Analysis of left anterior segmental bronchovascular patterns and its benefits for surgical implications: a retrospective cross-sectional study

Introduction

Recent landmark clinical studies, including the JCOG0802 (1) and CALGB140503 (2) trials, have demonstrated the efficacy of sub-lobar resections for tumours smaller than 2 cm and for clinical stage T1aN0 non-small cell lung cancer with ground-glass opacity components. Sublobar resections can be categorised into wedge resections, segmentectomies, and subsegmentectomies. Nowadays, the anatomic precise segmentectomies encompass a wide spectrum, varying from the resection of a single segment or subsegment to the combined resections of multiple segments or subsegments. Therefore, the surgical procedure relies heavily on a comprehensive understanding of the subsegmental anatomy of the lung.

Unlike traditional anatomy on cadavers, three-dimensional computed tomography bronchography and angiography (3-D CTBA) offer a non-invasive approach to understanding the complex organization of lung anatomy and indicate the significance of understanding these variations in surgical planning even when using only noncontrast computed tomography (CT) data. As one of the automated noncontrast CT reconstruction algorithm software, InferOperate Thoracic Surgery (3) has been widely accepted for its higher accuracy and automation.

The left upper lobe (LUL) has a higher prevalence of anatomical variations and severe complications (4). Within this lobe, the left anterior segment (LS³) lateral subsegment is located centrally at the junction of multiple subsegments, rendering surgical procedures in this area particularly challenging.

The watershed analysis (5) method has demonstrated its feasibility for wedge resections by temporarily blocking specific arteries. Moreover, another study (6) has reported that after splitting the left upper division and lingular, combined subsegmentectomy containing LS³ can be possible and safe. However, existing studies have not thoroughly investigated the variations in LS³.

Given the lack of comprehensive research on anatomical variations in this specific region, a comprehensive investigation is necessary to clarify the LS³ subsegmental bronchi and vessel patterns. For complex nodules located in the lateral subsegment of left anterior segment (LS³a), efficiently understanding and utilising these anatomical variations remains unresolved. To solve those problems, our study focuses on the distribution of LS³ bronchovascular patterns, particularly the incidence rate of an interlobar (IL) left anterior artery (A³), which has not been previously reported. We also hypothesize that preoperatively finding the anatomical variations may concise the operation and optimise the selection of the surgical approach for pulmonary nodules located in the LS³a. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-1397/rc).

Methods

Patients and 3-D CTBA

This retrospectively cross-sectional study enrolled 647 consecutive hospitalised patients who underwent thoracoscopic surgeries because of pulmonary nodules between March 2023 and August 2023 in the Thoracic Surgery Department of The First Affiliated Hospital of Nanjing Medical University. Subsequently, raw data from their CT scan were collected for reconstruction. 3-D CTBA models of patients were automatically processed and manually analysed using the InferOperate Thoracic Surgery software based on chest CT data. After excluding six cases due to malformation and another six cases for having surgical histories of LUL, 635 available cases (329 women and 306 men) were finally included, having a mean age of 50.1 years (range, 24–79 years). To control information bias, one surgeon underwent systematic training about the anatomical naming rules of the pulmonary system. After completing the training, the surgeon conducted systematic training and assessment for another two colleagues. Subsequently, three experienced thoracic surgeons confirmed the validity of all models. If different opinions arose, the final decision was made after discussion. Finally, distribution patterns of LS³ bronchus, arteries, and left upper division veins (LUDVs) were collected to classify and summarise. The nomenclature systems are illustrated below.

Inclusion criteria

• Chest CT exams with a slice thickness ≤1.5 mm, including non-contrast and contrast-enhanced CT scans.
• Availability of 3-D models of LUL bronchi and vessels.
• No previous history of left upper lung surgery.

Exclusion criteria

• Chest CT thickness exams over 1.5 mm.
• A history of left upper lung surgery.
• Dispute among experts regarding the validity of the reconstruction models.

Ethical statement

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Ethics Committee of Affiliated People’s Hospital of Jiangsu University (No. SQK-20220166-Y) and individual consent for this retrospective analysis was waived.

Definition of left upper lobar bronchi (B), arteries (A), and veins (V)

According to the nomenclature systems of Boyden (7), Yamashita (8), and the textbook of Chen et al. (9), the left apicoposterior segment (S¹⁺²) is composed of apical (S¹⁺²a), posterior (S¹⁺²b), and lateral (S¹⁺²c). The left anterior segment could be classified into lateral (S³a), medial (S³b), and superior (S³c) subsegments. Correspondingly, the bronchial names were labelled as B¹⁺²a, b, and c and B³a, b, and c, while the arterial names were labelled as A¹⁺²a, A¹⁺²b, A¹⁺²c, A³a, A³b, and A³c. According to Boyden (7), the first priority principle is superior and inferior, the second priority principle is posterior and anterior, and the last priority principle is lateral and internal. Sub-subsegments were designated with Roman numerals (i, ii, and iii), while the sub-subsegments were designated in the Greek alphabet (α, β, and γ). The apicoposterior segment vein (V¹⁺²) consisted of four branches: V¹⁺²a, V¹⁺²b, V¹⁺²c, and V¹⁺²d, although it may have a branch Vl. The anterior segment vein (V³) consisted of three branches: V³a, V³b and V³c. Normally, LUDV is comprised of three types: central vein, semi-central vein, and non-central vein.

Statistical analysis

All statistical analyses were conducted utilising Statistical Package for the Social Sciences software (version 27.0; Chicago, IL, USA). Descriptive data are summarised as the number of cases (%). We used the cross-tabulation method to analyse the correlations among the bronchus, artery, and vein. Fisher’s exact test was employed to assess the significance of associations between groups because over 20% of the cells contained counts less than 5, and some contained minimum expected counts below 1. In this study, a P<0.01 was considered statistically significant.

Results

Left B³ branching patterns

The left S³ branching patterns are detailed in Table 1, and images illustrating these patterns are displayed in Figure 1. In a study of 635 patients, six types of bronchial bifurcation and one trifurcation type were identified. The most common bifurcation type was B³a, B³b+c, found in 471 patients (74.17%). Meanwhile, there was only one instance of B³a, B³b+c belonging to left eparterial bronchus (11). Conversely, B³a+b, B³c, B³a+c, and B³b patterns were observed in 85 (13.39%) and 1 (0.16%), respectively. Interestingly, B³ai, B³aii+b+c type was reported for the first time in 17 (2.68%) patients. The BX³a, B³b+c type exhibited a descending pattern of B³a, potentially arising from the main trunk part of the left upper division bronchus either alone (4, 0.63%) or with B¹⁺²c (2, 0.31%). The trifurcation type B³a, B³b, and B³c was observed in 55 patients (8.66%).

Figure 1 Images of eight left anterior segmental bronchus (B³) branching patterns. (A) B³a+b, B³c; (B) B³a, B³b+c; (C) B³a+c, B³b; (D) BX³a (B¹⁺²c), B³b+c; (E) BX³a (Desc), B³b+c; (F) B³ai, B³aii+b+c; (G) B³a, B³b+c (left eparterial bronchus); (H) B³a, B³b, B³c. X, mean abnormal origin; B, bronchus; ³, anterior segment; ³a, lateral subsegment; ³b, medial subsegment; ³c, superior subsegment; ¹⁺², apicoposterior segment; ⁴, superior lingular segment; ⁵, inferior lingular segment; LPA, left pulmonary artery; Desc, descending of B³a to the trunk of left upper division; characters “i”, “ii” mean sub-subsegmental branches.

Left A³ branching patterns

This study represented the first anatomic distribution of the level of sub-sub-subsegmental pulmonary arteries, indicating that conventional classification systems are not applicable. We summarised 13 types based on the number of branches, the presence of IL arteries, and six different sources of arterial origin from 38 subtypes, as detailed in Table 2 and Table S1. Additionally, Figures 2,3 are the detailed images of these findings. In one branch type, only the mediastinal (M) type 1 was found in 463 patients (72.91%). The IL type that was reported (10) previously was not observed. While in the 164 cases (25.83%) of the two branch groups, type 2 comprised 36 cases that arose from the proximal left pulmonary artery (LPA). Apart from these two types, all other identified types displayed abnormal origins, which are detailed in Table 2. There were 8 cases (1.26%) that consisted of three A³ branches. Moreover, 28 cases (4.41%) exhibited abnormal branches originating from the mediastinal lingular artery (MLA). All 102 mediastinal and interlobar (MIL) type cases (16.06%) exhibited anatomic variations involving AX³a and its ramus.

Figure 2 Images of 18 M types of left segmental artery. (A) A³; (B) A³a+c, A³b; (C) A³a+b, A³c; (D) A³, A³ci; (E) A³aii+b, A³ai+c; (F) A³, A³aii; (G) A³a+c, AX³b; (H) A³, AX³aii; (I) AX³a, A³b+c; (J) A³a, AX³b+c; (K) AX³a+b, A³c; (L) A³, AX³ai+AX³aiiβ; (M) A³, AX³aiiα; (N) AX³a, A³b+c; (O) A³, AX³ci; (P) A³, AX³cii; (Q) AX³a+c, AX³b; (R) AX³a, A³b, A³c. A, artery; ³, anterior segment; 1, one branch; M, mediastinal; 2, two branches; 3, three branches; ¹⁺², apicoposterior segment; LA, lingular artery; MLA, mediastinal lingular artery; ³, anterior segment; X, mean abnormal origin; ³a, lateral subsegment; ³b, medial subsegment; ³c, superior subsegment; characters “i”, “ii”, “iii” mean sub-subsegmental branches; characters “α”, “β” mean sub-sub-subsegmental branch.

Figure 3 Images of 20 MIL types of A³. (A) A³, AX³aii (IL); (B) AX³a (IL), A³b+c; (C) A³, AX³ai (IL); (D) A³, AX³aiβ (IL); (E) A³, AX³aiiβ (IL); (F) AX³a, A³b+c (A¹⁺²c); (G) A³, AX³aii (A¹⁺²c); (H) A³, AX³aiβ (A¹⁺²c); (I) A³, AX³aii (LA); (J) AX³a (LA), A³b+c; (K) A³, AX³ai (LA); (L) A³, AX³aiiα (LA); (M) A³, AX³aiβ (LA); (N) A³, AX³aiiβ (LA); (O) AX³a (IL), AX³b (MLA), A³c; (P) A³aii+c, AX³ai (IL), AX³b (MLA); (Q) AX³a (IL), AX³b+ci (MLA), A³cii; (R) A³, AX³aii (IL), AX³aiii (LA); (S) AX³ai (IL), AX³aii ( LA), A³b+c; (T) A³aii+b, AX³ai (A¹⁺²c), A³c. A, artery; ³, anterior segment; 2, two branches of A³; MIL, mediastinal and interlobar; 3, three branches of A³; ¹⁺², apicoposterior segment; MLA, mediastinal lingular artery; IL, interlobar; LA, lingular artery; X, mean abnormal origin; ³a, lateral subsegment; ³b, medial subsegment; ³c, superior subsegment; characters “i”, “ii”, “iii” mean sub-subsegmental branches; characters “α”, “β” mean sub-sub-subsegmental branch.

Left A³a branching patterns

As detailed in Table 3, the A³a branching patterns consisted of single (560, 88.19%), double (74, 11.65%), and triple (1, 0.16%) patterns. The specifics are provided in Table S2. More specifics are listed in Table S2, including the same pictures as A³ and location information in column 5 of Table S2. The IL pulmonary artery (55, 8.66%) and lingular artery (LA) (49, 7.72%) represented the two most common abnormal origins types. The IL pulmonary artery consisted of two sub-types, including 39 cases (6.14%) originating from the IL portion of LPA singly or 16 cases (2.52%) sharing the same trunk with A¹⁺²c originating from LPA. Another two types, ectopic origins, accounted for 9 cases (1.42%) in A¹⁺²a+b and 14 cases (2.20%) in MLA, respectively. Moreover, the total incidence rate of sub-subsegmental arteries and its ramus was 11.7% (74/635).

Left upper division venous branching patterns

According to the nomenclature by Yamashita (8) and Wang et al. (12) regarding the left upper division venous pattern, the distribution of semi-central vein type (413, 65.04%), central vein type (178, 28.03%), and non-central vein type (44, 6.92%) are outlined in Table 4 and Figure 4. In addition to the five known types, we identified a rare new type named non-central O type, observed in two cases where V¹⁺²c abnormally drains blood to V⁶. Similarly, the semi-central A type was the most common, accounting for nearly half the total (317, 49.92%).

Figure 4 3-dimensional images of six left upper division venous branching patterns. (A) Cent-A; (B) Cent-B; (C) Semi-cent A; (D) Semi-cent B; (E) Non-cent; (F) Non-cent-O. V, vein; B, bronchus; S, segment; ¹⁺², apicoposterior segment; ³, anterior segment; V¹⁺²a, intra-segmental vein between apical subsegment and superior subsegment of S³; V¹⁺²b, intersubsegment vein between S¹⁺²a and S¹⁺²b; V¹⁺²c, intersubsegment vein between posterior subsegment of S¹⁺² and lateral subsegment of S¹⁺²c; V¹⁺²d, intersubsegment vein among lateral subsegment of S¹⁺² and lateral subsegment of S³ and lateral subsegment of superior lingular segment; V³a, intersubsegment vein between lateral subsegment of S³ and superior subsegment of S³, and between lateral subsegment of S³ and medial subsegment of S³; V³b, intersubsegment vein between medial subsegment of S³ and anterior subsegment of lingular; V³c, intersubsegment vein between medial subsegment of S³ and superior subsegment of S³; A, artery; ¹⁺²c, lateral subsegment; 6, superior segment of the left lower lobe.

Correlation analysis of A³a patterns, bronchus patterns, and venous patterns

Our statistics revealed that A³, A³a, and B³a and venous patterns of the left upper division comprised 38, 26, 7, and 6 different types, respectively. However, there are likely even more variations. Based on our clinical experience, we optimised these patterns. The classifications are detailed in Tables 1-4. The number of arterial branches and their original types were analysed independently. We analysed the correlations among all bronchovascular patterns. Significant relationships (all P<0.001) were observed between A³a branches number, origins, and subtype classification with A³ patterns, B³ patterns, and left upper division venous patterns, as outlined in Table 5.

Decision theory tree

According to the bronchovascular variation types of the left anterior segment, particularly the lateral subsegmental arteries and bronchus variants, we developed an anatomical variation-associated surgical decision tree for the ≤2 cm nodules located in the middle-lateral zone of the left lateral anterior subsegment. Our theory is illustrated in Figure 5. For subsegment A³a, classified as M type, the primary surgical plan recommended was intra-venous single-direction approach surgery regardless of the bronchus type. If A³a includes the IL ramus sharing the trunk with IL A¹⁺²c along with the bronchus type B³a+B¹⁺²c, B³b+c+B¹⁺²a+b, combined subsegmentectomy of S¹⁺²c+S³a was recommended. Conversely, split surgery of the left upper division and S³a subsegmentectomy were suggested. Moreover, trans-fissure single-direction approach surgery was recommended as an easier surgical procedure for single-branch IL A³a and descending B³a. Besides, there were two operation plans for IL A³a type and MIL A³a type. One plan was trans-fissure Watershed analysis method surgery, which simplifies the procedure and shortens operation time, while the other was split surgery of left upper division and S³a subsegmentectomy for experienced surgeons.

Figure 5 Decision theory tree of surgeries related lateral subsegment of LS³a based on arterial and bronchial patterns. ^†, both the type of IL LPA and LA is OK. A, artery; M, mediastinal; IL, interlobar; MIL, mediastinal and interlobar; LPA, left pulmonary artery; ¹⁺², apicoposterior segment; MLA, mediastinal lingular artery; LA, lingular artery; B, bronchus; ³, anterior segment; X, mean abnormal origin; ¹⁺²a, apical subsegment; ¹⁺²b, posterior subsegment; ¹⁺²c, lateral subsegment; ³a, lateral subsegment; ³b, medial subsegment; ³c, superior subsegment; S, segment.

Discussion

We conducted a comprehensive analysis of bronchovascular patterns using three-dimensional models of 635 patients, revealing new insights into the intricate relationship between these structures and the implications for surgical strategy. Our findings include identifying and reclassifying sub-subsegmental branches and the development of a decision tree to aid surgeons in surgical decision-making.

The current classification system of LUL pulmonary subsegments has predominantly relied on anatomical models developed by researchers such as Boyden (7) and Yamashita (8). As our comprehension of lung anatomy deepens, there is a significant shift from traditional anatomy toward imaging-based anatomy, particularly exemplified by the advancements in 3-D CTBA. This transition enables a more accurate understanding of bronchovascular anatomy in clinical practice.

A comparison with Li et al. (10) B³ patterns is presented in Table 1. We reported a novel bifurcation type of the B³ bronchus, specifically the B³ai and B³aii+b+c. We also identified a variety of arterial classifications at the sub-subsegmental level for the A³ artery, encompassing 38 distinct types, with 27 variations specifically for A³a. These arteries are categorised based on their branching origins, grouping into 13 unique combinations of sub-types. Following the five-type central vein model by Wang et al. (12) in Table 4, we uncovered a rare anatomical variation beyond these five classifications, namely the non-central O type.

Our detailed classifications of bronchial, arterial, and venous patterns at the LS³ sub-subsegment level present a comprehensive framework that enhances knowledge of LUL anatomy for surgeons. Onuki and colleagues (13) proposed that lung segments do not persistently exist as distinct units from the embryonic period and that these segments are artificially named. Throughout our research, we encountered numerous confusion regarding the nomenclature of pulmonary subsegments and their ramus. If two similar three-dimensional reconstruction models exhibited subtle variations in their superior-inferior, anterior-posterior or lateral-internal relationships, it may lead to entirely different names. Our new A³ and A³a pulmonary arterial classification method, based on branching number, presence or absence of IL artery, and anomalous branching origins, better reflects clinical applicability and is not influenced by the pulmonary nomenclature systems. However, the efficacy and clinical feasibility of these classifications require further validation through prospective studies.

The interrelationship between pulmonary arterial, bronchial, and pulmonary venous patterns has not been well documented. Isaka et al. (14) were the first to identify the relationship among pulmonary artery, pulmonary vein, and bronchus patterns in LUL. Their findings are consistent with our results; however, we specifically demonstrate the correlation between the A³a arterial branching types and the patterns of left upper division bronchi or veins. No other significant results were found among other S³ bronchovascular patterns. This phenomenon suggests that A³a is crucial in forming the left upper division.

The resection of pulmonary arteries and bronchi is essential for segmentectomy and subsegmentectomy, as it directly determines the extent of lung tissue excised. Previous studies (15-17) have reported that pulmonary veins in the target segment region should be divided, and the intersegmental veins should be preserved. This preservation serves as a boundary indicator of lung segments and subsegments, enhancing the precision of excisions.

The concept of three-dimensional guided, nodule-centered, and cone-shaped precise segmentectomy or combined subsegmentectomy has gained widespread recognition among thoracic surgeons. This approach has demonstrated satisfactory therapeutic outcomes and safety margins compared to lobectomy for pulmonary nodules in the middle-lateral region (18-20).

Wedge resection, the simplest form of pulmonary surgery, relies significantly on the accurate localisation of the tumour. The watershed analysis method (5), a contemporary approach to wedge resection, utilises three-dimensional reconstructions to identify target arteries. This method temporarily uses the colored ribbon to ligate the target artery with a slipknot and then administers indocyanine green for fluorescence visualisation. This approach more accurately delineates the resection boundaries while minimizing damage associated with invasive localisation techniques, resembling the anatomical wedge resection. However, the value of IL A³a in this study has been underestimated.

For complex nodules located between multiple lung segments or subsegments, combined resection techniques can efficiently preserve more functional lung tissue. Hong et al. (6) reported a new method, “Split”, combined subsegmentectomy for two cases of S³ combined with S¹⁺²c and one case of S³+S¹⁺²c+S¹⁺²a+bi. However, these procedures pose higher technical challenges and risks, leading thoracic surgeons to opt for lobectomy or broader wedge resections.

To address lung nodules in the complex region S³a, we developed a decision tree based on various surgical approaches and anatomical insights regarding the distribution of the LS³ pulmonary arteries and bronchi. This approach adheres to the principle of “nodule-centered, cone-shaped precise sub-lobar resection”. Our innovative framework incorporates the type of pulmonary arteries and bronchi and the location of the nodules using preoperative 3-D CTBA models, offering tailored surgical recommendations. By employing this theory, we can offer optimal surgical plans for patients based on their anatomical characteristics. This tailored strategy optimizes surgical options for different patients, providing feasible alternatives for those requiring simplified surgical procedures, and equips thoracic surgeons with critical anatomical knowledge. Furthermore, it aids in anticipating the complexity of the surgical procedure, thereby fostering additional preparedness and confidence among the surgical team. For instance, if a preoperative reconstruction reveals an IL A³a, the watershed analysis may be selected to minimise operative time and trauma. Conversely, anatomical variations such as a descending B³a in conjunction with IL A³a may favour a complete trans-fissure approach operation to avoid excessive damage to the anterior mediastinum.

However, this study has several potential limitations. First, the study relies solely on a fully automated noncontrast CT reconstruction algorithm (3), which may not achieve 100% accuracy and have bias between our results and real anatomical patterns. Second, the authors independently determined subsegment and sub-subsegment naming, which could be improved by more explicit procedures in future studies. Lastly, the findings of this study should be validated through multiracial research involving a larger patient cohort.

Conclusions

We identified a sub-subsegmental level within the left S³ bronchovascular distribution patterns and defined the frequency of anatomic variations in 635 patients using 3-D CTBA. This study also confirmed the correlation between A³a with B³, A³, and LUDV. The abnormal A³a, which originates from the IL portion of LPA, is common, having a frequency rate of 16.06% (102/635) among people. For surgeons intending to perform subsegmentectomies, the five ectopic origins of LA, IL LPA, IL A¹⁺²c, MLA, and A¹⁺²a+b are crucial for locating the A³a. Meanwhile, for those opting to select a concise wedge resection, the watershed analysis method and the occurrence of IL A³a can make the operation easier.

Acknowledgments

Funding: This work was supported by the Affiliated People’s Hospital of Jiangsu University (No. Y2022023) and the Key Project of Jiangsu Commission of Health (No. ZD2022055).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-1397/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-1397/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-1397/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-24-1397/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Ethics Committee of Affiliated People’s Hospital of Jiangsu University (No. SQK-20220166-Y) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Saji H, Okada M, Tsuboi M, et al. Segmentectomy versus lobectomy in small-sized peripheral non-small-cell lung cancer (JCOG0802/WJOG4607L): a multicentre, open-label, phase 3, randomised, controlled, non-inferiority trial. Lancet 2022;399:1607-17. [Crossref] [PubMed]
2. Altorki N, Wang X, Kozono D, et al. Lobar or Sublobar Resection for Peripheral Stage IA Non-Small-Cell Lung Cancer. N Engl J Med 2023;388:489-98. [Crossref] [PubMed]
3. Chen X, Wang Z, Qi Q, et al. A fully automated noncontrast CT 3-D reconstruction algorithm enabled accurate anatomical demonstration for lung segmentectomy. Thorac Cancer 2022;13:795-803. [Crossref] [PubMed]
4. Decaluwe H, Petersen RH, Hansen H, et al. Major intraoperative complications during video-assisted thoracoscopic anatomical lung resections: an intention-to-treat analysis. Eur J Cardiothorac Surg 2015;48:588-98; discussion 599. [Crossref] [PubMed]
5. Chu XP, Chen ZH, Lin SM, et al. Watershed analysis of the target pulmonary artery for real-time localization of non-palpable pulmonary nodules. Transl Lung Cancer Res 2021;10:1711-9. [Crossref] [PubMed]
6. Hong R, Chen C, Zheng W, et al. "Split" combined subsegmentectomy: A case series. Thorac Cancer 2022;13:423-9. [Crossref] [PubMed]
7. Boyden EA. Segmental anatomy of the lungs a study of the patterns of the segmental bronchi and related pulmonary vessels. New York: McGraw-Hill; 1955:185-200.
8. Yamashita HJ. Roentgenologic Anatomy of the Lung. Stuttgart: Georg Thieme Publishers; 1978.
9. Chen L, Zhu Q, Wu W. Atlas of thoracoscopic anatomi pulmonary subsegmentectomy. Amsterdam: Elsevier; 2023.
10. Li Z, Zhao Q, Wu W, et al. Analysis of bronchovascular patterns in the left superior division segment to explore the relationship between the descending bronchus and the artery crossing intersegmental planes. Front Oncol 2023;13:1183227. [Crossref] [PubMed]
11. Chassagnon G, Morel B, Carpentier E, et al. Tracheobronchial Branching Abnormalities: Lobe-based Classification Scheme. Radiographics 2016;36:358-73. [Crossref] [PubMed]
12. Wang J, Lin H, Bian C, et al. A modified system for classifying the bilateral superior pulmonary veins using three-dimensional computed tomography bronchography and angiography images. J Thorac Dis 2021;13:5933-41. [Crossref] [PubMed]
13. Onuki T, Kanzaki M, Wachi N, et al. Do the lung segments exist continuously from the early stage of the embryonic period as units? Interact Cardiovasc Thorac Surg 2016;23:26-30. [Crossref] [PubMed]
14. Isaka T, Mitsuboshi S, Maeda H, et al. Anatomical analysis of the left upper lobe of lung on three-dimensional images with focusing the branching pattern of the subsegmental veins. J Cardiothorac Surg 2020;15:273. [Crossref] [PubMed]
15. Oizumi H, Endoh M, Takeda S, et al. Anatomical lung segmentectomy simulated by computed tomographic angiography. Ann Thorac Surg 2010;90:1382-3. [Crossref] [PubMed]
16. Oizumi H, Kanauchi N, Kato H, et al. Anatomic thoracoscopic pulmonary segmentectomy under 3-dimensional multidetector computed tomography simulation: a report of 52 consecutive cases. J Thorac Cardiovasc Surg 2011;141:678-82. [Crossref] [PubMed]
17. Wu WB, Xu XF, Wen W, et al. Three-dimensional computed tomography bronchography and angiography in the preoperative evaluation of thoracoscopic segmentectomy and subsegmentectomy. J Thorac Dis 2016;8:S710-5. [Crossref] [PubMed]
18. Li Z, Xu W, Pan X, et al. Segmentectomy versus lobectomy for small-sized pure solid non-small cell lung cancer. Thorac Cancer 2023;14:1021-8. Erratum in: Thorac Cancer 2024;15:1038. [Crossref] [PubMed]
19. Li Z, Xu W, Wang J, et al. Three-Dimensional Guided Cone-Shaped Segmentectomy Versus Lobectomy for Small-sized Non-Small Cell Lung Cancer in the Middle Third of the Lung Field. Ann Surg Oncol 2023;30:6684-92. [Crossref] [PubMed]
20. Wu WB, Xia Y, Pan XL, et al. Three-dimensional navigation-guided thoracoscopic combined subsegmentectomy for intersegmental pulmonary nodules. Thorac Cancer 2019;10:41-6. [Crossref] [PubMed]