|Year : 2017 | Volume
| Issue : 4 | Page : 693-698
Comparison of different width detector on the gross tumor volume delineation of the solitary pulmonary lesion
Dongping Shang, Jinbo Yue, Jianbin Li, Jinghao Duan, Yong Yin, Jinming Yu
Department of Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong University, Jinan 250117, Shandong, China
|Date of Web Publication||13-Sep-2017|
Department of Radiation Oncology, Shandong Cancer Hospital Affiliated to Shandong University, 440 Jiyan Road, Jinan 250117, Shandong
Source of Support: None, Conflict of Interest: None
Purpose: To explore the impact of different width detector on the volume and geometric position of gross tumor volume (GTV) of the solitary pulmonary lesion (SPL), as well as the impact on scanning time and radiation dose during the simulation.
Materials and Methods: Twenty-three patients with SPL underwent three-dimensional computed tomography (3DCT) simulation using different width detector, followed by four-dimensional computed tomography (4DCT) scans. GTV16 and GTV4 derived from different width detectors were compared with internal gross tumor volume (IGTV) generated from 4DCT on the volume and geometric position. Fourteen patients with lesions located in the upper lobe were defined as Group A and nine patients in the middle or lower lobe were defined as Group B. The scanning time and radiation dose during the simulation with the different width detector were compared as well.
Results: The volumes of IGTV, GTV16, and GTV4 in Group A were 13.86 ± 14.42 cm3, 11.88 ± 11.93 cm3, and 11.64 ± 12.88 cm3, respectively, and the corresponding volumes in Group B were 12.84 ± 11.48 cm3, 6.90 ± 6.63 cm3, and 7.22 ± 7.15 cm3, respectively. No difference was found between GTV16 and GTV4 in Groups A and B (PA = 0.11, PB = 0.86). Either GTV16 or GTV4 was smaller than IGTV (P16 = 0.001, P4 = 0.000). The comparison of the centroidal positions in x, y, and z directions for GTV16, GTV4, and IGTV showed no significant difference both in Groups A and B (Group A: Px = 0.19, Py = 0.14, Pz = 0.47. Group B: Px = 0.09, Py = 0.90, Pz = 0.90). The scanning time was shorter and radiation dose patient received was lower using 16 × 1.5 mm detector combination than 4 × 1.5 mm detector (P = 0.000).
Conclusions: Different width detector had no impact on the volume and geometric position of GTV of SPL during 3DCT simulation. Using wide detector would save time and decrease radiation dose compared with the narrow one. 3DCT simulation using either 16 × 1.5 mm detector or 4 × 1.5 mm detector could not cover all tumor motion information that 4DCT offered under free breathing conditions.
Keywords: Detector width, gross tumor volume, lung carcinoma, respiratory movement, tomography, X-ray computed
|How to cite this article:|
Shang D, Yue J, Li J, Duan J, Yin Y, Yu J. Comparison of different width detector on the gross tumor volume delineation of the solitary pulmonary lesion. J Can Res Ther 2017;13:693-8
|How to cite this URL:|
Shang D, Yue J, Li J, Duan J, Yin Y, Yu J. Comparison of different width detector on the gross tumor volume delineation of the solitary pulmonary lesion. J Can Res Ther [serial online] 2017 [cited 2017 Dec 11];13:693-8. Available from: http://www.cancerjournal.net/text.asp?2017/13/4/693/214464
| > Introduction|| |
Radiation therapy (RT) was one of the most important modalities for inoperable patients with pulmonary lesion. Some reasons for inoperability may be pulmonary compliance, cardiovascular disease, or poor performance status. However, during RT of pulmonary carcinomas, the tumor movement as a result of respiration could cause motion artifacts, distortion of tumor shape, and changes of tumor volume in tumor delineation. Pulmonary tumor delineation complicated by respiration-induced motion remains special challenge for radiation oncologists. The conventional approach for pulmonary tumor delineation employs additional margins to gross tumor volume (GTV) to ensure adequate tumor coverage based on the motion amplitudes., In recent years, the four-dimensional computed tomography (4DCT) has been used to contain tumor motion information and is becoming the gold standard for simulation and radiotherapy planning. It has been implemented in many clinics to evaluate the displacement of tumor and organ during breathe freely. However, because 4DCT for simulation and radiotherapy planning is not available in many radiotherapy centers, especially in developing country, the three-dimensional computed tomography (3DCT) is still commonly used for simulation and radiotherapy planning at present. Multi-row detector CT scanner has been applied to 3DCT radiotherapy simulation. The increase in the row of detector would accelerate the scanning speed and shorten simulation duration. Improvement in CT devices and techniques allows accurate target delineation with a concomitant reduction in dose delivery to adjacent organ at risk and normal tissue. Limiting dose to organs at risk and normal tissues requires the margin on the planning target volume is reduced. Consequently, precise patient imaging and GTV delineation are critical components in the treatment planning process for pulmonary carcinomas. The CT scanner for simulation in different medical institution is strikingly different [mainly referred to the width of the detector, [Figure 1]a and [Figure 1]b. Whether the difference in detector width would affect the target volume was unclear.
|Figure 1: Diagram of detector combination 4 × 1.5 mm (a) and 16 × 1.5 mm (b). The widths of the detector combinations were 6 mm and 24 mm, respectively|
Click here to view
In this study, we proposed and investigated the impact of the different width detector on volume, geometric position, and space matching relationship of GTVs of solitary pulmonary lesion (SPL) based on the standard of internal gross tumor volume (IGTV) generated from 4DCT and GTVabc, which was contoured on the images acquired using the active breath-controlled (ABC) technique. Then, we analyzed the difference of the scanning time and radiation dose that patient received using different width detector during 3D simulation.
| > Materials and Methods|| |
A total of 23 patients with peripheral stage I lung cancer or pulmonary metastasis were selected and treated consecutively with radiotherapy at Shandong Cancer Hospital from September 2013 to March 2015. All patients had SPL with no adhesion between the tumor edge and pleura. Fourteen patients with lesions located in the upper lobe were defined as Group A and nine patients in the middle or lower lobe were defined as Group B. This study was approved by Shandong Cancer Hospital Review Board, and all patients gave written informed consent to participate.
Computed tomography simulation and image acquisition
For two groups, patients were positioned supine in vacuum bag with their arms above head and were asked to breathe freely. With Philips Brilliance Big Bore CT scanner (Philips Medical Systems, Inc., Cleveland, OH, USA), the simulation scans were performed using the two different width detector combinations, i.e., 16 × 1.5 mm and 4 × 1.5 mm on the same region. The volume coverage spans with the two different width detectors of 16 × 1.5 mm and 4 × 1.5 mm were 24 mm and 6 mm in y-axis (superior-inferior [SI]) direction, respectively. The CT scanner was operated in a helical mode with an X-ray tube voltage peak of 120 kV, 200 mAs, and a slice thickness of 3 mm and an interval of 3 mm. In this study, all patients underwent proposed 4DCT scans for the region of interest under free breathing conditions and received 3DCT scans under ABC breath-holding status. During acquisition of the 4DCT images, the Real-Time Position Management system (Varian Medical Systems, Palo Alto, CA, USA) was used to monitor and record respiratory information. This system consists of a box with reflectors placed on the patient's abdomen near the umbilicus. Its motion was tracked by an infrared camera attached at the end of the tail of the scanning couch. The respiratory signal was transferred to the CT scanner, where the respiratory phases were matched with the acquired images. The multiple images at different table position were sorted together into the same phase images. Thus, ten sequence images of different respiratory phases from 0% to 90% were reconstructed on 4D raw data. The ABC breath-holding scans were performed at the deep inhalation phase with 75% trigger threshold.
Delineations of different target volumes
The images obtained by two different width detectors of 16 × 1.5 mm and 4 × 1.5 mm were loaded into Eclipse treatment planning system. In the same CT window setting (window width 1600 HU; window level − 600 HU), GTV16 and GTV4 of the two groups were delineated slice-by-slice on the two corresponding sets of images by the same clinical experienced radiation oncologist, respectively [Figure 2]. The volumes and isocenter coordinates in each set of images were calculated in the treatment planning system. The volumes of GTV16 and GTV4 derived from the two different width detectors were compared in both forms of the absolute value and the spatial conformity, which in the current study was defined as myocardial infarction (MI) = (GTV16∩GTV4)/(GTV4∪GTV16) (the symbols ∩ and ∪ were intersection and union, respectively). GTV0%, GTV10%,…, GTV90% were delineated on the ten series images of 4DCT and then were combined to produce IGTV [Figure 2]. GTVabc was delineated on the images obtained under ABC breath-hold conditions [Figure 2].
|Figure 2: Representative axial (A) and coronal (B) views of a corresponding contoured 16 × 1.5 mm detector-derived GTV16, 4 × 1.5 mm detector-derived GTV4, four-dimensional computed tomography-derived internal gross tumor volume, and active breath controlled technique-derived GTVabcof one patient. Internal gross tumor volume was notably larger than any other GTVs. GTV=Gross tumor volume|
Click here to view
For Groups A and B, the statistical analyses were done by the SPSS version 16.0 software package (SPSS Inc., Chicago, IL, USA). The Wilcoxon test was used to compare the tumor volumes and the isocenter coordinates of GTV16, GTV4, and IGTV. A Pearson correlation analysis was used to study the correlation between MI and GTVabc. The radiation dose that patient received during the simulation scan with the different width detector was compared by paired t-test. Variation was considered statistically significant if P < 0.05.
| > Results|| |
Detailed information about all analyzed patients and the tumor volumes is presented in [Table 1].
Difference on volume between gross tumor volume16, gross tumor volume4, and internal gross tumor volume
The volumes of IGTV, GTV16, and GTV4 in Group A were 13.86 ± 14.42 cm3, 11.88 ± 11.93 cm3, and 11.64 ± 12.88 cm3, respectively. For Group B, the volumes were 12.84 ± 11.48 cm3, 6.90 ± 6.63 cm3, and 7.22 ± 7.15 cm3, respectively. No difference was found between GTV16 and GTV4 both in Groups A and B (PA = 0.11, PB = 0.86). Either GTV16 or GTV4 was smaller than IGTV for the total 23 patients (P16 = 0.001, P4 = 0.000). The ratios of IGTV to GTV16 and GTV4 were 1.14 ± 0.07 and 1.24 ± 0.14 in Group A, respectively, while it was 2.02 ± 0.44 and 1.97 ± 0.26 in Group B. The MI between GTV16 and GTV4 was 0.62 ± 0.22 for the total 23 patients. There was a large interpatient's variation in MI, and the variation range was from 0.05 to 0.88. The correlation between MI and the volume of GTVabc is illustrated in [Figure 3]. There was a positive correlation between MI and GTVabc; the correlation coefficient was r = 0.478, P = 0.021. The MI between GTV16 and GTV4 correlated with the lesion location as well. It was 0.71 ± 0.17 in Group A compared with 0.50 ± 0.23 in Group B.
|Figure 3: Correlation of myocardial infarction and GTVabc (r = 0.48). GTV=Gross tumor volume|
Click here to view
Difference on geometric center between gross tumor volume16, gross tumor volume4, and internal gross tumor volume
Detailed information on the center of tumor coordinate for GTV16, GTV4, and IGTV is presented in [Table 2]. No differences were found on the geometric center in x, y, and z directions between them for the two groups [Table 3].
|Table 2: Center of tumor coordinate for GTV16, GTV4, and internal gross tumor volume|
Click here to view
|Table 3: Statistical comparison of tumor coordinates for GTV16, GTV4, and internal gross tumor volume included in Group A and Group B|
Click here to view
Difference on scanning time and radiation dose between different width detectors
For the same region, it took shorter time using the detector of 16 × 1.5 mm than 4 × 1.5 mm detector to complete the simulation (11.90 ± 0.72 s vs. 43.26 ± 2.90 s, P = 0.000). The radiation dose using the detector of 16 × 1.5 mm was lower than 4 × 1.5 mm detector (422.87 ± 25.75 mGy*cm vs. 515.03 ± 34.49 mGy*cm). The radiation dose patient received using the two kinds of combination was different significantly (P = 0.000).
| > Discussion|| |
In this study, we found that the different width detector had no impact on the volume and geometric position of GTV of SPL located in the upper or lower lobe during 3DCT simulation. CT scanning using either 16 × 1.5 mm detector or 4 × 1.5 mm detector might not cover all motion information compared with 4DCT scans.
The tumor simulation was completed momentarily by modern CT scanners. Therefore, GTV16 and GTV4 could not completely encompass respiration-induced tumor motion. We initially compared the volumes of GTV16, GTV4 with IGTV according to the lesions' location. The result showed that the motion magnitude contained in 3DCT images acquired synchronously within the transitory tube exposure time was more limited using the different width detector.
Using the ABC technique, breath-holding could be controlled by a valve. The tumor motion magnitude was much smaller than that of a free breathing study. ABC technique could effectively eliminate the interference of the respiratory movement. It is useful to obtain the accurate organ shape and volume relatively. In this study, the volumes of GTVabc were 11.53 ± 13.09 cm3 and 7.23 ± 7.42 cm3 in Groups A and B, respectively. Under normal free breathing, the volume and shape of tumor were subjected to breathing-induced motion. Due to the same slice thickness by the two kinds of detector combinations, the more images on which tumor appeared, the more tumor information was acquired by the detectors, the bigger GTV would be delineated. We analyzed the tumor motion information contained in GTV3D by comparing the volume of GTV16 with GTVabc and GTV4 with GTVabc. Comparison of GTV16 with GTVabc showed that there were all 16 patients whose GTV16 was larger than GTVabc, while seven smaller among the patients enrolled in this study. The ratios of GTV16 to GTVabc were 1.20 ± 0.19 and 0.99 ± 0.18 in Groups A and B, respectively. The similar result was shown in the comparison of GTV4 to GTVabc. There were 15 patients whose GTV4 was larger than GTVabc and eight patients smaller. The ratios of GTV4 to GTVabc were 1.08 ± 0.15 and 1.02 ± 0.31 in Groups A and B, respectively. The reason why part of GTV16 and GTV4 was larger than GTVabc could be explained plausibly. If the patients were breathing in and the tumor moved to the side of foot while the scanning couch moved to the side of head, that was to say, the tumor and scanning couch moved in the opposite direction during 3DCT simulation. Because of the detector located in the permanent position in y direction, the velocity that the tumor went through the detector dropped relative to the velocity during breath-hold status (V = Vcouch − Vtumor, Vtumor represented the velocity of tumor itself under free breathing conditions and Vcouch represented the velocity of scanning couch). The low translational velocity was benefit to obtain more tumor motion information. Hence, some of GTV16 and GTV4 acquired in the inspiration phase were larger than GTVabc. On the contrary, if the patients were breathing out and the tumor moved to the side of head when the scanning couch moved to the side of head, the tumor and scanning couch moved in the same direction. The velocity of tumor going through the detector increased relative to the velocity during breath-hold status (V = Vcouch + Vtumor). The duration that the detector acquired tumor images shortened. For these cases, the volumes of GTV16 and GTV4 were smaller than GTVabc, and the ratios of GTV16/GTVabc and GTV4/GTVabc were both <1.0. The quantitative value of ratio, no matter > or <1.0, might be expected depending on the velocity of the tumor going through the detector. The lower the velocity of tumor relative to the static detector was, the more motion information was contained in the GTV, and the higher the ratio of GTV16 or GTV4 to GTVabc would get. Oppositely, the faster the velocity was, the lower the ratio would be. Due to the fact that the two kinds of scan modes using different width detector were both performed in normal respiration, tumors moved in the same or opposite directions with the scanning couch randomly. Therefore, the whole result of this study demonstrated that there was no significant difference in volume and geometric position of the GTV16 and GTV4 generated from the different width detector.
4DCT provided a patient-specific characterization of respiration-induced tumor motion. The fused IGTV generated by the ten respiratory phases was the summation of the tumor position during the entire respiratory cycle. It provided not only the motion information but also defined the center location of tumor. 4DCT had been the clinical standard for quantifying respiratory motion in RT treatment. The ten sequence images generated by 4D raw data allowed visualization of motion and deformation of tissue during the breathing cycle. IGTV generated from 4DCT images contained the entire respiratory motion. The result of our study indicated that there were significant differences in volumes between IGTV and GTV16 (P16 = 0.001) and IGTV and GTV4 (P4 = 0.000). For 23 patients, the ratios of IGTV to GTV16 and GTV4 were 1.46 ± 0.51 and 1.54 ± 0.64, respectively. The result was consistent with other studies comparing IGTV and GTV3D in volumes. The difference in volume between IGTV and GTV3D was caused by the tumor motion in three dimensions. The ratio was related to the tumor motion amplitude, and the amplitude was related to the properties of tumor, such as tumor location (upper vs. lower lobe, peripheral vs. central tumor location) and stage (early vs. locally advanced stage). The tumor located in the lower lobes shifted greatly than tumor located in the upper and middle lobe. Tumors in the periphery of the lung moved 2.5 times as much as central tumors. The motion in the anterior-posterior, SI, and right-left directions could differ greatly. The motion amplitude of early-stage nonsmall cell lung cancer was larger than that of the locally advanced lung cancer. There were large interpatient variations in tumor movement also., Therefore, the optimal value for accurately defining target volume should be individualized using 4DCT technique. 3DCT images were acquired in a very short time using modern CT scanners. GTV16 and GTV4 were the temporary shape and location of the breathing cycle. The mere use of conventional margin to account for tumor motion cannot compensate for the additional effects that the implementation of 4DCT in target volume definition could offer. In this study, the centroids position of GTV16, GTV4, and IGTV had no statistical differences no matter where the tumor located in, but the volumes of GTV16 and GTV4 were much smaller than IGTV. Either GTV16 or GTV4 during 3DCT simulation could not cover all motion information compared with 4DCT scans.
Although there was no difference in volume and the centroidal coordinates between GTV16 and GTV4, there was still respiration-induced change in shape of tumor. We conducted target registration on GTV16 and GTV4. The MI between GTV16 and GTV4 varied largely from 0.05 to 0.88 after GTV was rectified. The MI value of 1.0 indicated a perfect similarity between the two GTVs, and the minimum was 0 if the two volumes were completely nonoverlapping. The MI was concerned with not only the tumor motion amplitude and morphological changes within the breathing cycle but also the tumor volume. In this study, the MI had a linear correlation with the volume of GTVabc; the correlation coefficient was r = 0.478. The bigger the tumor volume was, the higher MI between GTV16 and GTV4 would get.
The range covered by two different width detectors in y-axis direction was different. It covers the range of 24 mm by the detector combination of 16 × 1.5 mm in each cycle, while the range was only 6 mm by 4 × 1.5 mm detector. Therefore, it would take different time to complete the same simulation by different width detector. Given a tumor was approximately 24 mm in y-axis direction, it took one cycle to complete the scan using the detector of the 16 × 1.5 mm. However, it would take four cycles or more time to complete the scan using 4 × 1.5 mm detector combination.
The CT dose index of 4 × 1.5 mm detector was higher than that of 16 × 1.5 mm detector combination (15.9 mGy vs. 11.8 mGy). What's more, it took more time to complete the same simulation by 4 × 1.5 mm combination. Hence, the radiation dose that patients received using the narrow detector was higher than that of the wide detector brought about.
The present study did have limitations among them that the number of cases was limited, especially the lesions located in the lower lobe. Henceforth, more lesions located in middle and lower lobe will be enrolled in the further study.
| > Conclusions|| |
Our results demonstrated that different width detector had no significant impact on the volume and isocenter coordinates of GTV of SPL during 3DCT simulation. Using wide detector will save time and decrease radiation dose compared with narrow one. 3DCT simulation using either 16 mm × 1.5 mm detector or 4 mm × 1.5 mm detector could not cover all tumor motion information that 4DCT offered under free breathing conditions.
Financial support and sponsorship
This study is supported by Shandong Provincial Key Program of Research and Development (2015GSF118170). The study is supported partly by the Projects of Medical and Health Technology Development Program in Shandong Province (2016WS0561).
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Ram TS, Badkul R, Maraboyina S, Wang F. A comparative study to evaluate the efficacy of on board imaging with cone beam CT using target registration in patients with lung tumors undergoing stereotactic body radiation therapy and comparison with ExacTrac using skeletal registration on Novalis Tx. J Cancer Res Ther 2011;7:304-7.
Descovich M, McGuinness C, Kannarunimit D, Chen J, Pinnaduwage D, Pouliot J, et al.
Comparison between target margins derived from 4DCT scans and real-time tumor motion tracking: Insights from lung tumor patients treated with robotic radiosurgery. Med Phys 2015;42:1280-7.
Cai W, Hurwitz MH, Williams CL, Dhou S, Berbeco RI, Seco J, et al.
3D delivered dose assessment using a 4DCT-based motion model. Med Phys 2015;42:2897-907.
Callahan J, Kron T, Siva S, Simoens N, Edgar A, Everitt S, et al.
Geographic miss of lung tumours due to respiratory motion: A comparison of 3D vs. 4D PET/CT defined target volumes. Radiat Oncol 2014;9:291.
Wang L, Hayes S, Paskalev K, Jin L, Buyyounouski MK, Ma CC, et al.
Dosimetric comparison of stereotactic body radiotherapy using 4D CT and multiphase CT images for treatment planning of lung cancer: Evaluation of the impact on daily dose coverage. Radiother Oncol 2009;91:314-24.
Jang SS, Huh GJ, Park SY, Yang PS, Cho EY. The impact of respiratory gating on lung dosimetry in stereotactic body radiotherapy for lung cancer. Phys Med 2014;30:682-9.
Dou TH, Thomas DH, O'Connell DP, Lamb JM, Lee P, Low DA. A method for assessing ground-truth accuracy of the 5DCT technique. Int J Radiat Oncol Biol Phys 2015;93:925-33.
Cole AJ, O'Hare JM, McMahon SJ, McGarry CK, Butterworth KT, McAleese J, et al.
Investigating the potential impact of four-dimensional computed tomography (4DCT) on toxicity, outcomes and dose escalation for radical lung cancer radiotherapy. Clin Oncol (R Coll Radiol) 2014;26:142-50.
Kumar AS, Singh IR, Sharma SD, John S, Ravindran BP. Radiation dose measurements during kilovoltage-cone beam computed tomography imaging in radiotherapy. J Cancer Res Ther 2016;12:858-63.
Ezhil M, Vedam S, Balter P, Choi B, Mirkovic D, Starkschall G, et al.
Determination of patient-specific internal gross tumor volumes for lung cancer using four-dimensional computed tomography. Radiat Oncol 2009;4:4.
Dieleman EM, Senan S, Vincent A, Lagerwaard FJ, Slotman BJ, van Sörnsen de Koste JR. Four-dimensional computed tomographic analysis of esophageal mobility during normal respiration. Int J Radiat Oncol Biol Phys 2007;67:775-80.
Fukumitsu N, Hayashi Y. Application of a deformable registration technique to investigate breath-hold reproducibility. Jpn J Radiol 2014;32:700-7.
Adamson J, Zhuang T, Yin FF. Contour based respiratory motion analysis for free breathing CT. Comput Biol Med 2011;41:908-15.
Lewis JH, Jiang SB. A theoretical model for respiratory motion artifacts in free-breathing CT scans. Phys Med Biol 2009;54:745-55.
Hof H, Rhein B, Haering P, Kopp-Schneider A, Debus J, Herfarth K. 4D-CT-based target volume definition in stereotactic radiotherapy of lung tumours: Comparison with a conventional technique using individual margins. Radiother Oncol 2009;93:419-23.
Tai A, Liang Z, Erickson B, Li XA. Management of respiration-induced motion with 4-dimensional computed tomography (4DCT) for pancreas irradiation. Int J Radiat Oncol Biol Phys 2013;86:908-13.
Sharma PK, Srivastava R, Munshi A, Chomal M, Saini G, Garg M, et al.
Comparison of the gross tumor volume in end-expiration/end-inspiration (2 Phase) and summated all phase volume captured in four-dimensional computed tomography in carcinoma lung patients. J Cancer Res Ther 2016;12:47-52.
Kruis MF, van de Kamer JB, Belderbos JS, Sonke JJ, van Herk M. 4D CT amplitude binning for the generation of a time-averaged 3D mid-position CT scan. Phys Med Biol 2014;59:5517-29.
Nakamura M, Narita Y, Matsuo Y, Narabayashi M, Nakata M, Yano S, et al.
Geometrical differences in target volumes between slow CT and 4D CT imaging in stereotactic body radiotherapy for lung tumors in the upper and middle lobe. Med Phys 2008;35:4142-8.
Shang DP, Liu CX, Yin Y. A comparison of the different 3D CT scanning modes on the GTV delineation for the solitary pulmonary lesion. Radiat Oncol 2014;9:211.
Li FX, Li JB, Zhang YJ, Liu TH, Tian SY, Xu M, et al.
Comparison of the planning target volume based on three-dimensional CT and four-dimensional CT images of non-small-cell lung cancer. Radiother Oncol 2011;99:176-80.
Weiss E, Wijesooriya K, Dill SV, Keall PJ. Tumor and normal tissue motion in the thorax during respiration: Analysis of volumetric and positional variations using 4D CT. Int J Radiat Oncol Biol Phys 2007;67:296-307.
Britton KR, Starkschall G, Tucker SL, Pan T, Nelson C, Chang JY, et al.
Assessment of gross tumor volume regression and motion changes during radiotherapy for non-small-cell lung cancer as measured by four-dimensional computed tomography. Int J Radiat Oncol Biol Phys 2007;68:1036-46.
Jan N, Hugo GD, Mukhopadhyay N, Weiss E. Respiratory motion variability of primary tumors and lymph nodes during radiotherapy of locally advanced non-small-cell lung cancers. Radiat Oncol 2015;10:133.
Liu HH, Balter P, Tutt T, Choi B, Zhang J, Wang C, et al.
Assessing respiration-induced tumor motion and internal target volume using four-dimensional computed tomography for radiotherapy of lung cancer. Int J Radiat Oncol Biol Phys 2007;68:531-40.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]