Improving the preoperative diagnosis of invasive lung adenocarcinoma by combining multiplanar volume rendering (MPVR)-based solid component measurement with pathological analysis

Introduction

Lung cancer is the most commonly diagnosed cancer and the leading cause of cancer-related death worldwide (1). Lung adenocarcinoma (LUAD) is the most common histopathological subtype of lung cancer (2) and is classified into adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC). AIS and MIA can generally be managed with limited resection, whereas IAC often requires lobectomy with lymph node evaluation or dissection (3,4). Therefore, the accurate preoperative identification of IAC is crucial for clinical decision-making. However, the early diagnosis of LUAD remains challenging, and many patients present with advanced-stage disease or metastasis, complicating treatment (5).

Computed tomography (CT) is the most commonly used noninvasive tool for the preoperative diagnosis of early lung cancer. LUAD frequently presents as ground-glass nodules (GGNs). Based on the presence and size of invasive components, GGNs are classified as AIS, MIA, or IAC (6). The CT-measured solid component of GGNs is correlated with tumor invasiveness (7). Moreover, integrating solid component measurements with other CT features has achieved diagnostic accuracies of up to 80.3% (8). However, CT-based solid components may represent noninvasive pathological changes, such as fibrosis or scarring, rather than true invasive components (9). In addition, their visualization and measurement can be influenced by imaging acquisition and reconstruction techniques.

Multiplanar volume rendering (MPVR) is a three-dimensional CT post-processing technique that transforms volumetric CT datasets into reconstructed images based on voxel attenuation information and predefined visualization parameters. Unlike conventional axial CT images, MPVR simultaneously integrates information from multiple planes and displays the spatial distribution of tissue densities in pulmonary nodules. By assigning different colors to specific CT attenuation ranges, MPVR enhances the visualization of subtle density variations and internal architectural heterogeneity, potentially facilitating the identification of invasive components that may not be readily distinguishable on routine CT images. Parameter adjustment, particularly the optimization of CT attenuation thresholds, can further improve the delineation of invasive regions by emphasizing clinically relevant density ranges. Our previous study demonstrated that optimized MPVR parameters achieved a diagnostic accuracy of 80.28% for identifying IAC in GGNs (10). However, misclassification still occurred. Therefore, this study aimed to explore the pathological basis of MPVR misclassification, further optimize MPVR thresholds, and integrate conventional CT features to improve the accuracy of preoperative IAC diagnosis. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-1-0292/rc).

Methods

This retrospective study included 799 patients from The First Affiliated Hospital of Chongqing Medical University. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and was approved by the Ethics Committee of The First Affiliated Hospital of Chongqing Medical University (No. ZZ2025-009-01), which waived the requirement of informed consent due to the retrospective nature of this study.

Patients

The Pathology Information System was searched to identify patients who underwent surgical resection of pulmonary lesions classified as AIS, MIA, or IAC between October 2024 and December 2025. Preoperative imaging was manually reviewed on a Picture Archiving and Communication System (PACS) workstation. A total of 1,021 patients who underwent CT examinations within 2 weeks before surgery were identified. The inclusion criteria were as follows: (I) GGNs with a diameter <3 cm; (II) thin-section CT images with a slice thickness ≤1.25 mm; and (III) CT images without significant motion or respiratory artifacts and hematoxylin and eosin (H&E)-stained slides of adequate quality without major tissue folding and tearing. Ultimately, 799 patients with LUAD were included in the study. Of these, the data of 602 patients collected retrospectively between October 2024 and September 2025 constituted the training set, while the data of 197 patients collected prospectively from October 2025 to December 2025 constituted the test set.

The pathological growth patterns of LUAD include lepidic, acinar, papillary, micropapillary, cribriform, complex glandular, and solid patterns (11). The presence of any pattern other than lepidic, or evidence of vascular or pleural invasion, indicates tumor invasiveness. An invasive component measuring ≤5 mm is classified as MIA, while an invasive component >5 mm is classified as IAC (11). On CT, ground-glass opacities predominantly correspond to lepidic growth, whereas solid components suggest other patterns and may indicate invasiveness (12). Based on CT morphology, part-solid nodules (PSNs) were considered radiologically invasive-appearing lesions, whereas pure GGNs (pGGNs) were considered radiologically noninvasive-appearing lesions. Pathologically, AIS was classified as noninvasive, while MIA and IAC were classified as invasive lesions.

For this study, true-positive, false-positive, true-negative and false-negative classifications were defined according to the concordance between the conventional CT morphological classification (pGGN vs. PSN) and pathological diagnosis. These classifications were based solely on conventional CT appearance and were not derived from MPVR assessment or the overall radiological classification. According to the concordance between CT appearance and pathological findings, patients were categorized into four groups: the false-negative group, comprising 96 pGGNs with MIA or IAC; the true-negative group, comprising 122 pGGNs with AIS; the false-positive group, comprising 132 PSNs with AIS; and the true-positive group, comprising 449 PSNs with MIA or IAC. Representative cases from the four groups are shown in Figure 1. Clinical data, preoperative CT images, and postoperative histopathological slides were systematically collected for all study patients.

Figure 1 Representative case images for each group. (A1-A4) A false-negative case in which plain CT showed a pGGN, MPVR showed a PSN, and the pathological diagnosis was MIA. (B1-B4) A true-negative case in which plain CT and MPVR showed a pGGN, and the pathological diagnosis was AIS. (C1-C4) A false-positive case in which both plain CT and MPVR showed a PSN, and the pathological diagnosis was AIS. (D1-D4) A true-positive case in which both plain CT and MPVR showed a PSN, and the pathological diagnosis was IAC. Hematoxylin and eosin staining; magnification, 10× or 200×. AIS, adenocarcinoma in situ; CT, computed tomography; IAC, invasive adenocarcinoma; pGGN, pure ground-glass nodule; MIA, minimally invasive adenocarcinoma; MPVR, multiplanar volume rendering; PSN, part-solid nodule.

CT examinations

Chest CT examinations were performed using the SOMATOM Perspective (Siemens Healthineers, Erlangen, Germany), Discovery CT750 HD (GE Healthcare, Milwaukee, WI, USA), SOMATOM Definition Flash (Siemens Healthineers, Erlangen, Germany), SOMATOM Force (Siemens Healthineers, Erlangen, Germany), and Aquilion ONE pureViSION (Canon Medical System, Japan) CT scanners. All patients were positioned supine with their arms raised and instructed to hold their breath after deep inspiration to ensure optimal image acquisition. The scan range extended from the thoracic inlet to the costophrenic angles. The scanning parameters were as follows: tube voltage, 110–120 kV; tube current, 50–140 mAs (reference mAs determined using automatic tube current modulation); slice thickness, 5 mm; pitch, 1–1.1; collimation, 0.6 or 0.625 mm; reconstruction slice thickness and interval, 0.625, 1, or 1.25 mm; matrix size, 512×512; and rotation time, 0.5 seconds.

CT image analysis

MPVR reconstruction and analysis were performed using the post-processing module integrated into the Carestream Vue PACS (Carestream Health, USA). Pulmonary nodules were segmented from the surrounding lung parenchyma using a threshold of −800 Hounsfield units (HU), while solid components were distinguished from ground-glass opacity in the PSN view using a threshold of −350 HU (13,14). To optimize lesion visualization, low-density lung parenchyma was rendered with high transparency, while vascular structures and solid tumor components were progressively enhanced using opacity adjustment and color mapping. Transfer curve parameters were interactively adjusted to maximize the visual differentiation between ground-glass and solid components. The representative MPVR transfer function and opacity mapping settings are detailed in Figure S1.

The following CT features were assessed: (I) lesion diameter (the mean of the longest axial diameter and its perpendicular diameter); (II) lesion location (right upper, middle, or lower lobe; left upper or lower lobe); (III) lobulation; (IV) vacuole sign; (V) spiculation; (VI) pleural indentation; (VII) air bronchogram; and (VIII) the maximum diameter of the solid component (red area) measured on MPVR images. Lobulation was defined as a sudden bulging of the lesion margin; spiculation as linear projections extending from the nodule margin into the surrounding lung parenchyma; pleural indentation as a visible area of retraction or localized depression on the pleura; air bronchogram as the presence of air-filled bronchi against a background of consolidated lung parenchyma (15,16).

All nodules were independently re-measured by a third observer who was blinded to the previous results, and any discrepancies were resolved by consensus. Solid component size was measured on two-dimensional lung window images and MPVR images.

Histopathological examination

Two senior pathologists, each with over 10 years of diagnostic experience, performed the histological preparation and evaluation of all cases. Disagreements were resolved by consensus. A total of 620 H&E-stained slides were digitized into whole-slide images and analyzed using QuPath software to assess tumor growth patterns (lepidic, acinar, papillary, complex glandular, cribriform, micropapillary, and solid) and non-tumor components in lesions (e.g., inflammation and fibrosis) (17). The remaining 179 cases could not be digitized due to technical scanning failures and were thus reviewed manually under a microscope using a calibrated eyepiece micrometer. Measurements were obtained using a microscopic scale. Both methods used standardized measurement tools to ensure comparable accuracy. Histopathological reporting and pathological staging were performed according to the 8th edition of the American Joint Committee on Cancer staging system (11).

Statistical analysis

All statistical analyses were performed using R language (version 4.0). Continuous variables are expressed as median (interquartile range), while categorical variables are presented as frequencies and percentages. The intraclass correlation coefficient (ICC) and Cohen’s kappa coefficient were employed to assess interobserver agreement for continuous and categorical variables, respectively. Interobserver agreement, as indicated by ICC, was classified as poor (<0.500), moderate (0.500–0.740), good (0.750–0.890), or excellent (≥0.900) (18). The agreement based on kappa coefficients was categorized as poor (<0.000), slight (0.000–0.200), fair (0.210–0.400), moderate (0.410–0.600), substantial (0.610–0.800), or almost perfect (0.810–1.000) (18). Continuous data were analyzed using the Mann-Whitney U test. Categorical data were analyzed using Pearson’s Chi-squared test or Fisher’s exact test.

Univariate logistic regression was performed to identify potential predictors of IAC. Variables with P<0.05 were entered into a multivariate logistic regression model. Model performance was evaluated using receiver operating characteristic (ROC) curve analysis, with calculation of the area under the curve (AUC), sensitivity, and specificity. ROC curves for individual variables were generated based on predicted probabilities from univariate logistic regression models. Five-fold cross-validation was performed within the training set for internal validation and assessment of model stability, without variable selection or parameter tuning. Comparisons between ROC curves were performed using the DeLong test.

Results

Patients’ clinical characteristics and CT features

The patient selection process is shown in Figure 2. The clinical characteristics and CT features of all patients are summarized in Table 1, with no significant differences observed between the training and test sets. Interobserver agreement is presented in Table 2 and was good for all evaluated features. CT feature comparisons between the true-positive and false-positive groups and between the true-negative and false-negative groups are presented in Table 3.

Figure 2 The patient selection process. AIS, adenocarcinoma in situ; CT, computed tomography; IAC, invasive adenocarcinoma; LUAD, lung adenocarcinoma; MIA, minimally invasive adenocarcinoma; pGGNs, pure ground-glass nodules; PSNs, part-solid nodules.

Compared with the patients in the false-positive group, those in the true-positive group were older [58.0 (52.0–67.0) vs. 49.0 (42.0–57.0) years, P<0.001]. The true-positive group also had a higher proportion of women (71.2% vs. 59.9%, P=0.024), a larger MPVR-measured solid component diameter [8.2 (5.5–12.3) vs. 4.9 (4.2–5.6) mm, P<0.001], and higher rates of spiculation (39% vs. 22%, P<0.001) and pleural indentation (29.0% vs. 12.9%, P<0.001) than the false-positive group. Compared with the patients in the true-negative group, those in the false-negative group were older [56.0 (47.0–65.0) vs. 50.0 (44.0–58.0) years, P=0.008]. Although the median CT-measured and MPVR-measured solid component diameters were both 0.0 mm in the two groups, their overall distributions differed significantly (both Mann-Whitney U tests, P<0.05).

Histopathological features of different LUAD subtypes

In the false-positive group, 57 cases (43.2%) showed densely packed tumor structures, 52 (39.4%) had marked thickening of fibrous septa, 16 (12.1%) exhibited prominent fibrosis, and seven (5.3%) had intra-alveolar hemorrhage (Figure 3). In the false-negative group, 66 cases (68.8%) were MIA, 21 (21.9%) were well-differentiated IAC, and nine (9.3%) were moderately differentiated IAC. Additionally, all cases in the false-negative group had invasive components measuring <2 mm.

Figure 3 Representative case images of different pathological components in the false-positive group. (A1-A4) Axial CT (A1) and MPVR (A2) images show a PSN in the right lobe. pathological sections (A3,A4) reveal densely structured lesions. (B1-B4) Axial CT (B1) and MPVR (B2) images show a PSN in the right lobe; pathological sections (B3,B4) display significantly thickened alveolar septa. (C1-C4) Axial CT (C1) and MPVR (C2) images show a PSN in the left lobe; pathological sections (C3,C4) exhibit fibrosis. (D1-D4) Axial CT (D1) and MPVR (D2) images show a PSN in the right lobe; pathological sections (D3,D4) show evident hemorrhage. Hematoxylin and eosin, 10× or 200×. CT, computed tomography; MPVR, multiplanar volume rendering; PSN, part-solid nodule.

Model construction and evaluation

A logistic regression model was built using sex, age, CT-measured solid component diameter, MPVR-measured solid component diameter, spiculation, and pleural indentation. In the training set, the model achieved an AUC of 0.816 [95% confidence interval (CI): 0.78–0.851], with a sensitivity of 79.3% and a specificity of 70.2%. Five-fold cross-validation further confirmed the robustness and generalizability of the model, yielding a mean AUC of 0.807 (95% CI: 0.769–0.833), with a standard deviation of 0.043 across folds. In the test set, the model achieved an AUC of 0.778 (95% CI: 0.712–0.844), with a sensitivity of 75.0% and a specificity of 69.3%. ROC curve analysis in the training and test sets showed that the combined model achieved superior diagnostic performance compared with each individual feature, and that MPVR-measured solid component diameter outperformed CT-measured solid component diameter (Figure 4; all P<0.05).

Figure 4 Receiver operating characteristic curve analysis of the training (A) and test (B) sets. CT, computed tomography; MPVR, multiplanar volume rendering.

Discussion

This study retrospectively analyzed cases of misdiagnosis and missed diagnosis by MPVR and examined the pathological components of all lesions. The findings revealed that prominent fibrosis, inflammation, and densely packed tumor cells were major contributors to MPVR misclassification. In addition, a combined model incorporating sex, age, MPVR-measured solid component diameter, CT-measured solid component diameter, spiculation, and pleural indentation achieved superior diagnostic accuracy compared with any individual feature.

The appearance of ground-glass opacity and solid components on CT is closely associated with pathological growth patterns. Noninvasive growth is characterized by a lepidic pattern, with limited tumor cell proliferation and minimal involvement of alveolar spaces, typically appearing as ground-glass components on imaging. Invasive growth patterns, including acinar, papillary, micropapillary, and solid patterns, are associated with marked tumor cell proliferation that significantly disrupts the lung architecture, resulting in increased density and solid components on CT. When invasive components are minimal or small, false-negative cases may occur (19-21).

AIS is confined to the alveolar epithelium and exhibits a pure lepidic pattern, usually appearing as pGGNs. MIA is predominantly lepidic with ≤5 mm of invasion and typically presents as PSNs. IAC has invasive components >5 mm and may exhibit mixed growth patterns (including acinar, papillary, micropapillary, or solid patterns), also appearing as PSNs (11,22,23). However, solid components on imaging can also be influenced by fibrosis, dense noninvasive tumor cells, hemorrhage, and alveolar collapse, which cannot be reliably distinguished based on CT attenuation alone and may lead to false-positive findings. Therefore, while solid components are helpful for predicting the likelihood of IAC, they do not provide perfect accuracy; current imaging approaches aim to improve visualization and risk stratification to minimize preoperative misdiagnosis.

As a post-processing imaging technique, MPVR enables more accurate visualization and measurement of solid components in GGNs. On conventional CT, the appearance of solid components can be affected by vessels, bronchi, and other structures, and their boundaries are often indistinct, leading to measurement bias. In addition, the question of whether solid components should be measured using lung or mediastinal window settings has long been debated (24,25). MPVR addresses these limitations by assigning different colors, opacity levels, and grayscale shading to voxel properties, allowing independent adjustment of tissue visualization and clearer differentiation of structures, offering clear advantages over conventional CT.

Previous studies have shown that spiculation and pleural indentation on CT can aid in the diagnosis of IAC (14,26). However, these signs are not present in all cases, and their interpretation is highly subjective, depending on radiologists’ experience and varying criteria, which may limit their diagnostic value. In addition to conventional CT features, artificial intelligence has shown promise in the diagnosis of IAC (27,28); however, its interpretability and generalizability across populations require further validation. Conversely, MPVR offers strong interpretability and may facilitate preoperative visualization of invasive components in GGNs, making it a valuable research tool.

Although this study further explored the preoperative predictive value of MPVR by correlating imaging features with pathological components and included prospective validation, several limitations remain. First, postoperative pathological assessment relied on a single H&E-stained paraffin section, providing only a static and localized view of the tumor. Second, the relatively limited sample size may have introduced bias. Third, although CT and MPVR findings are useful for preoperative risk stratification and estimation of the likelihood of invasive behavior, they cannot establish a definitive diagnosis of LUAD or determine its histological subtype with certainty. Pathological examination remains the reference standard for diagnosis and histological classification. Fourth, CT images were acquired using different scanner platforms and reconstruction protocols. Although all examinations met the predefined imaging quality criteria, variations in acquisition and reconstruction parameters, including reconstruction slice thickness and interval, may have influenced MPVR visualization and measurement of solid components, thereby introducing measurement variability. Future studies using standardized CT acquisition and reconstruction protocols are warranted to further evaluate the robustness and reproducibility of MPVR measurements. Finally, as this was a single-center study, multicenter validation is still required.

Conclusions

MPVR-measured solid components in GGNs are useful for differentiating IAC from AIS/MIA and may facilitate clinical decision-making. Integrating sex, age, MPVR-measured solid component diameter, CT-measured solid component diameter, spiculation, and pleural indentation further improves the preoperative diagnostic accuracy for IAC.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-1-0292/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-1-0292/dss

Funding: This work was supported by the National Key R&D Program of China (No. 2020YFA0714002), the Chongqing Health Appropriate Technology Promotion Project (No. 2023jstg044), the Key Project of Technological Innovation and Application Development of Chongqing Science and Technology Bureau (No. CSTC2021jscxksbN0030), and the Joint Project of Chongqing Health Commission and Science and Technology Bureau (No. 2022ZDXM006).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-1-0292/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 and its subsequent amendments. The study was approved by the ethics committee of The First Affiliated Hospital of Chongqing Medical University (No. ZZ2025-009-01). Informed consent was waived due to the retrospective nature of this study.

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. Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, Jemal A. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Thai AA, Solomon BJ, Sequist LV, Gainor JF, Heist RS. Lung cancer. Lancet 2021;398:535-54. [Crossref] [PubMed]
3. Zhang Y, Ma X, Shen X, Wang S, Li Y, Hu H, Chen H. Surgery for pre- and minimally invasive lung adenocarcinoma. J Thorac Cardiovasc Surg 2022;163:456-64. [Crossref] [PubMed]
4. Sun G, Sun Y, Zou Z, Xu S. Analysis of Segmental Lymph Node Metastasis and Clinical Features in cT1N0M0 Lung Adenocarcinoma. Biomed Res Int 2020;2020:2842604. [Crossref] [PubMed]
5. Lu J, Tang H, Yang X, Liu L, Pang M. Diagnostic value and imaging features of multi-detector CT in lung adenocarcinoma with ground glass nodule patients. Oncol Lett 2020;20:693-8. [Crossref] [PubMed]
6. Su H, Chen L, Wu J, Cheng Z, Li J, Ren Y, et al. Proteogenomic characterization reveals tumorigenesis and progression of lung cancer manifested as subsolid nodules. Nat Commun 2025;16:2414. [Crossref] [PubMed]
7. Jing W, Liu M, Li W, Li D, Wu Y, Lv F. Prognostic implication of consolidation-to-tumor ratio in early lung adenocarcinoma: a retrospective cross-sectional study. Quant Imaging Med Surg 2024;14:3366-80. [Crossref] [PubMed]
8. Zhang Y, Shen Y, Qiang JW, Ye JD, Zhang J, Zhao RY. HRCT features distinguishing pre-invasive from invasive pulmonary adenocarcinomas appearing as ground-glass nodules. Eur Radiol 2016;26:2921-8. [Crossref] [PubMed]
9. Yamada N, Kusumoto M, Maeshima A, Suzuki K, Matsuno Y. Correlation of the solid part on high-resolution computed tomography with pathological scar in small lung adenocarcinomas. Jpn J Clin Oncol 2007;37:913-7. [Crossref] [PubMed]
10. Li WJ, Chu ZG, Li D, Jing WW, Shi QL, Lv FJ. Accuracy of solid portion size measured on multiplanar volume rendering images for assessing invasiveness in lung adenocarcinoma manifesting as subsolid nodules. Quant Imaging Med Surg 2024;14:1971-84. [Crossref] [PubMed]
11. Nicholson AG, Tsao MS, Beasley MB, Borczuk AC, Brambilla E, Cooper WA, Dacic S, Jain D, Kerr KM, Lantuejoul S, Noguchi M, Papotti M, Rekhtman N, Scagliotti G, van Schil P, Sholl L, Yatabe Y, Yoshida A, Travis WD. The 2021 WHO Classification of Lung Tumors: Impact of Advances Since 2015. J Thorac Oncol 2022;17:362-87. [Crossref] [PubMed]
12. Jiang L, Zhou Y, Miao W, Zhu H, Zou N, Tian Y, Pan H, Jin W, Huang J, Luo Q. Artificial intelligence-assisted quantitative CT parameters in predicting the degree of risk of solitary pulmonary nodules. Ann Med 2024;56:2405075. [Crossref] [PubMed]
13. Catelli C, Guerrini S, D’Alessandro M, Mazzei MA, Fiorelli A, Rosso L, Nosotti M, Marulli G, Dell’Amore A, Margaritora S, Leonardi B, Brascia D, Rea F, Lloret Madrid A, Giraudo C, Mathieu F, Congedo MT, Lococo F, Luzzi LGordon Study Group. Lymph Node Involvement and Radiologic Predictors of Invasiveness in Pure Ground-Glass Adenocarcinomas: The GORDON Study. Ann Thorac Surg 2026;121:1318-27. [Crossref] [PubMed]
14. Hu F, Huang H, Jiang Y, Feng M, Wang H, Tang M, Zhou Y, Tan X, Liu Y, Xu C, Ding N, Bai C, Hu J, Yang D, Zhang Y. Discriminating invasive adenocarcinoma among lung pure ground-glass nodules: a multi-parameter prediction model. J Thorac Dis 2021;13:5383-94. [Crossref] [PubMed]
15. Zwirewich CV, Vedal S, Miller RR, Müller NL. Solitary pulmonary nodule: high-resolution CT and radiologic-pathologic correlation. Radiology 1991;179:469-76. [Crossref] [PubMed]
16. Hansell DM, Bankier AA, MacMahon H, McLoud TC, Müller NL, Remy J. Fleischner Society: glossary of terms for thoracic imaging. Radiology 2008;246:697-722. [Crossref] [PubMed]
17. Bankhead P, Loughrey MB, Fernández JA, Dombrowski Y, McArt DG, Dunne PD, McQuaid S, Gray RT, Murray LJ, Coleman HG, James JA, Salto-Tellez M, Hamilton PW. QuPath: Open source software for digital pathology image analysis. Sci Rep 2017;7:16878. [Crossref] [PubMed]
18. Cannella R, Ronot M, Sartoris R, Cauchy F, Hobeika C, Beaufrere A, Trapani L, Paradis V, Bouattour M, Bonvalet F, Vilgrain V, Dioguardi Burgio M. Enhancing capsule in hepatocellular carcinoma: intra-individual comparison between CT and MRI with extracellular contrast agent. Diagn Interv Imaging 2021;102:735-42. [Crossref] [PubMed]
19. Carretta A, Bandiera A, Muriana P, Viscardi S, Ciriaco P, Gajate AMS, Arrigoni G, Lazzari C, Gregorc V, Negri G. Prognostic role of positron emission tomography and computed tomography parameters in stage I lung adenocarcinoma. Radiol Oncol 2020;54:278-84. [Crossref] [PubMed]
20. Solis LM, Behrens C, Raso MG, Lin HY, Kadara H, Yuan P, Galindo H, Tang X, Lee JJ, Kalhor N, Wistuba II, Moran CA. Histologic patterns and molecular characteristics of lung adenocarcinoma associated with clinical outcome. Cancer 2012;118:2889-99. [Crossref] [PubMed]
21. Yoon HJ, Park H, Lee HY, Sohn I, Ahn J, Lee SH. Prediction of tumor doubling time of lung adenocarcinoma using radiomic margin characteristics. Thorac Cancer 2020;11:2600-9. [Crossref] [PubMed]
22. Roy E, Shrager J, Benson J, Trope WL, Bhandari P, Lui N, Liou D, Backhus L, Berry MF. Risk of adenocarcinoma in patients with a suspicious ground-glass opacity: a retrospective review. J Thorac Dis 2022;14:4236-45. [Crossref] [PubMed]
23. Zhu J, Wang W, Xiong Y, Xu S, Chen J, Wen M, Zhao Y, Lei J, Jiang T. Evolution of lung adenocarcinoma from preneoplasia to invasive adenocarcinoma. Cancer Med 2023;12:5545-57. [Crossref] [PubMed]
24. Feng W, He T, Zhang Y, Mu Y, Pu Y, Liu Y, et al. Comparing the Efficacy of Empirical Anti-Infective Therapy and Follow-Up Observation for Newly Diagnosed Pulmonary Ground-Glass Nodules With Suspected Inflammatory Etiology: A Multicenter Prospective Observational Study Protocol. Health Sci Rep 2024;7:e70201. [Crossref] [PubMed]
25. Zhang C, Luan K, Li S, Wang Z, Chen S, Zhang W, Zhao C, Liu A, Jiao W. Different nodal upstaging rates and prognoses for patients with clinical T1N0M0 lung adenocarcinoma classified according to the presence of solid components in the lung and mediastinal windows. J Thorac Dis 2023;15:3612-26. [Crossref] [PubMed]
26. Cha MJ, Lee KS, Kim TJ, Kim HS, Kim TS, Chung MJ, Kim BT, Kim YS. Solitary Nodular Invasive Mucinous Adenocarcinoma of the Lung: Imaging Diagnosis Using the Morphologic-Metabolic Dissociation Sign. Korean J Radiol 2019;20:513-21. [Crossref] [PubMed]
27. Wu G, Woodruff HC, Shen J, Refaee T, Sanduleanu S, Ibrahim A, Leijenaar RTH, Wang R, Xiong J, Bian J, Wu J, Lambin P. Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter Study. Radiology 2020;297:451-8. [Crossref] [PubMed]
28. Zhu M, Yang Z, Wang M, Zhao W, Zhu Q, Shi W, Yu H, Liang Z, Chen L. A computerized tomography-based radiomic model for assessing the invasiveness of lung adenocarcinoma manifesting as ground-glass opacity nodules. Respir Res 2022;23:96. [Crossref] [PubMed]