Dual-energy CT multiparametric analysis improves diagnostic accuracy and reduces complications in percutaneous transthoracic needle biopsy
Jinhui Yao
, Jie Sun, Ning Han, Xuetao Zhang, Chong Lei, Jin Du
Contributions: (I) Conception and design: J Yao, J Sun, C Lei; (II) Administrative support: X Zhang; (III) Provision of study materials or patients: J Sun, C Lei, J Du, N Han; (IV) Collection and assembly of data: J Yao, J Sun; (V) Data analysis and interpretation: J Yao, C Lei; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.
Background: Lung cancer is a leading cause of cancer death worldwide. To plan treatment, doctors need an accurate tissue diagnosis. Percutaneous transthoracic needle biopsy (PTNB) is one of the main ways to obtain tissue samples, but the procedure often causes problems such as pneumothorax and bleeding in the lungs. Conventional contrast-enhanced computed tomography (CECT) is commonly used before biopsy to visualize the lesion’s morphology and location, but it does not reliably delineate small vessels or internal tumor heterogeneity. Dual-energy computed tomography (DECT) can provide more information by using two energy levels to create virtual monochromatic images and iodine maps. These data facilitate the selection of a safer and more accurate biopsy path by more clearly revealing blood supply and tissue composition. This study aimed to evaluate whether DECT multiparametric analysis can improve diagnostic accuracy and reduce procedure-related complications compared with CECT in PTNB.
Methods: A retrospective analysis was conducted on 165 patients who underwent PTNB. The patients were divided into two groups: Group A (n=98) underwent CECT, while Group B (n=67) received DECT with advanced material decomposition analysis. Comprehensive quantitative parameters, including virtual monochromatic imaging and iodine concentration maps, were utilized in Group B to guide individualized biopsy trajectory planning and risk assessment.
Results: In Group A, positive pathological results were obtained in 80 cases (81.6%), with pneumothorax and pulmonary hemorrhage occurring in 33 (33.7%) and 24 (24.5%) cases, respectively. In Group B, 62 cases (92.5%) yielded a positive pathological result, while 13 patients (19.4%) developed pneumothorax and 7 (10.4%) experienced pulmonary hemorrhage. The differences in diagnostic yield and complication rates between the two groups were statistically significant (P<0.05).
Conclusions: DECT-guided multiparametric quantitative analysis significantly improves the accuracy of PTNB and reduces procedure-related complications, demonstrating its clinical value in thoracic interventional imaging.
Keywords: Dual-energy computed tomography (DECT); image-guided percutaneous thoracic needle biopsy (image-guided PTNB); lung cancer; pathological diagnosis; complications
Submitted May 28, 2025. Accepted for publication Oct 21, 2025. Published online Nov 21, 2025.
doi: 10.21037/qims-2025-1246
Introduction
Lung cancer is one of the major malignant tumors that poses a threat to human health worldwide. The incidence rate (18.06%) and mortality rate (23.9%) of lung cancer rank first among all malignant tumors in China (1). Histopathological confirmation of suspected lesions combined with immunohistochemical profiling is considered the gold standard for accurate clinical decision-making. This serves as the cornerstone for individualized therapeutic strategies in contemporary oncological practice (2,3).
Percutaneous transthoracic needle biopsy (PTNB) is currently one of the primary methods for acquiring pathological specimens in lung cancer diagnosis, with pneumothorax (27%) and intrapulmonary hemorrhage (25.9%) being its most frequently observed complications (4,5). In clinical practice, contrast-enhanced computed tomography (CECT) is routinely used preoperatively to assess lesion characteristics, including anatomy, morphology, dimensions, and relationships with surrounding structures, reducing intraoperative bleeding risks (6).
Spectral computed tomography (CT), also known as dual-energy computed tomography (DECT), makes use of the fact that different materials attenuate X-rays differently based on their atomic numbers and densities. By using two different X-ray energy levels (typically 80 and 140 kVp), it can generate images that enhance the contrast of the image and provide more detailed information about tissue composition compared to traditional CT.
This technique enhances the visualization of fine vasculature, which may be inadequately depicted on conventional CECT scans (7). Additionally, DECT provides visualization of intratumoral iodine distribution, areas with high iodine concentration indicate a higher density of blood vessels, and tumor tissue activity in these areas tends to be relatively high (8). Therefore, we envision using DECT examination data to optimize the surgical plan to investigate whether it can reduce the incidence of intraoperative complications in PTNB. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1246/rc).
Methods
Research methodology
This study retrospectively analyzed data from 165 patients who underwent PTNB at our hospital, from April 2019 to October 2024. Inclusion criteria: (I) lesions larger than 3 cm in maximum diameter; (II) lesions confirmed as solid masses; (III) complete medical records and imaging data; and (IV) only one biopsy performed per puncture site.
All enrolled patients underwent CT scans within 1 week preoperatively. Group A (n=98) underwent CECT, while Group B (n=67) underwent DECT. All patients demonstrated normal preoperative coagulation parameters and platelet counts. Patients on long-term antithrombotic therapy stopped taking their medications for at least 5 days before the procedure to ensure stable coagulation.
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the ethics committee of Tangshan Gongren Hospital (No. 2023-040) and individual consent for this retrospective analysis was waived.
Biopsy method
All patients underwent CECT or DECT examination 1 week before surgery. Group A underwent preoperative CECT using Philips Brilliance iCT (Philips Healthcare, Best, The Netherlands), while Group B received DECT via GE Revolution CT (GE Healthcare, Chicago, USA). Procedures were planned on dedicated workstations by interventional radiologists. For Group B, 55 keV images were used for optimized vascular visualization, and 70 keV images for enhanced lesion delineation, with the shortest safe route avoiding vasculature selected. In Group A, biopsy targets were selected in areas of maximum enhancement; in Group B, regions with higher iodine values were prioritized. The surgical design and procedure were jointly completed by two experienced interventional radiologists (with ≥5 years of professional experience) and assisting medical staff. Under the guidance of a Philips Brilliance 64-slice CT, they performed all operations using an 18 G × 15 cm automatic biopsy needle (Biopince, Argon Medical, USA). After disinfection and local anesthesia, the biopsy needle was delivered to the lesion site via thoracentesis. After CT confirmation of the location, specimens were collected, fixed with formalin, and subjected to pathological analysis. A pressure dressing was applied postoperatively, and a CT scan was performed 30 minutes later to inspect for any complications.
The criteria for determining a positive diagnosis based on biopsy results are as follows: (I) for patients who underwent surgical treatment, the postoperative pathology is consistent with the post-PTNB pathology; (II) for patients who did not undergo surgery but chose other treatment methods, the follow-up results after 6 months of continuous treatment are consistent with the post-PTNB pathology.
Statistical methods
Statistical analysis was performed using SPSS 26.0 software. Non-normally distributed data (depth of puncture, procedural duration, number of needle passes) were expressed as median (interquartile range) and analyzed using the Mann-Whitney U test. Normally distributed data (needle insertion angle) were presented as mean ± standard deviation and compared using t-test. Categorical variables (gender, biopsy results, complications) were described as counts (percentages) and analyzed by χ2 test, with P<0.05 considered statistically significant.
Results
In Group A, pneumothorax occurred in 33 cases (33.7%), including one severe case requiring intervention for approximately 1,200 mL of air, while the remaining cases showed 30% or less lung involvement without significant symptoms. Intrapulmonary hemorrhage was observed in 24 cases (24.5%), with 7 patients experiencing hemoptysis and 6 developing minor hemothorax. All patients demonstrated favorable recovery during follow-up. Pathological analysis revealed malignant pulmonary tumors or metastases in 80 cases (81.6%) and benign or non-malignant lesions in 18 cases (18.4%). The median duration of the procedure and the number of needle passes were 19 minutes and 5, respectively.
In Group B, pneumothorax occurred in 13 cases (19.4%), all with 30% or less lung involvement and no severe symptoms; none required chest tube drainage. Intrapulmonary hemorrhage was noted in 7 cases (10.4%), including 3 with hemoptysis and 2 with minor hemothorax. All patients achieved complete recovery. Pathological diagnosis confirmed malignant pulmonary tumors or metastases in 62 cases (92.5%) and benign or non-malignant lesions in 5 cases (7.5%). The median duration of the surgery and the number of needle insertion attempts were 17 minutes and 5, respectively (Tables 1-5, Figures 1,2).
Table 1
| Variables | Total (n=165) | Group A (n=98) | Group B (n=67) | Z/χ2 | P |
|---|---|---|---|---|---|
| Gender | 0.214 | 0.644 | |||
| Man | 102 (61.8) | 62 (63.3) | 40 (59.7) | ||
| Woman | 63 (38.2) | 36 (36.7) | 27 (40.3) | ||
| Age, years | – | 68 (61 to 74.75) | 64 (60 to 70) | −1.848 | 0.065 |
| Pathological results | 3.945 | 0.047 | |||
| Negative | 23 (13.9) | 18 (18.4) | 5 (7.5) | ||
| Positive | 142 (86.1) | 80 (81.6) | 62 (92.5) |
Data are shown as n (%) or median (interquartile range). Group A: underwent CECT; Group B: underwent DECT. CECT, contrast-enhanced computed tomography; DECT, dual-energy computed tomography.
Table 2
| Variables | Total (n=165) | Group A (n=98) | Group B (n=67) | Z/χ2 | P |
|---|---|---|---|---|---|
| Pneumothorax | 4.03 | 0.045 | |||
| No pneumothorax | 119 (72.1) | 65 (66.3) | 54 (80.6) | ||
| Pneumothorax | 46 (27.9) | 33 (33.7) | 13 (19.4) | ||
| Intrapulmonary hemorrhage | 5.143 | 0.023 | |||
| No | 134 (81.2) | 74 (75.5) | 60 (89.6) | ||
| Yes | 31 (18.8) | 24 (24.5) | 7 (10.4) |
Data are shown as n (%). Group A: underwent CECT; Group B: underwent DECT. CECT, contrast-enhanced computed tomography; DECT, dual-energy computed tomography.
Table 3
| Variables | Group A (n=98) | Group B (n=67) |
|---|---|---|
| Adenocarcinoma | 45 | 38 |
| Squamous cell carcinoma | 19 | 12 |
| Small cell lung cancer | 7 | 6 |
| Large cell lung carcinoma | – | 1 |
| Adenosquamous carcinoma of the lung | – | 1 |
| Malignant neoplasm of lung, unspecified | 2 | – |
| Sarcoma | – | 1 |
| Lymphoma | 2 | – |
| Metastatic tumor of non-pulmonary origin | ||
| Tumors of reproductive system origin | 3 | – |
| Tumors of breast origin | 1 | 1 |
| Cancer of unknown primary | 1 | 2 |
| Fibrotic lung tissue | 3 | – |
| Fibroma | 1 | – |
| Inflammatory lung lesion | 6 | 3 |
| Tuberculosis | 2 | 1 |
| Necrosis | 5 | 1 |
| Pathological nature unclear due to sampling issues | 1 | – |
Group A: underwent CECT; Group B: underwent DECT. CECT, contrast-enhanced computed tomography; DECT, dual-energy computed tomography.
Table 4
| Variables | Group A | Group B | Z/t | P |
|---|---|---|---|---|
| Non-enhanced phase | ||||
| Average CT value (HU) | 43 (36 to 49) | 40 (36 to 45.5) (70 keV) | ||
| Arterial phase | ||||
| Average CT value (HU) | 65 (53 to 80.75) | 72.2 (66 to 84.2) (70 keV) | ||
| IC (μg/cm3) | 21.6 (16.9 to 28.8) | |||
| NIC | 0.16 (0.13 to 0.22) | |||
| Spectral curve slope | −3.34 (−2.89 to −4.09) | |||
| Venous phase | ||||
| Average CT value (HU) | 67.5 (55.25 to 77) | 70.7 (60.9 to 84.7) (70 keV) | ||
| IC (μg/cm3) | 21.0 (16.8 to 24.9) | |||
| NIC | 0.43 (0.37 to 0.52) | |||
| Spectral curve slope | −3.59 (−2.89 to −3.94) | |||
| Effective atomic number | 7.28 (7.185 to 7.385) | |||
| Pulmonary emphysema | 26 (26.5) | 16 (23.9) | 0.701 | |
| The maximum diameter of the tumor (mm) | 50.88±19.02 | 50.12±17.05 | −0.120 | 0.905 |
Data are presented as mean ± standard deviation, n (%) or median (interquartile range). Pour: NIC = IC of the lesion area / IC of the aorta at the same level; spectral curve slope= (HU70 keV − HU40 keV)/[70−40]. Group A: underwent CECT; Group B: underwent DECT. CECT, contrast-enhanced computed tomography; CT, computed tomography; DECT, dual-energy computed tomography; HU, Hounsfield units; IC, iodine concentration; NIC, normalized iodine concentration.
Table 5
| Variables | Group A | Group B | Z/t | P |
|---|---|---|---|---|
| Needle insertion angle (°) | 89.99±20.36 | 93.06±12.33 | −1.204 | 0.230 |
| Pleura to puncture endpoint distance (mm) | 32 (23 to 42) | 34 (23.5 to 43) | −0.05 | 0.960 |
| Skin surface to puncture endpoint distance (mm) | 68.5 (54.25 to 74) | 66 (51 to 80) | −0.199 | 0.842 |
| Puncture procedure duration (min) | 19 (15 to 26) | 17 (13 to 21) | −2.452 | 0.014 |
| Number of needle passes | 5 (4 to 7) | 5 (4 to 6) | −2.266 | 0.023 |
Data are presented as mean ± standard deviation or median (interquartile range). Group A: underwent CECT; Group B: underwent DECT. CECT, contrast-enhanced computed tomography; DECT, dual-energy computed tomography.
Discussion
Lung cancer represents the most prevalent malignant tumor, and PTNB serves as a crucial clinical method for obtaining pathological tissues. However, the frequent presence of necrosis, pneumonia, or atelectasis within lung tumors complicates the distinction of viable tumor tissue. This difficulty, in turn, compromises biopsy accuracy and can yield unreliable pathological results (9). CECT has limitations in differentiating various intratumoral components. In clinical surgery with complex tumor composition, CECT struggles to achieve effective distinction (10,11). DECT uses high- and low-energy data to differentiate tissues by atomic number and quantifies material properties, such as iodine concentration. This helps identify viable tumor areas and delineate tumor boundaries from surrounding atelectasis, providing similar functional information to PET-CT (12,13). DECT not only delineates lesion boundaries from surrounding tissues, but also assists in differentiating benign and malignant lesions (14). Group B demonstrated 10.9% higher malignancy detection rate compared to Group A, with statistical significance (P<0.05). DECT quantitative parameters (e.g., low keV images) help radiologists to identify representative regions of the tumor due to enhancement of the contrast. These parameters distinguish enhancing tumor tissue from necrosis and adjacent lung consolidation, which supports more accurate and targeted biopsy sampling.
Pneumothorax represents the most common complication of PTNB. Evidence suggest that its occurrence may correlate with traversal through areas of pulmonary emphysema during the puncture procedure, while the incidence probability demonstrates associations with procedural duration, needle path length, number of passes, and insertion angle (4,15,16). Although most patients experience spontaneous resolution without intervention, a subset requires postprocedural drainage (17). In this study, one patient in Group A developed intraoperative pneumothorax of approximately 1,200 mL and received postoperative treatment, indicating the necessity of preventing PTNB-related pneumothorax. The incidence of pneumothorax in Group B was 14.3% lower than in Group A, showing statistical significance (P<0.05). This reduction might relate to shorter puncture duration and fewer needle passes. No significant differences existed between Groups A and B in puncture angle or depth parameters (including skin-to-target and pleura-to-target distances). However, Group B demonstrated statistically significant reductions in both procedural time (P<0.05) and number of passes (P<0.05) compared with Group A. Therefore, we believe that DECT can help optimize the puncture path to reduce the duration of surgery and the number of needle insertion for patients (Figures 3-5).
Figures 1,2 show that longer procedural time and a more needle passes are associated with higher complication rates. Longer procedures and multiple punctures can damage pulmonary vessels and alveolar structures. Compared with Group A, Group B had shorter procedure time and fewer passes. These improvements likely contributed to the lower pneumothorax and hemorrhage rates observed in Group B.
Pulmonary hemorrhage is the second most common complication of PTNB. According to the study of Wu et al., malignant tumors often contain abundant blood vessels, so the puncture path should avoid these vascular areas (15). DECT uses instantaneous high-low voltage switching technology (140 and 80 kV) for imaging, reconstructing images at energy levels from 40 to 140 keV. Machida et al. have demonstrated that reducing keV levels increases iodine’s X-ray absorption, thereby enhancing vascular contrast and improving the visualization of small-caliber vessels (18).
Therefore, DECT enables the visualization of vascular networks through low-keV imaging, which can be seamlessly integrated with lesion to optimize surgical path planning and avoid small vessels, thereby minimizing the risk of hemorrhage. In this study, Group B had 13.4% fewer cases of bleeding in the lungs compared to Group A with a significant difference (P<0.05).
Inadequacy and reflection
Due to limitations at this stage of our research, only a preliminary analysis of different pathological types of pulmonary tumors was conducted. In this cohort, adenocarcinoma, squamous cell carcinoma, and small cell lung cancer were markedly more prevalent than other pathological subtypes. Given this imbalance in subtype distribution, further analysis of complication rates and diagnostic accuracy across all subtypes may lead to bias and reduce the reliability of the findings. Therefore, detailed evaluation of complication patterns among different tumor types was not pursued in the present study. Further investigation on this topic is ongoing and will be reported in a separate study.
Moreover, despite the surgeon’s considerable experience, variations in their individual preferences and judgments when choosing biopsy plans could not be eliminated. Although we attempted to mitigate this potential bias by maximizing the sample size, the final cohort was still limited due to factors such as patients refusing to continue treatment, transferring to other hospitals, and unclear medical records.
Conclusions
In conclusion, DECT improves the positive rate of lung cancer diagnosis in PTNB, and helps mitigate the risk of pulmonary hemorrhage, demonstrating its valuable role in clinical practice.
Acknowledgments
We sincerely appreciate Dr. Lu Yu from Tangshan Women and Children’s Hospital, Hebei Province, for his invaluable assistance during the manuscript writing process. We also extend our gratitude to Dr. Sheng Ren from the Institute of Medical Biotechnology, Chinese Academy of Medical Sciences, for his support in translation and refinement.
Footnote
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1246/rc
Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1246/dss
Funding: This study was supported by
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1246/coif). J.Y., J.S., and X.Z. report that this study was supported by the Health Commission of Hebei Province (No. 2024-2019). The other 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 Tangshan Gongren Hospital with the approval document number 2023-040 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
- Xia C, Dong X, Li H, Cao M, Sun D, He S, Yang F, Yan X, Zhang S, Li N, Chen W. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chin Med J (Engl) 2022;135:584-90. [Crossref] [PubMed]
- Lam DC, Liam CK, Andarini S, Park S, Tan DSW, Singh N, Jang SH, Vardhanabhuti V, Ramos AB, Nakayama T, Nhung NV, Ashizawa K, Chang YC, Tscheikuna J, Van CC, Chan WY, Lai YH, Yang PC. Lung Cancer Screening in Asia: An Expert Consensus Report. J Thorac Oncol 2023;18:1303-22. [Crossref] [PubMed]
- Chen P, Liu Y, Wen Y, Zhou C. Non-small cell lung cancer in China. Cancer Commun (Lond) 2022;42:937-70. [Crossref] [PubMed]
- Huo YR, Chan MV, Habib AR, Lui I, Ridley L. Pneumothorax rates in CT-Guided lung biopsies: a comprehensive systematic review and meta-analysis of risk factors. Br J Radiol 2020;93:20190866. [Crossref] [PubMed]
- Khan MF, Straub R, Moghaddam SR, Maataoui A, Gurung J, Wagner TO, Ackermann H, Thalhammer A, Vogl TJ, Jacobi V. Variables affecting the risk of pneumothorax and intrapulmonal hemorrhage in CT-guided transthoracic biopsy. Eur Radiol 2008;18:1356-63. [Crossref] [PubMed]
- Guo Z, Shi H, Li W, Lin D, Wang C, Liu C, et al. Chinese multidisciplinary expert consensus: Guidelines on percutaneous transthoracic needle biopsy. Thorac Cancer 2018;9:1530-43. [Crossref] [PubMed]
- Gruschwitz P, Petritsch B, Schmid A, Schmidt AMA, Grunz JP, Kuhl PJ, Heidenreich JF, Huflage H, Bley TA, Kosmala A. Noise-optimized virtual monoenergetic reconstructions of dual-energy CT angiographies improve assessability of the lower leg arterial segments in peripheral arterial occlusive disease. Radiography (Lond) 2023;29:19-27. [Crossref] [PubMed]
- Aoki M, Takai Y, Narita Y, Hirose K, Sato M, Akimoto H, Kawaguchi H, Hatayama Y, Miura H, Ono S. Correlation between tumor size and blood volume in lung tumors: a prospective study on dual-energy gemstone spectral CT imaging. J Radiat Res 2014;55:917-23. [Crossref] [PubMed]
- Rivera MP, Mehta AC, Wahidi MM. Establishing the diagnosis of lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2013;143:e142S-65S.
- van Elmpt W, Landry G, Das M, Verhaegen F. Dual energy CT in radiotherapy: Current applications and future outlook. Radiother Oncol 2016;119:137-44. [Crossref] [PubMed]
- Vlahos I, Jacobsen MC, Godoy MC, Stefanidis K, Layman RR. Dual-energy CT in pulmonary vascular disease. Br J Radiol 2022;95:20210699. [Crossref] [PubMed]
- Schmid-Bindert G, Henzler T, Chu TQ, Meyer M, Nance JW Jr, Schoepf UJ, Dinter DJ, Apfaltrer P, Krissak R, Manegold C, Schoenberg SO, Fink C. Functional imaging of lung cancer using dual energy CT: how does iodine related attenuation correlate with standardized uptake value of 18FDG-PET-CT? Eur Radiol 2012;22:93-103. [Crossref] [PubMed]
- Hartley-Blossom ZJ, Digumarthy SR. Dual-Energy Computed Tomography Applications in Lung Cancer. Radiol Clin North Am 2023;61:987-94. [Crossref] [PubMed]
- Vulasala SSR, Wynn GC, Hernandez M, Kadambi I, Gopireddy DR, Bhosale P, Virarkar MK. Dual-Energy Imaging of the Chest. Semin Ultrasound CT MR 2022;43:311-9. [Crossref] [PubMed]
- Wu CC, Maher MM, Shepard JA. Complications of CT-guided percutaneous needle biopsy of the chest: prevention and management. AJR Am J Roentgenol 2011;196:W678-82. [Crossref] [PubMed]
- Hong W, Yoon SH, Goo JM, Park CM. Cone-Beam CT-Guided Percutaneous Transthoracic Needle Lung Biopsy of Juxtaphrenic Lesions: Diagnostic Accuracy and Complications. Korean J Radiol 2021;22:1203-12. [Crossref] [PubMed]
- Sharma A, Shepard JO. Lung Cancer Biopsies. Radiol Clin North Am 2018;56:377-90. [Crossref] [PubMed]
- Machida H, Tanaka I, Fukui R, Shen Y, Ishikawa T, Tate E, Ueno E. Dual-Energy Spectral CT: Various Clinical Vascular Applications. Radiographics 2016;36:1215-32. [Crossref] [PubMed]
