and radiological resolution during the 6-month follow-up; and (d) cases with inconclusive pathology and microbiological findings were considered undetermined unless clarified by surgical resection or percutaneous biopsy or if they showed no resolution during the 6-month CT follow-up.

Statistical analysis

Data were analyzed using SPSS version 29.0 and GraphPad Prism version 10.6.0. Normality was assessed using the Shapiro-Wilk test. Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range) and compared using Student's t-test, the Mann–Whitney U test, or the Kruskal–Wallis H test, as appropriate. Categorical variables were presented as percentages and analyzed using the chi-square test. A p-value <0.05 was considered statistically significant. The diagnostic rate was calculated as the ratio of confirmed cases to the total number of valid examinations.

Results

Baseline characteristics

A total of 185 patients with PPLs were enrolled, including 131 males and 54 females. The cohort was divided into the rEBUS group (n = 74), the HBN-rEBUS group (n = 57), and the HBN-rEBUS-ROSE group (n = 54). Comparative analysis revealed no significant differences among the three groups with regard to age, sex, lesion location, or lesion size (p > 0.05). Detailed baseline characteristics are presented in Table 1.

Table 1.

Baseline characteristics of patients with peripheral pulmonary lesions.

	rEBUS group	HBN-rEBUS group	HBN-rEBUS-ROSE group	p-value
Cases, no.	74	57	54
Age, years, mean ± SD	64.11 ± 11.128	60.68 ± 13.585	64.20 ± 11.348	0.196
Male, n (%)	47 (63.5%)	42 (73.7%)	42 (77.8%)	0.182
Female, n (%)	27 (36.5%)	15 (26.3%)	12 (22.2%)
Lesion size, mm, median (IQR)	33.75 (22.88–41.13)	29.00 (20.25–43.75)	30.00 (23.76–45.63)	0.507
Location of lesion, n (%)				0.927
Right upper lobe	19 (25.7%)	17 (29.8%)	12 (22.2%)
Right middle lobe	7 (9.5%)	6 (10.5%)	6 (11.1%)
Right lower lobe	13 (17.6%)	9 (15.8%)	14 (25.9%)
Left upper lobe	23 (31.1%)	17 (29.8%)	13 (24.1%)
Left lower lobe	12 (16.2%)	8 (14.0%)	9 (16.7%)

HBN: hand-drawn bronchial navigation; rEBUS: radial endobronchial ultrasound; ROSE: rapid on-site evaluation.

Procedural outcomes and complications

As shown in Table 2, there were no significant differences between the rEBUS group and the HBN-rEBUS group in the number of biopsies performed or the incidence of common complications, such as pneumothorax and hemorrhage (p > 0.05). In the rEBUS group, lesions were not detected in four patients, resulting in a detection rate of 94.59% (70/74), whereas the HBN-rEBUS group achieved a 100% detection rate (57/57). Among patients with successfully detected lesions, the HBN-rEBUS group demonstrated significantly fewer ultrasound probe passes and a shorter time to lesion detection than the rEBUS group (p < 0.001).

Table 2.

Comparison of operational data between the rEBUS and HBN-rEBUS groups.

	rEBUS group	HBN-rEBUS group	p-value
Cases, no.	70	57
Number of ultrasound probe passes time, median (IQR), time	3 (2–4)	1 (1–2)	<0.001
Time to lesion localization,a median (IQR), s	130 (78.75–228.75)	28 (15.00–56.50)	<0.001
Number of biopsies, median (IQR), Time	8 (6–8)	8 (6–8)	0.211
Pneumothorax situation, no. (%)
No	68 (97.1%)	55 (96.5%)	0.069
Yes	2 (2.9%)	2 (3.5%)
Bleeding situation, no. (%)
No	37 (52.9%)	34 (59.6%)	0.443
Yes	33 (47.1%)	23 (40.4%)

HBN; hand-drawn bronchial navigation; rEBUS: radial endobronchial ultrasound; ROSE: rapid on-site evaluation.

^aThe timing starts when the bronchoscope and ultrasound probe reach the orifice of the target pulmonary segment bronchus and ends when the lesion is detected, including the total time of multiple repeated explorations.

As shown in Table 3, when the HBN-rEBUS group was compared with the HBN-rEBUS-ROSE group, both achieved a 100% lesion detection rate. No significant differences were observed in the number of ultrasound probe passes, detection time, or the incidence of pneumothorax (p > 0.05). However, significant differences were observed in the number of biopsies (p < 0.001) and the incidence of hemorrhage (p = 0.001).

Table 3.

Comparison of operational data between the HBN-rEBUS and HBN-rEBUS-ROSE groups.

	HBN-rEBUS group	HBN-rEBUS-ROSE group	p-value
Cases, no.	57	54
Number of ultrasound probe passes time, median (IQR), time	1 (1–2)	1 (1–2)	0.296
Time to lesion localizationa, median (IQR), s	28.00 (15.00–56.50)	24.5 (12.00–47.00)	0.438
Number of biopsies, median (IQR), Time	8 (6–8)	5 (5–6)	<0.001
Pneumothorax situation, no. (%)			1.000
No	55 (96.5%)	53 (98.1%)
Yes	2 (3.5%)	1 (1.9%)
Bleeding situation, no. (%)			0.001
No	34 (59.6%)	47 (87.0%)
Yes	23 (40.4%)	7 (13.0%)

HBN: hand-drawn bronchial navigation; rEBUS: radial endobronchial ultrasound.

^aThe timing starts when the bronchoscope and ultrasound probe reach the orifice of the target pulmonary segment bronchus and ends when the lesion is detected, including the total time of multiple repeated explorations.

The overall incidence of pneumothorax was 2.8% (5/181) across all groups: 2.9% (2/70) in the rEBUS group, 3.5% (2/57) in the HBN-rEBUS group, and 1.9% (1/54) in the HBN-rEBUS-ROSE group. Two cases (one in the rEBUS group and one in the HBN-rEBUS group) required closed thoracic drainage, whereas three improved with oxygen therapy alone. The overall incidence of hemorrhage was 34.81% (63/181), with rates of 47.1% (33/70) in the rEBUS group, 40.4% (23/57) in the HBN-rEBUS group, and 13.0% (7/54) in the HBN-rEBUS-ROSE group. Severe bleeding occurred in only one case (rEBUS group) and was controlled intraoperatively with balloon tamponade. The remaining cases involved moderate bleeding that was successfully managed with bronchoscopic occlusion and local epinephrine injection, requiring no further postoperative intervention.

Final diagnostic outcomes

As shown in Figure 3, a total of 185 patients were enrolled and divided into three groups. In the rEBUS group (n = 74), lesions were not detected in four patients, and an additional four patients were lost to follow-up, leaving 66 patients for the final analysis. In the HBN-rEBUS group (n = 57), three patients were lost to follow-up, resulting in 54 patients included in the statistical analysis. In the HBN-rEBUS-ROSE group (n = 54), four patients were lost to follow-up, leaving 50 patients for the final analysis. Thus, a total of 170 patients were included in the final diagnostic analysis.

Figure 3.

Flowchart of patient enrollment and exclusion.

As shown in Table 4, the overall diagnostic yield for malignant lesions across the three groups was 35.88% (61/170), and that for benign lesions was 34.12% (58/170). In the rEBUS group, 40 cases yielded a confirmed diagnosis (24 malignant and 16 benign), whereas 26 remained undiagnosed. In the HBN-rEBUS group, 36 cases resulted in a confirmed diagnosis (12 malignant and 24 benign), whereas 18 remained undiagnosed. In the HBN-rEBUS-ROSE group, a confirmed diagnosis was established in 43 cases (25 malignant and 18 benign), whereas seven remained undiagnosed.

Table 4.

Final diagnosis results of the three groups of patients.

	rEBUS group	HBN-rEBUS group	HBN-rEBUS-ROSE group
Cases, no.	66	54	50
Malignant	24	12	25
Nonsmall cell lung cancer	20	10	22
Small cell lung cancer	3	0	0
Metastatic cancer	1	0	3
Carcinoma in situ	0	1	0
Lymphoma	0	1	0
Benign	16	24	18
General inflammatory lesionsa	3	6	7
Tuberculosis	6	3	2
Organizing pneumonia	2	6	2
Mycosis		4	1
Lung abscess	4	3	5
IgG4-related disease		1
Actinomycete infection		1
Hamartoma			1
Atypical adenomatous hyperplasia	1
Undiagnosed	26	18	7
Not sure	8	3	4

^aConsidering the irrigation fluid and targeted next-generation sequencing (tNGS) results, bacterial infection is suspected and after anti-infective treatment, the lesion was absorbed on re-examination.

Diagnostic efficacy

As shown in Figure 4, the overall diagnostic accuracy for the entire cohort was 70.00% (119/170). The specific diagnostic accuracy rates were 60.60% (40/66) for the rEBUS group, 66.7% (36/54) for the HBN-rEBUS group, and 86.00% (43/50) for the HBN-rEBUS-ROSE group. There was no statistically significant difference in diagnostic accuracy between the rEBUS and HBN-rEBUS groups (χ² = 0.470, p = 0.493). However, the HBN-rEBUS-ROSE group demonstrated significantly higher diagnostic accuracy than both the rEBUS group (χ² = 9.012, p = 0.003) and the HBN-rEBUS group (χ² = 5.314, p = 0.021).

Figure 4.

Comparison of diagnostic rates of three diagnostic methods. *indicates a statistically significant difference with p < 0.05, ns indicates no statistically significant difference (p > 0.05).

Discussion

PPLs are anatomically defined as lesions located distal to the segmental bronchi, surrounded by lung parenchyma, and not visible on routine bronchoscopy. Based on etiology, they are classified as benign or malignant space-occupying lesions. ¹⁵ Because of their deep location and complex airway anatomy, the diagnosis of PPLs has long been a clinical focus in respiratory medicine. Although rEBUS has significantly improved diagnostic accessibility for PPLs, the conventional procedure relies on physician experience for image interpretation and lesion localization without additional navigation capabilities, resulting in low exploration efficiency and a high false-negative rate. ¹⁶ Although bronchoscopic navigation systems (e.g. virtual navigation bronchoscopy (VBN) and ENB) exhibit high diagnostic rates, these devices and systems are expensive and technically demanding. Studies have reported issues such as high consumable use and prolonged training periods,^17,18 which limit the widespread application of navigation technology in primary healthcare institutions. This retrospective case–control study systematically evaluated the application value of HBN combined with ROSE in the rEBUS diagnosis of PPLs. The results demonstrated that this combined approach achieved improvements in localization precision, operational efficiency, and diagnostic accuracy without additional equipment costs, providing a feasible pathway for the diagnosis of PPLs in primary hospitals.

HBN technology originates from the precise integration of CT imaging and bronchoscopy operational perspectives. It is based on a hand-drawn navigation method gradually developed by the Kurimoto team during the application of radial ultrasound, ¹⁹ which transforms complex three-dimensional airway structures into intuitive two-dimensional navigation maps. By clearly delineating each step from the bronchus to the lesion, it reduces the impact of individual differences on localization efficiency. This enables physicians with varying levels of experience to rapidly target the lesion. The results of this study demonstrated that the HBN-rEBUS group achieved a 100% lesion detection rate, higher than that of the rEBUS group, with significantly fewer ultrasound probe examinations and a shorter lesion detection time than the rEBUS group. This effectively avoids the preoperative preparation time and intraoperative system synchronization required for devices such as electromagnetic navigation, virtual navigation, and robotic navigation, ensuring the smoothness and efficiency of the entire procedure and thereby reducing the overall surgical time. Additionally, the significant reduction in ultrasound bronchoscopy examinations implies more economical consumption of the ultrasound microprobe lifespan. This cost-saving effect on consumables is expected to lower the clinical application threshold for rEBUS. Furthermore, the median number of ultrasound probe examinations for lesion localization in both the HBN-rEBUS group and the HBN-rEBUS-ROSE group was one, suggesting that, in primary hospitals where the high cost of equipment and consumables limits the widespread adoption of ultrasound bronchoscopy, HBN may serve as an effective navigation pathway for traditional TBLB performed solely with chest CT.

In terms of diagnostic efficacy, although the HBN-rEBUS group had a higher diagnostic rate (66.7%) compared to the rEBUS group (60.60%), the difference between the two groups was not statistically significant. This may be attributed to the following limitations in the process of performing lung biopsies after detecting lesions with rEBUS: 1. The accuracy of lung biopsy under rEBUS guidance is influenced by the multiple bifurcations distal to the bronchial bundle and the operator's experience; ²⁰ 2. There is an issue of inaccurate localization of “peripheral” lesions; ²¹ and 3. Insufficient sample volume is a concern in rEBUS-guided lung biopsy. ²² This study found that when HBN-guided rEBUS was combined with ROSE technology for lung biopsy, the diagnostic accuracy improved to 86.00%, showing a statistically significant difference compared with the HBN-rEBUS group (p = 0.021). A prospective, multicenter, randomized controlled clinical trial reported that guide sheath (GS)-assisted ENB-guided rEBUS enhanced the ability to localize PPLs, achieving a diagnostic detection rate of 82.9%. ²³ Another meta-analysis revealed that the diagnostic rate for CT-guided percutaneous lung biopsy or needle aspiration (CT-TBNA) was 88.9%, robot-assisted bronchoscopy (RAB) was 84.8%, and virtual navigation bronchoscopy was 75.1%. ¹⁶ A previous study also showed that the diagnostic rate of HBN was higher than that of VBN (75% vs. 61.9%), although the difference was not statistically significant. ⁶ Overall, the diagnostic accuracy of HBN-guided rEBUS combined with ROSE technology is comparable to that of CT-TBNA, GS-ENB, and RAB and is higher than that of VBN. Its advantages lie in improving diagnostic accuracy while markedly reducing procedure time and lowering surgical costs.

The overall proportion of malignant lesions in this study was 35.88% (61/170), whereas the proportion of benign lesions was 34.12% (58/170), consistent with the epidemiological distribution of benign and malignant peripheral lung lesions in the VBN-rEBUS group reported by Zhou et al. ²⁴ In terms of pathological subtypes, nonsmall cell lung cancer (NSCLC) was the predominant malignant subtype across all three groups, accounting for 86.89% (53/61) of all malignant cases, consistent with the pathological characteristics of lung cancer. ²⁵ By comparing the diagnostic results of rEBUS alone, rEBUS combined with HBN, and rEBUS combined with HBN and ROSE, this study explored the optimization value of multimodal collaboration for diagnosing peripheral lung lesions in most primary hospitals under available equipment conditions. The results showed a significant gradient increase in diagnostic accuracy rate of effective cases across the three groups, while the nonspecific diagnosis rate gradually decreased from 39.39% to 14.00%, further confirming the diagnostic advantages of multimodal collaboration.

In terms of safety, the current main diagnostic methods for PPLs in internal medicine involve percutaneous and bronchoscopic approaches. A retrospective study ²⁶ performed CT-guided percutaneous lung biopsy on 1063 patients, revealing that although this method demonstrates high diagnostic efficacy, it is associated with a relatively high incidence of complications, including pneumothorax (31.7%), hemorrhage (12.04%), hemoptysis (1.60%), and hemothorax (0.38%). In the present study, the overall incidence of pneumothorax was 2.76% (5/181), with only two cases requiring intervention via closed thoracic drainage. Follow-up of the five patients with pneumothorax revealed that all lesions were located outside the lung fields, adjacent to the pleura. Therefore, bronchoscopic lung biopsy may be the preferred approach for PPLs with bronchial access, particularly in patients with lesions distant from the chest wall who cannot tolerate the risk of pneumothorax. Regarding bleeding risk, the overall incidence in this study was 34.81%, although most cases were moderate. Local injection of epinephrine saline under bronchoscopy effectively controlled bleeding, with only one patient in the rEBUS group experiencing massive hemorrhage that required hemostasis using a bronchial occlusion balloon. Moreover, the use of ROSE technology significantly reduced the number of intraoperative biopsies, lowering the bleeding rate to 13.0%. This benefit is attributed to the “real-time intraoperative quality control” feature of ROSE technology, which minimizes unnecessary biopsies and bleeding risks by allowing immediate cessation of the procedure upon obtaining qualified specimens, thereby avoiding airway injury caused by excessive sampling.

This study was a single-center retrospective controlled study and has several limitations. First, the sample size was relatively small, and subgroup analyses based on lesion density and distance from the pleura were not performed, making it difficult to determine the diagnostic performance of the combined approach in PPLs with different characteristics. Additionally, the statistical results for some rare pathological subtypes may be biased, affecting the generalizability of the conclusions. Second, our center currently lacks electronic navigation systems such as VBN, ENB, and RAB; therefore, no concurrent comparative studies were performed. Third, the sequential implementation of the different techniques introduced potential temporal bias. Despite standardized training and efforts to minimize learning-curve effects, minor improvements in operator proficiency over time cannot be completely excluded and may have influenced the intergroup comparisons of diagnostic performance. Future research should focus on three areas. First, expanding the sample size and conducting randomized controlled trials to further validate these findings and perform subgroup analyses according to lesion characteristics. Second, participating in multicenter studies incorporating other electronic navigation systems, such as VBN and ENB, as control groups. Third, establishing a standardized training system for HBN, systematically training personnel in primary hospitals without rEBUS to perform HBN-guided TBLB, and collecting relevant data to validate the clinical utility of HBN in blind TBLB procedures.

Conclusions

The application of HBN combined with ROSE technology in rEBUS-guided lung biopsy for the diagnosis of PPLs may indicate significantly improve lesion localization efficiency, reduce biopsy-related complications, and enhance diagnostic accuracy. This protocol is simple to operate, cost-effective, and does not require high-end equipment, making it particularly suitable for primary healthcare institutions lacking expensive navigation devices. It may provide a replicable clinical pathway for the precise diagnosis and treatment of patients with PPLs, demonstrating significant clinical value. Future large-scale, multicenter prospective studies may be conducted to validate its long-term diagnostic efficacy and health economic benefits.