Prediction and intraoperative validation of pulmonary interlobar fissure development based on high-resolution thin-slice computed tomography and multiplanar reconstruction three-dimensional reconstruction technology: a prospective study Authors: Xiuxin Mo, Hongbing Zhang, Chengcheng Chen, Jun Chen Published: 2026 ABSTRACT BackgroundPulmonary interlobar fissure development is an important anatomical factor influencing the complexity of thoracoscopic pulmonary resection and surgical strategy selection. Accurate preoperative assessment of fissure completeness may facilitate operative planning and reduce intraoperative risks. This prospective diagnostic accuracy study aimed to evaluate the accuracy of high-resolution thin-slice computed tomography (HRCT) combined with multiplanar reconstruction (MPR) three-dimensional (3D) technology in predicting the development of pulmonary interlobar fissures, and to explore regional differences in predictive performance through intraoperative validation.MethodsA total of 447 patients who underwent unilateral pulmonary surgery between September 2023 and April 2024 were enrolled. Preoperatively, three radiologists independently scored the development of operative-side interlobar fissures (divided into 5 segments: right horizontal fissure, upper/lower segments of right oblique fissure, and upper/lower segments of left oblique fissure) using HRCT and MPR images, with the average score as the preoperative result. Intraoperatively, three thoracic surgeons independently assessed the actual condition of the fissures, and the average score was recorded as the intraoperative result. The consistency between preoperative and intraoperative scores was analyzed using concordance rate, proportion of differences within 1 grade, and Kappa coefficient.ResultsThe overall concordance rate between preoperative and intraoperative scores was 55.42%, with 88.67% of cases showing a score difference within 1 grade (Kappa =0.48, P<0.01), indicating moderate agreement. Subgroup analysis revealed that the left lung had higher predictive accuracy than the right lung: the concordance rate was 59.54% vs. 52.87%, and the proportion of differences within 1 grade was 90.24% vs. 84.48% (left lung Kappa =0.556; right lung Kappa =0.437, both P<0.01). Among specific segments, the lower segment of the left oblique fissure exhibited the highest concordance rate (60.93%) and Kappa value (0.563), while the right horizontal fissure had the lowest concordance rate (52.59%) and Kappa value (0.414).ConclusionsHRCT and MPR 3D reconstruction technology can effectively predict the development of pulmonary interlobar fissures, with superior accuracy in the left lung compared to the right lung. These findings provide valuable reference for preoperative planning of thoracic surgery, particularly in assessing complex fissure segments such as the right horizontal fissure. Highlight box Key findings • High-resolution thin-slice computed tomography (HRCT) combined with multiplanar reconstruction (MPR) showed moderate agreement with intraoperative assessment of pulmonary interlobar fissure development (concordance 55.42%; 88.67% within one grade; Kappa =0.48). • Predictive accuracy was higher in the left lung than in the right lung (Kappa 0.556 vs. 0.437). • The right horizontal fissure demonstrated the lowest predictive performance. What is known and what is new? • Interlobar fissure development affects surgical difficulty and outcomes in video-assisted thoracoscopic surgery (VATS), and preoperative computed tomography (CT) evaluation is commonly used, but discrepancies with intraoperative findings remain. • This prospective study validates HRCT combined with MPR against intraoperative findings and demonstrates regional differences in predictive accuracy. What is the implication, and what should change now? • HRCT with MPR can assist preoperative assessment of fissure development for surgical planning. • Imaging findings should be interpreted cautiously for complex regions such as the right horizontal fissure. Introduction The development of the pulmonary interlobar fissures is a critical factor influencing the complexity of thoracic surgical procedures (1). Accurate assessment of interlobar fissural development before pulmonary surgery is essential for effective surgical planning (2). Variations in fissural complexity and individual differences may increase the difficulty of surgery, intraoperative risks, and postoperative complications, particularly in cases of incomplete fissure development (3,4). With the widespread use of chest computed tomography (CT), the detection rate of pulmonary nodules has significantly increased, leading to a trend toward more minimally invasive and precise surgical techniques (5). Video-assisted thoracoscopic surgery (VATS) has gained widespread application due to its advantages, including minimal trauma, fewer postoperative complications, lower mortality, and comparable tumor resection outcomes to traditional open lobectomy (6,7). The development of the interlobar fissures has a certain impact on VATS complications. This study aims to evaluate the predictive capability of high-resolution thin-slice computed tomography (HRCT) and multiplanar reconstruction (MPR) techniques in assessing interlobar fissure development, providing valuable references for clinical practice. We present this article in accordance with the STARD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-0778/rc). Methods Study population This was a prospective diagnostic accuracy study. This study included 447 patients who underwent pulmonary surgery between September 2023 and April 2024. This study was approved by the Institutional Review Board of Weifang Second People’s Hospital (approval number: KY2024-077-01). All procedures were conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from all participants prior to enrollment, including consent for the use of clinical data and imaging materials for research purposes. For participants with limited decision-making capacity, informed consent was obtained from their legally authorized representatives. All patients underwent unilateral pulmonary surgery. Inclusion criteria were: patients scheduled for VATS or open pulmonary surgery (including lobectomy, segmentectomy, and wedge resection) with preoperative HRCT and MPR images of sufficient quality to allow clear visualization and grading of the target interlobar fissures (i.e., no significant motion artifacts, severe image noise, or incomplete coverage of the operative lung lobe that would impair fissure assessment). With reference to other relevant studies (8), the exclusion criteria for this study are formulated as follows: patients unable to effectively assess pulmonary interlobar fissures—including those with lesion-related obstruction or destruction (e.g., central lung cancer) that precludes fissure identification via preoperative imaging (HRCT, MPR) or intraoperative observation, those with underlying diseases (pulmonary/pleural inflammation, emphysema meeting specific criteria (e.g., areas with CT value ≤−950 HU accounting for >10% of lung volume) or bronchiectasis), those with a history of ipsilateral lung surgery leading to anatomical changes and inaccurate assessment, and those with poor preoperative CT quality (artifacts, noise, incomplete coverage) that hinders fissure grading—and patients with missing key data (incomplete preoperative assessment data, including visual CT, MPR analysis or intraoperative grading data). Participants formed a consecutive series of all eligible patients during the study period. Sample size was based on consecutive enrollment without a priori power calculation. This study was not registered in a public registry. Imaging evaluation CT examinations were conducted using a Siemens dual-source 64-slice spiral CT scanner. Patients were placed in the supine position, instructed to hold their breath after deep inspiration, and scanned from the lung apex to the diaphragm. The dual-source mode was not required for pulmonary scanning. The scanning parameters were as follows: tube voltage of 120 kV, tube current of 71–200 mAs, rotation time of 0.5 s per rotation, and pitch of 1.2. Image post-processing: all images were uploaded to the workstation for reconstruction. Axial images were reconstructed with a slice thickness and interval of 1.25 mm, while sagittal and coronal images were reconstructed with a slice thickness and interval of 1.5 mm. Image analysis was performed on the workstation, with a window width of 1,500 HU and a window level of −750 HU. The interlobar fissures assessed depended on the specific surgical approach. Observers primarily examined sagittal images to evaluate the target interlobar fissure. For areas where the fissure appeared discontinuous, axial and coronal images were used to confirm findings, ensuring accuracy. Image evaluation: observers conducted a comprehensive evaluation of the interlobar fissures using axial, sagittal, and coronal views. Measurements were performed when necessary. Two experienced radiologists independently analyzed the images. In cases of disagreement, the fissure discontinuity was reviewed, and consensus was reached through discussion with a third radiologist. Radiologists were blinded to intraoperative findings. Imaging scoring criteria: the development of interlobar fissures was assessed using MPR images. Fissure completeness was graded on multiplanar CT images using a four-level classification based on the continuity of the fissure and its relationship to the pulmonary artery, ranging from complete fissures to absent fissures (3). Fissures that extended continuously from the lung apex to the base without visible interruption were classified as Grade 1. If the fissure extended medially toward the artery but did not reach the mediastinum, it was classified as Grade 2. Fissures that were indistinct or terminated within the lung tissue before reaching the pulmonary artery were classified as Grade 3 or 4. Specifically, fissures terminating before reaching the pulmonary artery were graded as 3, while the absence of visible fissures was graded as 4. Scoring followed a previously described multidetector CT-based fissure assessment method (9), in which the right oblique fissure was divided into upper (posterior) and lower (anterior) segments, with the right horizontal fissure as the dividing marker. Similarly, the left oblique fissure was divided into upper (posterior) and lower (anterior) segments, with the fissure location at the pulmonary artery serving as the marker (Figure 1). A total of five interlobar fissure segments were evaluated and graded individually. Right lung (left side of the image): with the right horizontal fissure as the dividing point, the right oblique fissure is split into upper and lower segments, which are named the upper segment of the right oblique fissure and the lower segment of the right oblique fissure respectively. Left lung (right side of the image): taking the position of the fissure at the pulmonary artery as the dividing line, the left oblique fissure is divided into the upper segment of the left oblique fissure and lower segment of the left oblique fissure, respectively. Surgical scoring system During surgery, three thoracic surgeons with attending or higher qualifications independently assessed the patients’ interlobar fissure completeness using the same four-grade classification as described for CT evaluation (3). Surgeons were blinded to preoperative imaging scores. The final intraoperative score was determined by averaging the three surgeons’ scores, rounded to the nearest integer. Intraoperative visual assessment was chosen as the reference standard because it represents direct anatomical observation of fissure completeness. Statistical analysis Data were presented as frequencies and percentages. The preoperative and intraoperative fissure integrity scores were compared using the Wilcoxon signed-rank test. Statistical analyses were performed using SPSS 18.0, with P<0.05 indicating statistical significance. Preoperative fissure scores were compared with intraoperative scores for each patient to calculate the concordance rate and evaluate scoring consistency. The Kappa coefficient was used to assess the agreement between preoperative and intraoperative scores, reflecting the accuracy of HRCT and MPR three-dimensional (3D) reconstruction techniques in predicting fissure integrity. Kappa values were interpreted as follows (10): 0.01: very weak agreement; 0.01–0.20: weak agreement; 0.21–0.40: fair agreement; 0.41–0.60: moderate agreement; 0.61–0.80: substantial agreement; 0.81–1.00: almost perfect agreement. Results Patient characteristics A total of 447 patients were included in the study, comprising 181 males and 266 females. The average patient age was 63 years (range, 45–80 years). Among them, 232 patients underwent right lung surgery, while 215 underwent left lung surgery (Table 1). Table 1Clinical and demographic characteristics of patientsParametersData, N=447Male181 (40.5)Age (years)63±9.2Smoking status Never smoked212 (47.4) Current smoker78 (17.4) Former smoker157 (35.2)Surgical indications Lung nodule376 (84.1) Lung mass42 (9.4) Known lung cancer†29 (6.5)Surgical procedure VATS441 (98.7) Thoracotomy6 (1.3)Surgical type Lobectomy173 (38.7) Wedge resection127 (28.4) Segmentectomy147 (32.9)Surgical side Right232 (51.9) Left215 (48.1)Use of contrast-enhanced CT Yes416 (93.1) No31 (6.9)Data are presented as n (%) or mean ± standard deviation. †, patients with known lung cancer are those whose preoperative biopsy pathology confirmed the diagnosis of lung cancer. CT, computed tomography; VATS, video-assisted thoracoscopic surgery. Clinical and demographic characteristics of patients Data are presented as n (%) or mean ± standard deviation. †, patients with known lung cancer are those whose preoperative biopsy pathology confirmed the diagnosis of lung cancer. CT, computed tomography; VATS, video-assisted thoracoscopic surgery. The cohort included 447 patients, with 181 males (40.5%) and 266 females (59.5%). The mean age was 63 years (range, 45–80 years). Regarding smoking status, 212 patients (47.4%) had never smoked, 78 (17.4%) were current smokers, and 157 (35.2%) were former smokers. The primary surgical indication was lung nodules (376 cases, 84.1%), followed by lung masses (42 cases, 9.4%) and known lung cancer (29 cases, 6.5%). Among surgical procedures, 441 cases (98.7%) were performed via VATS and 6 (1.3%) via thoracotomy. Surgical types included lobectomy (173 cases, 38.7%), wedge resection (127 cases, 28.4%), and segmentectomy (147 cases, 32.9%). Surgeries were distributed between the right lung (232 cases, 51.9%) and left lung (215 cases, 48.1%). Contrast-enhanced CT was used in 416 cases (93.1%), with 31 cases (6.9%) not using contrast. The accuracy assessment of predicting the development of pulmonary lobar fissures Preoperative CT scores were compared with intraoperative scores, and the results are summarized as follows (Figure 2). No adverse events were reported from performing preoperative CT or intraoperative fissure assessment. This figure shows the comparative distribution of preoperative prediction results and intraoperative findings for pulmonary lobar fissure segments. Scatter points for the upper and lower segments of the left oblique fissure are concentrated near the diagonal line, indicating high consistency between preoperative and intraoperative assessments. In contrast, scatter points for the right horizontal fissure are more dispersed, further confirming the higher difficulty in predicting the right horizontal fissure. Overall, the total concordance rate between preoperative and intraoperative scores for the whole lung was 55.42%, with 88.67% of cases showing a score difference within 1 grade (Kappa =0.48, P<0.01). Subgroup analysis revealed that the left lung had a higher overall concordance rate (59.54%) than the right lung (52.87%). Among specific segments, the lower segment of the left oblique fissure exhibited the highest concordance rate (60.93%) and Kappa value (0.563), while the right horizontal fissure had the lowest concordance rate (52.59%), with 17.67% of cases showing a score difference exceeding 1 grade. Similarity analysis Exact similarity calculation: each pair of data points from two datasets was compared item by item. The number of exact matches was counted and divided by the total number of data points to obtain the exact similarity ratio. Similarity within a difference of 1: for each pair of data points, the absolute difference was calculated. If the difference was less than or equal to 1, the pair was considered similar. The number of similar pairs was counted and divided by the total number of data points to determine the similarity ratio within a difference of 1. Similarity representation: the final similarity results were multiplied by 100 to convert them into percentage form. Exact similarity reflects complete consistency between the two datasets, while similarity within a difference of 1 accounts for minor discrepancies, representing overall data consistency (Table 2, Figure 3). Table 2Preoperative interlobar fissure classification & intraoperative grading and scoring similarity analysisLocationScoring agreement rateProportion of scoring differences within 1 gradeKappaStandard error (assuming the null hypothesis)P Right horizontal fissure52.59%82.33%0.4140.060<0.01Upper segment of the right oblique fissure55.17%85.34%0.4350.061<0.01Lower segment of the right oblique fissure50.86%85.78%0.4520.060<0.01Upper segment of the left oblique fissure58.14%90.70%0.5420.064<0.01Lower segment of the left oblique fissure60.93%89.77%0.5630.064<0.01Right lung52.87%84.48%0.4370.035<0.01Left lung59.54%90.24%0.5560.061<0.01Whole lung55.42%88.67%0.4800.060<0.01 Preoperative interlobar fissure classification & intraoperative grading and scoring similarity analysis The bar chart demonstrates that the complete matching rate of interlobar fissure predictions exceeded 50% for all segments, while the proportion of predictions within one grade difference surpassed 80%. The accuracy of interlobar fissure predictions for the left lung was superior to that of the right lung. Specifically, the bar chart shows that the complete matching rate and the proportion of ‘differences within 1 grade’ in all left lung segments are higher than those in the right lung. Notably, the complete matching rate of the lower segment of the left oblique fissure exceeds 60%, whereas the proportion of ‘differences within 1 grade’ in the right horizontal fissure is only 82.33%, suggesting that clinical attention should be focused on preoperative assessment of the right horizontal fissure. Discussion The primary objective of this study was to evaluate the accuracy of HRCT and MPR 3D reconstruction techniques in predicting the development of pulmonary interlobar fissures. In clinical practice, incomplete interlobar fissures increase the complexity of VATS procedures, potentially leading to prolonged surgical times, increased blood loss, and higher postoperative complication risks. Patients with incomplete fissures exhibited higher rates of postoperative air leakage, resulting in prolonged chest drainage and extended hospital stays (11-13). Additional studies (11,14,15) indicate that incomplete fissures elevate the risk of postoperative air leakage by providing collateral ventilation pathways between adjacent lobes through fused parenchyma. The integrity of interlobar fissures also influences the localization of pulmonary lesions, distribution of pleural effusions, and the spread of inflammation and tumors, thereby impacting diagnosis and treatment (3,16). Therefore, accurate assessment of interlobar fissure development is crucial for surgical planning and risk management. HRCT and MPR provide high-resolution imaging and detailed anatomical information. MPR, in particular, offers critical insights into lesion localization, which is essential for preoperative risk assessment and surgical planning (17,18). However, discrepancies may arise between preoperative imaging and intraoperative findings, potentially attributable to anatomical variations or limitations in imaging technology (19). Despite these challenges, preliminary findings suggest that HRCT enhances diagnostic accuracy and surgical precision (20). This study analyzed data from 447 cases, revealing that a majority of preoperative and intraoperative fissure scores matched exactly, with most samples showing grade differences of ≤1. A significant correlation was observed between preoperative CT and intraoperative scores (P<0.01), with kappa coefficients indicating moderate agreement, suggesting HRCT and MPR offer predictive value in assessing fissure integrity preoperatively. A notable discrepancy was observed between preoperative CT and intraoperative scores for the right horizontal fissure, with a lower concordance rate and a smaller proportion of cases exhibiting grade differences of ≤1. This inconsistency is likely due to the size and irregularity of the right horizontal fissure, which often develops incompletely and presents complex morphology (21). Its close proximity to the right middle lobe bronchus and vasculature further obscures boundaries on axial CT slices. Additionally, overlapping lung tissue and lower tissue density at the horizontal fissure further reduce the accuracy of HRCT in identifying complete fissures (3). Variations in respiratory status and patient positioning during CT may also obscure fissure visualization, complicating assessment (22). Specifically, the right horizontal fissure lies in a region highly susceptible to diaphragmatic motion during breathing, leading to motion artifacts that further blur fissure edges on preoperative scans. The upper segment of the right oblique fissure predominantly scored between grades 2 and 3, reflecting the complexity of the anatomical structure, particularly in the upper portion, where developmental differences are more pronounced. The right lung’s overall anatomical structure is more intricate, with the upper oblique fissure closely associated with pulmonary vasculature and airways, increasing interpretation difficulty (2). In contrast, fissure predictions for the left lung demonstrated higher accuracy, with a higher concordance rate for the upper segment of the left oblique fissure and a larger proportion of cases showing grade differences of ≤1. The left lung’s simpler anatomy, absence of a horizontal fissure, and fewer anatomical variations provide a more stable reference for imaging evaluation (2), reducing assessment complexity and enhancing consistency between preoperative and intraoperative scores (23). The tendency for imaging-based scores to be lower than intraoperative assessments corresponds to the phenomenon of “imaging overestimation” (19), which is defined as a discrepancy in pulmonary interlobar fissure assessment where high-resolution imaging (e.g., HRCT) overestimates the degree of fissure development compared with direct intraoperative observation. This overestimation occurs because HRCT’s high spatial resolution magnifies subtle structural features (such as minor clefts or linear densities) that lack functional significance in vivo (21). Unlike intraoperative observation—where surgeons directly assess parenchymal fusion and rely on visual feedback to judge fissure integrity—HRCT may misclassify these minor structural features as more mature or complete fissures. This discrepancy underscores the need to correlate preoperative imaging findings with real-time surgical exploration, especially for complex fissure segments such as the right horizontal fissure (19). Recently, quantitative CT analysis software such as VIDA (VIDA Diagnostics, Coralville, IA, USA) and StratX (StratX Lung Analysis Platform, Pulmonx Corporation, Redwood City, CA, USA) has been developed to evaluate fissure integrity objectively (9). These automated tools are currently used to determine patient eligibility for bronchoscopic lung volume reduction (BLVR) with one-way endobronchial valves, where complete fissure integrity is a critical selection criterion to prevent interlobar collateral ventilation (24). The StratX platform utilizes validated algorithms to automatically segment lung lobes and calculate a Fissure Completeness Score (FCS), classifying fissures as complete (≥95% intact), partially complete (80–95% intact), or incomplete (<80% intact) (25). Similarly, VIDA software adopts a 90% threshold to distinguish complete from incomplete fissures (26). Studies have demonstrated that quantitative CT analysis can reduce inter-observer variability and provide standardized, reproducible assessments compared to visual analysis (9,27). While the present study employed a semi-quantitative MPR approach without these commercial automated tools, the fundamental methodology—volumetric assessment of fissure completeness based on high-resolution CT data—aligns with the principles underlying these software systems. Future integration of such automated quantitative tools may further standardize fissure evaluation, minimize subjective bias, and enhance predictive accuracy, particularly for complex segments like the right horizontal fissure where visual assessment remains challenging (9). This study adopted a modified fissure scoring method combining segmental assessment with a four-grade completeness classification. Three radiologists performed preoperative scoring, while three thoracic surgeons conducted intraoperative assessments, minimizing individual bias and enhancing score reliability. Averaging scores further ensured objectivity and consistency (28). Additionally, this study implemented a refined fissure grading system, assessing upper and lower segments of the right and left oblique fissures separately. This granular approach, unlike previous studies (3) that evaluated fissure completeness as a whole, highlights regional differences, enhancing predictive accuracy. However, this method introduces greater complexity and workload. Despite the inclusion of 447 patients, the sample size may limit the generalizability of results. Although independent assessments by multiple physicians reduced subjectivity, intraoperative grading by surgeons involved in the procedures may introduce bias. Furthermore, patient positioning and respiratory status during CT scans may affect image quality, posing challenges to fissure visualization. Lastly, the 4-grade fissure scoring system may lack the granularity needed for precise evaluation, potentially impacting scoring accuracy. Conclusions This study demonstrated a significant correlation (P<0.01) between preoperative HRCT/MPR scores and intraoperative fissure scores, underscoring the clinical value of HRCT and MPR in preoperative fissure assessment. For incomplete fissures, preoperative CT evaluation aids surgical planning, reducing the risk of postoperative air leakage and complications. Future research should expand sample sizes, incorporate multicenter studies, refine scoring methods, and consider additional variables to enhance accuracy and clinical applicability. Supplementary The article’s supplementary files as Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-0778/prf