Characterization of 3D printing techniques: Toward patient specific quality assurance spine-shaped phantom for stereotactic body radiation therapy
Characterization of 3D printing techniques: Toward patient specific quality assurance spine-shaped phantom for stereotactic body radiation therapy
Min-Joo Kim 0 1
Seu-Ran Lee 0 1
Min-Young Lee 0 1
Jason W. Sohn 1
Hyong Geon Yun 1
Joon Yong Choi 1
Sang Won Jeon 1
Tae Suk Suh 0 1
0 Department of Biomedical Engineering, Research Institute of Biomedical Engineering, College of Medicine, The Catholic University of Korea , Seoul , Korea , 2 Department of Radiation Oncology, Yonsei Cancer Center, Yonsei University College of Medicine , Seoul , Korea , 3 Department of Radiation Oncology, School of Medicine, Case Western Reserve University , Cleveland , Ohio, United States of America, 4 Department of Radiation Oncology, College of Medicine, DongGuk University Hospital , Goyang , Korea
1 Editor: Chun Kee Chung, Seoul National University College of Medicine , REPUBLIC OF KOREA
Development and comparison of spine-shaped phantoms generated by two different 3Dprinting technologies, digital light processing (DLP) and Polyjet has been purposed to utilize in patient-specific quality assurance (QA) of stereotactic body radiation treatment. The developed 3D-printed spine QA phantom consisted of an acrylic body phantom and a 3Dprinted spine shaped object. DLP and Polyjet 3D printers using a high-density acrylic polymer were employed to produce spine-shaped phantoms based on CT images. Image fusion was performed to evaluate the reproducibility of our phantom, and the Hounsfield units (HUs) were measured based on each CT image. Two different intensity-modulated radiotherapy plans based on both CT phantom image sets from the two printed spine-shaped phantoms with acrylic body phantoms were designed to deliver 16 Gy dose to the planning target volume (PTV) and were compared for target coverage and normal organ-sparing. Image fusion demonstrated good reproducibility of the developed phantom. The HU values of the DLP- and Polyjet-printed spine vertebrae differed by 54.3 on average. The PTV Dmax dose for the DLP-generated phantom was about 1.488 Gy higher than that for the Polyjetgenerated phantom. The organs at risk received a lower dose for the 3D printed spineshaped phantom image using the DLP technique than for the phantom image using the Polyjet technique. Despite using the same material for printing the spine-shaped phantom, these phantoms generated by different 3D printing techniques, DLP and Polyjet, showed different HU values and these differently appearing HU values according to the printing technique could be an extra consideration for developing the 3D printed spine-shaped phantom depending on the patient's age and the density of the spinal bone. Therefore, the 3D printing technique and materials should be carefully chosen by taking into account the condition of the patient in order to accurately produce 3D printed patient-specific QA phantom.
Competing interests: The authors have declared
that no competing interests exist.
Nowadays, advanced radiotherapy such as stereotactic body radiation therapy (SBRT)
delivered high radiation dose into a small size of the tumor region using highly elaborated radiation
fluence, and thus, patient-specific quality assurance (QA) plays an important role [
Tumors of the spine are an especially common malignancy, and spinal tumors could cause
neurological disabilities including pain. The SBRT could be performed in patients with spinal
tumors, and the target volume for radiation treatment such as spine SBRT unavoidably
includes the spinal cord during the radiation treatment planning process [3±5]. Thus, spine
SBRT requires the inclusion of a steep dose gradient in delivered dose distribution, high
prescription dose, small size of radiation fields, and extra image guidance [
] to exclude the
spinal cord from the delivered dose distribution. To verify and increase the accuracy of spine
SBRT and also most of the radiation treatment, the QA process using a specialized
patient-specific phantom has become increasingly important since patient-specific QA using a highly
customized patient-specific phantom could clearly determine the accuracy of radiation treatment
Accurate and cost-effective production of these patient-specific phantoms requires the use
of the latest technology, the three-dimensional (3D) printing technique. The advent of 3D
printers has led to a growing interest in various fields. In particular, the characteristics of 3D
printing, such as the versatility and variety of materials for 3D printing as well as the ability to
customize products with the desired geometrical features are being promoted to utilize this
latest technology in various fields and these merits of 3D printing have been recently integrated
into the field of medical physics, especially in the development of bolus, compensators, and
Various types of 3D printing technologies are available to researchers, including
stereolithography (SLA or SL), fused deposition modelling (FDM), selective laser sintering (SLS),
Polyjet printing, and digital light processing (DLP) [
]. SLA was the first 3D printing
technology to be developed and it involves focusing a concentrated beam of ultraviolet (UV) light on
the surface of a vat filled with a liquid photopolymer [
]. Recently, Bache et al. made a positive
mold in the shape of a rodent using this technology [
]. FDM is also a relatively old 3D
printing technology; however, and it is commonly used due to its lowest cost. This type of 3D
printing uses thermoplastic materials such as acrylonitrile butadiene styrene (ABS, density 1.05 g/
cm3) or poly lactic acid (PLA, density 1.25 g/cm3) [
]. There are a few papers involving FDM
3D printing and these studies have resulted in the development of various types of phantoms
for application in the field of medical physics such as the development of a phantom including
], low-density materials to simulate the patient's lung [13±15], and in magnetic
resonance imaging (MRI) for breast phantoms [
]. The advanced printing method is SLS,
which involves the use of a high-power laser to fuse plastic or metal powders into the desired
3D shape [
]. Madamesila et al. evaluated low-density phantoms by comparing FDM and SLS
]. Lastly, DLP and Polyjet printing techniques appear to be appropriate for construction of
patient-specific QA phantoms consisting of high-density materials with high output
resolution. There have been a few publications regarding the Polyjet technique: Design of a
patientspecific phantom for liver [
], neurovascular model [
], imaging phantom for the evaluation
of a new CT reconstruction algorithm [
]. On the other hand, to the best of our
knowledge, the DLP printing technique has never been used in the medical physics field.
While 3D printing has been utilized in various fields to produce highly customized objects
as described above, the phantom developed from the 3D printing technique is available only
for the soft tissue or the applied low-density materials in the medical physics field. In this
work, 3D printed objects that mimic the human spine using high-density materials with DLP
2 / 12
and Polyjet printing techniques were produced and each characteristic of these two different
patient-specific spine phantom sets was investigated by comparing treatment radiation doses
and Hounsfield units (HUs) using the corresponding CT image sets.
Materials and methods
3D printer technology
Two types of 3D printers were used in this study: a DLP printer and a Polyjet printer. The DLP
printing technology uses a more conventional light source, such as an arc lamp with a
liquidcrystal display panel or a deformable mirror device which is applied to the entire surface of the
vat of photopolymer resin in a single pass. The DLP printing technique also produces highly
accurate parts with excellent resolution. The main advantages of this technique are that DLP
technique can harden a whole layer in a fraction of the time and it takes to a laser to trace
around and fill in each item on the print bed. With use of this production method, DLP can
provide higher printing speeds at a relatively low cost [
]. The Polyjet 3D printer is similar to
an ink jet printer; however, this technique applies resins instead of ink. The resin is laid down
on a print bed layer by layer and then it is hardened using UV light. Some Polyjet machines
can make a combination of hybrid materials because of their multiple print heads. Printing
materials used for Polyjet technique are very diverse, ranging from hard plastics to soft rubber
The characteristics of each printer used in this study are summarized in Table 1. For the
DLP technique, Titan 1 3D printer manufactured by Kudo3D (Pleasanton, CA, USA) was
used in this investigation. This printer yields resolution values of 30 to 70 μm in the x and y
directions, and a resolution of 5 μm in the z-direction [
]. In case of the Polyjet technique,
the phantom was printed using an Objet Connex 3D printer (Objet Geometries, Rehovot,
Israel), which provides a z resolution of 16±30 μm and accuracy of 20±80 μm [
]. For both
technology, the same main material, acrylic polymer with a density of 1.29±1.39 g/cm3 [
was applied and the density of this acrylic polymer is most similar to the density of the human
Spine modelling process for 3D printing
Fig 1 shows the workflow for 3D printing to generate the human spine. Volumetric CT data
were exported as 1.25 mm axial slices in the DICOM (.dcm) format to Seg3D and ImageVis3D
(Scientific Computing and Imaging Institute, University of Utah, Salt Lake City, UT, USA),
which are open-source software packages for volumetric image segmentation, volume
rendering, and visualization. The spines were segmented and a 3D iso-surface was generated for each
spine by generating a series of triangles using the above two software packages which can
convert and export the stereolithography (STL) file format. The STL files were imported into
Autodesk Meshmixer software (Autodesk Inc., San Rafael, CA, USA) to divide each spine and
to convert the files into design web format (DWF) file formats. The DWF files were imported
into SolidWorks 3D CAD software (SolidWorks Corp., Concord, MA, USA) to simulate spinal
3 / 12
Fig 1. Diagram of the workflow framework that we followed for producing 3D printed spine-shaped
phantom from CT data.
cancer by producing holes on the first lumbar vertebra (L1). The edited files were exported in
the STL file format. Lastly, the twelfth thoracic vertebra (T12), L1 with tumors, and second
lumber vertebra were divided laterally exactly into half using the Autodesk Meshmixer
software to possibly measure the doses delivered to the inside of the vertebrae by inserting a
Development of a patient-specific spine QA phantom
The body phantom was constructed with an acrylic and it was the size of the human abdomen
on average [
], and it consisted of five slabs and the gap between two slabs would be utilized
for subsequent application by inserting dosimetry films. Since the human spine is located 2 cm
from the back on average, the body phantom was also punctured at intervals of 2 cm from the
back to insert a cylindrical phantom which contains the 3D printed spine. Carrageenan, a
substance applied to MRI phantoms [
], was used to fix the 3D printed spine phantom in the
cylindrical phantom. In the pre-study results, a 1% concentration of carrageenan yielded an
average HU of 8±10 on CT images and it was demonstrated that this concentration of
carrageenan could simulate the spinal cord and cancer in patient-specific QA phantoms. Also,
carrageenan at this concentration could become solidified after a certain period of time within 1
hours. Each spinal canal of the 3D printed spine was aligned at the centre of the cylindrical
phantom. The two different sets of the 3D printed spine phantom created by the DLP and
Polyjet printing technique were able to replace each other in the body phantom and these two
different 3D printed spine phantoms inserted in the body phantom were called the DLP
printed phantom and the Polyjet printed phantom in our study. Fig 2 shows the corresponding
developed phantom images.
4 / 12
Fig 2. (a) Acrylic body phantom (b) 3D printed spine phantom generated by the DLP technique (left) and the Polyjet technique (right) fixed to a cylindrical
phantom with carrageenan (c) Top view of the developed 3D printed spine quality assurance (QA) phantom, which consists of five slabs (d) Front view of the
developed 3D printed spine QA phantom.
To evaluate the reproducibility and applicability of the developed SBRT QA phantoms and
to compare the HU values of two different 3D printed spine phantoms, image fusion and
calculation of HU values between the two different CT image sets were performed. MIM software
(MIM Software, Cleveland, OH, USA) for image fusion was used. To measure the HU values,
ovals of the same size (200 mm2) were drawn at the same location of the body of the
3Dprinted spine phantom from T11 to L3 on each CT image, and then, the averages and standard
deviations were calculated using the open-source program ImageJ (http://rsb.info.nih.gov/ij/).
Treatment planning and evaluation
To investigate the acceptability of the developed SBRT QA phantom for purpose of radiation
dosimetry, radiation treatment planning, dose calculation was performed. The developed
phantom was scanned using a CT simulator (BrightSpeed, GE Healthcare, Fairfield, CT, USA)
with a slice thickness of 1.25 mm and at 120 kVp, 200 mAs, 120 mA, a 50 cm field of view, and
a 512ⅹ512 image matrix, which resulted in a voxel size of 0.977ⅹ0.977ⅹ1.5 mm using an axial
acquisition. In our study, all treatment plans were generated using the Eclipse treatment
planning system (Varian Associates, Palo Alto, CA, USA). All planning systems used a 2 mm dose
grid for dose calculations and 6 MV photon beams, and dose calculation was performed by
5 / 12
using the pencil beam convolution. The clinical target volume (CTV) was contoured, and a
planning target volume (PTV) was created by adding a 3 mm margin. The spinal cord
planning risk volume (PRV) was generated by adding a 2 mm margin to the delineated spinal cord
]. The PRV never overlapped with the PTV. The spinal cord and partial spinal cord
PRV was limited to 5 mm above and below the PTV . The PTV margin and spinal cord
PRV were reflected in set-up errors and motion errors on spine SBRT [
A single-fraction SBRT treatment plan was developed using seven static 6-MV beams in a
fixed gantry and sliding-window intensity-modulated radiation treatment (IMRT). All plans
included a prescribed dose of 16 Gy to the PTV with an aim that at least 90% of the PTV
received more than the prescribed dose and that the following dose constraints were satisfied:
a maximum PTV dose of 23 Gy, a maximum spinal cord dose of 10 Gy, and a maximum
partial cord PRV dose of 14 Gy [
To evaluate the SBRT QA plan, dose volume histograms were constructed. For the PTV,
Dmax, Dmean, D95%, and the relative volume of the PTV receiving at least 16 Gy (V16Gy) were
estimated. The conformity index (CI) evaluates the appropriateness of the PTV for the
prescription isodose volume in the treatment plans. The calculation for the CI is shown below
CI VPTV VTV
For the spinal cord, PRV, Dmax, and the relative volume of the spinal cord receiving at least
10 Gy (V10Gy) were evaluated.
Development and evaluation of the patient-specific spine QA phantom
Since the developed phantom is a body phantom with a form of replacing 3D printed spine
phantoms which were made by using the DLP and Polyjet 3D printing methods, the ability to
reproducibly manufacture phantoms is important. Reproducibility of the developed phantoms
was verified by the evaluating the degree of overlap for the two different 3D printed spine
phantom sets. The fused CT image sets of 3D printed L1 spine phantom which has the hole at
vertebra body for simulation of spine tumor showed in Fig 3. The centers and the edge of 3D
printed spin phantom were well matched each other without any distortion on the fused CT
image set. Especially, vertebral foramen and the holes at vertebra body of 3D printed spine
phantom on the fused CT image sets were also well matched as shown in Fig 3. Also, the CT
images of carrageenan surrounding the spine were homogeneous in all slices and showed HU
values of 8±10, as already observed in our pre-study. The mean HU values and standard
deviations of carrageenan surrounding the Polyjet-printed spine and the DLP-printed spine were
8.98±1.05 and 10.88±1.67, respectively.
Table 2 summarizes the means and standard deviations of the HU values at the vertebral
body from our results and it also presents these values according to the age of the patient and
the type of printing technique from the study by Lee [
]. The difference between the HU
values of the two 3D printed spine vertebrae (DLP and Polyjet) was 54.3 on average.
Dosimetric results of the patient-specific spine QA phantom according to the 3D printing technology
The radiation treatment planning results of the two phantom sets applied to the same SBRT
plan is shown in Table 3. All plan results of both phantoms satisfied the dose criteria described
6 / 12
Fig 3. The result of image fusion.
in the ongoing RTOG 0631 study [
]. For the spinal cord PRV, the RTOG 0631 criteria stated
that Dmax should be less than 14 Gy, and the Dmax values for both the Polyjet and DLP printed
phantom sets were indeed smaller than this value. However, the Dmax for treatment planning
using the DLP printed phantom image was about 1.488 Gy higher than that for treatment
planning using the Polyjet method. The conformity index values for dose distribution using the
DLP and Polyjet phantom image sets were 0.957 was 0.962, respectively.
7 / 12
Fig 4 shows the graph of dose to relative volume for the CTV, PTV, spinal cord PRV, and
spine without PTV. Even the same plan technique was applied for the DLP and Polyjet printed
phantom image sets, the PTV and CTV doses were greater for the DLP printed phantom plan
than for the Polyjet printed phantom plan. The DLP printed phantom delivered lower doses to
the organs at risk (OAR) than the Polyjet printed phantom, as shown in Fig 4.
Even though anatomical phantoms are becoming more advanced, they have several limitations
such as high cost and they are not fully customized for each patient [
]. However, with
initiation of several 3D printing agencies, customers will soon be able to order conveniently and
Fig 4. Dose volume histogram for the planning target volume (PTV) and organs at risk (OAR) from 3D
printed spine phantom sets (solid line for digital light projection (DLP) printed phantom, and dotted
line for Polyjet printed phantom) from spine stereotactic body radiation therapy (SBRT) treatment
8 / 12
inexpensively produced patient-specific phantoms on demand. Therefore, the application of
3D printing technology in the medical field including the radiation treatment field has
received focused attention. In this study, the applicability of patient-specific QA phantoms
which simulated the human spine, one of the anatomic structures, the representative organ
having a relatively high density than the other organs, was evaluated by generating 3D printed
spine phantoms using two 3D printing techniques, DLP and Polyjet.
The developed phantom consists of a main body phantom and a cylindrical phantom to be
inserted into the body phantom, and a 3D printed spine phantom was placed at the centre of
the cylindrical phantom. This type of developed phantom that depends on the insertion
method could cause a misalignment issue during the fabrication of the phantom. Thus,
evaluation of the reproducibility in the fabrication of a patient-specific 3D printed QA phantom is
important before using the developed phantom. As shown in Fig 3, there was neither image
distortion nor inhomogeneity, and this result could be interpreted as follows: 3D printed
spines were well fixed within carrageenan and the concentration of carrageenan was
appropriate to cause solidification. These results demonstrated that 3D printed spine phantoms within
body phantoms could be applied to different patients with good reproducibility.
The DLP and Polyjet technologies using an acrylic polymer are applicable to simulation of
high-density organs because both these technologies use high density materials and produce
stiff objects without formation of air bubbles inside the phantom. However, there are
significant differences between the HU values for the two printed phantom image sets and the
difference between the HU values of the 3D printed spine QA phantoms from the two printing
methods in the current investigation was 54.3, which could actually cause a difference in the
delivered radiation dose. This difference in the HU value depending on the 3D printing
technique in spite of the application of the same printing material has been discussed below. First,
additional materials were used for the Polyjet printing method. The Polyjet technology utilized
a material consisting of an acrylic polymer and additional materials, including TangoPlus
(Stratasys, MN, USA) and VeroWhite (Stratasys, MN, USA), to create a translucent object. On
the other hand, the material used for the DLP technology contained only an acrylic polymer
]. However, this effect may be negligible since only small quantities of these additional
materials were used for Polyjet printing. Second, different printing techniques may result in
products with different degrees of solidity or stiffness. Thus, different HU values could be
obtained. The solidity of the product is related to the resolution of the printing technique in
the z-direction. The DLP technique affords a higher resolution than the Polyjet technique, as
shown in Table 1. In other words, the HU values resulting from the DLP printed phantom are
also higher than the HU values resulting from the Polyjet printed phantom, which implies that
products obtained with use of the former technique are relatively stiff. However, dependence
of this different HU value on the 3D printing technique could be a helpful approach to produce
the most highly customized patient-specific spine QA phantom. As shown in Table 2, the HU
value of a vertebral body of a patient between 50 and 59 years of age was similar to the HU
value of a DLP printed phantom, and the HU values for a patient between 60 and 69 years of
age were well matched with the HU values of a Polyjet printed phantom. [
the related research had reported that the average HU value of spine depending on different
age groups decreased as the patients' age increased and the differences in HU values from L1
to 4th Lumbar were significant among different age groups. [
27, 35, 36
] Therefore, our finding
suggests that each 3D printing technique has its own special advantage in terms of the HU
value that depends on the patient's age and it also demonstrated that careful selection of the
3D printing technique and printing materials is required since even when the same material,
which is known to have a similar density to the human spine, was applied for 3D printing, the
calculated HU values from two different 3D printed spine phantom sets were different.
9 / 12
The dosimetric results from the two types of phantom image sets using the SBRT plan are
shown in Fig 4 and Table 3. The dosimetric results demonstrated that the both 3D printed
spine phantom could be applied in patient specific QA process since both dosimetric results
were within a dose criteria guided by RTOG 0631 for all calculated dose index (Dose to PTV,
PRV spinal cord, spinal cord) without unexpected dose gradient within phantom body and
peripheral region of spine phantom and shown acceptable conformity index even the HU
value of 3D printed spine phantom was slightly different. Furthermore, as shown in Fig 3, the
good reproducibility of 3D printed spine phantom could be able to produce appropriate spine
phantoms for therapeutic radiation purposes.
A limitation of our study was that the CT image set for the 3D printed spine phantom
modelling was not acquired from a real patient with a spinal tumor and the CT data set for 3D
printing modelling process was acquired from a normal patient. However, the main purpose
of our study was to investigate the 3D printing technology for development of a
patient-specific QA phantom for cases involving the spine, which is known as the organ having a relatively
high density, and to compare the 3D printing techniques. For this purpose, the CT image from
a normal patient, who has a standard shape of the spine and homogeneous composition, was
useful for the image modelling process and 3D printing.
In conclusion, this study confirmed that a 3D-printed phantom simulating a high-density
(about 1.4 g/cm3) organ can be created based on CT images and that a developed 3D printed
spine phantom could be utilized in patient-specific QA for SBRT. Additionally, a careful
decision regarding the appropriate printing technique according to the patient's condition
is required since there was a difference in the HU value of about 54.3 following application
of different printing technologies (DLP and Polyjet) even though the same material which
has the same density had been utilized. In further studies, our methods will be applied to
CT images of patients with actual spinal tumors and the appropriate printing technique and
materials suggested in our results will be used for patient-specific QA for spine SBRT
including dosimetric measurements to carry out an end-to-end test in order to increase the
accuracy of spine SBRT.
S1 File. https://figshare.com/s/3ce045497691c13cf1fc.
Conceptualization: MJK SRL.
Formal analysis: SRL.
Funding acquisition: TSS.
Investigation: SRL MYL HGY JYC SWJ.
Methodology: MJK SRL.
Resources: SRL MYL HGY JYC SWJ.
Supervision: JWS TSS.
10 / 12
Writing ± original draft: MJL SRL.
Writing ± review & editing: MJK.
11 / 12
1. Benedict SH , Yenice KM , Followill D , Galvin JM , Hinson W , Kavanagh B , et al. Stereotactic body radiation therapy: The report of AAPM Task Group 101. Med Phys 2010 ; 37 ( 8 ): 4078 ± 101 . https://doi.org/10. 1118/1.3438081 PMID: 20879569
2. Gallo JJ , Kaufman I , Powell R , Pandya S , Somnay A , Bossenberger T , et al. Single-fraction spine SBRT end-to-end testing on tomotherapy, vero, truebeam, and cyberknife treatment platforms using a novel anthropomorphic phantom . J Appl Clin Med Phys 2015 ; 16 ( 1 ): 170 ± 82 .
3. Gerszten PC , Ozhasoglu C , Burton SA , Vogel WJ , Atkins BA , Kalnicki S , et al. Cyberknife frameless stereotactic radiosurgery for spinal lesion: clinical experience in 125 cases . Neurosurgery 2014 ; 55 ( 1 ): 89 ± 99 .
4. Bartels RH , Yvette M , Winette TA . Spinal extradural metastasis: review of current treatment options . CA Cancer J Clin 2008 ; 58 ( 4 ): 245 ± 59 . https://doi.org/10.3322/CA. 2007 .0016 PMID: 18354080
5. Gagnon GJ , Nasr NM , Liao JJ , Molzahn I , Marsh D , McRae D , et al. Treatment of spinal tumors using cyberknife fractionated stereotactic radiosurgery: pain and quality-of-life assessment after treatment in 200 patients . Neurosurgery 2009 ; 64 ( 2 ): 297 ± 306 . https://doi.org/10.1227/01.NEU. 0000338072 .30246. BD PMID: 19057426
6. Avanzo M , Romanelli P . Spinal radiosurgery: technology and clinical outcomes . Neurosur rev 2009 ; 32 ( 1 ):1± 12 .
7. Ehler ED , Barney BM , Higgins PD , Dusenbery KE . Patient specific 3D printed phantom for IMRT quality assurance . Phys Med Biol 2014 ; 59 ( 19 ): 5763 ± 73 . https://doi.org/10.1088/ 0031 -9155/59/19/5763 PMID: 25207965
8. Heyns M , Breseman K , Lee C , Bloch BN , Jaffe C , Xiang H . Design of a Patient-Specific Radiotherapy Treatment Target . 2013 ; 171 ± 2 .
9. Leary M , Kron T , Keller C , Franich R , Lonski P , Subic A , et al. Additive manufacture of custom radiation dosimetry phantoms: An automated method compatible with commercial polymer 3D printers . Mater Des 2015 ; 86 : 487 ± 99 .
10. Hull CW . Apparatus for production of three-dimensional objects by stereolithography . Patent: US 4 , 575 , 330 , Mar. 11 , 1986 .
11. Bache ST , Juang T , Belley MD , Koontz BF , Adamovics J , Yoshizumi TT , et al. Investigating the accuracy of microstereotactic-body-radiotherapy utilizing anatomically accurate 3D printed rodentmorphic dosimeters . Med Phys 2015 ; 42 ( 2 ): 846 ± 55 . https://doi.org/10.1118/1.4905489 PMID: 25652497
12. Crump SS . Apparatus and method for creating three-dimensional objects . Patent: US 5 , 121 , 329 , Jun. 9, 1992 .
13. Jung J , Song SY , Yoon SM , Kwak JW , Yoon KJ , Choi WS , et al. Verification of Accuracy of CyberKnife Tumor-tracking Radiation Therapy Using Patient-specific Lung Phantoms . Int J Radiat Oncol Biol Phys 2015 ; 92 ( 4 ): 745 ± 53 . https://doi.org/10.1016/j.ijrobp. 2015 . 02 .055 PMID: 25936598
14. Madamesila J , McGeachy P , Villarreal Barajas JE , Khan R . Characterizing 3D printing in the fabrication of variable density phantoms for quality assurance of radiotherapy . Phys Med 2015 in press
15. Yoon K , Kwak J , Cho B , Song SY , Lee SW , Ahn SD , et al. Development of New 4D Phantom Model in Respiratory Gated Volumetric Modulated Arc Therapy for Lung SBRT . K J Med Phys (Progress in Medical Physics) 2014 ; 25 ( 2 ): 100 .
16. Burfeindt M , Mays R , Van Veen B. MRI-derived 3-D-printed breast phantom for microwave breast imaging validation . IEEE AWPL . 2012 ; 11 : 1610 ± 1613 . https://doi.org/10.1109/LAWP. 2012 .2236293 PMID: 25132808
17. Deckard CR . Method and apparatus for producing parts by selective sintering . Patent: US 4 , 863 , 538 , Sep. 5, 1989 .
18. Gear JI , Long C , Rushforth D , Chittenden SJ , Cummings C , Flux GD . Development of patient-specific molecular imaging phantoms using a 3D printer . Med Phys 2014 ; 41 ( 8 ): 082502 . https://doi.org/10. 1118/1.4887854 PMID: 25086556
19. Ionita CN , Mokin M , Varble N , Bendnarek DR , Xiang J , Snyder KV , et al. Challenges and limitations of patient-specific vascular phantom fabrication using 3D Polyjet printing . Proc SPIE Int Soc Opt Eng 2014 ; 9038 : 90380M .
20. Solomen J , Mileto A , Ramirez-Giraldo JC , Samei E. Diagnostic performance of an advanced modeled iterative reconstruction algorithm for low-contrast detectability with a third-generation dual-source multidetector CT scanner . Radiology 2015 ; 275 ( 3 ): 735 ±45 https://doi.org/10.1148/radiol.15142005 PMID: 25751228
21. Solomon J , Samei E. Quantum noise properties of CT images with anatomical textured backgrounds across reconstruction algorithms: FBP and SAFIRE . Med Phys 2014 ; 41 ( 9 ): 091908 . https://doi.org/10. 1118/1.4893497 PMID: 25186395
22. Kudo3D, Titan1, http://www.kudo3d. com/recommended-printing-parameters-exposure-time-liftingheight-lifting-speed/
23. Yamane M , Kawaguchi T , Kagayama S , Higashiyama, Suzuki, Imaeda M , et al. Apparatus and method for forming three-dimensional article . Patent: US 5 , 059 , 266 . Oct. 22 , 1991 .
24. Objet , Connex, Connex 350 2011. Available form: http://www.cmu.edu/ices/advanced-manufacturinglaboratory/objet-connex-350-description.pdf
25. National Industrial Chemicals Notification and Assessment Scheme, Modified acrylic polymer . 1996 Available form: https://www.nicnas.gov.au/__data/assets/pdf_file/0017/9071/NA307FR.PDF
26. Lee S , Chung CK , Oh SH , Park SB . Correlation between Bone Mineral Density Measured by DualEnergy X-Ray Absorptiometry and Hounsfield Units Measured by Diagnostic CT in Lumbar Spine . J Korean Neurosurg Soc 2013 ; 54 ( 5 ): 384 ±9. https://doi.org/10.3340/jkns. 2013 . 54 .5.384 PMID: 24379944
27. Du JY , Aichmair A , Ueda H , Girardi FP , Cammisa FP , Lebl DR . Vertebral body Hounsfield units as a predictor of incidental durotomy in primary lumbar spinal surgery . Spine 2014 ; 39 ( 9):E593±8 . https:// doi.org/10.1097/BRS.0000000000000255 PMID: 24503684
28. Xu XG . An exponential growth of computational phantom research in radiation protection, imaging, and radiotherapy: a review of the fifty-year history . Phys Med Biol 2014 ; 59 ( 18 ):R233± 302 . https://doi.org/ 10.1088/ 0031 -9155/59/18/R233 PMID: 25144730
29. Kato H , Kuroda M , Yoshimura K , Yoshida A , Hanamoto K , Kawasaki S , et al. Composition of MRI phantom equivalent to human tissues . Med Phys 2005 ; 32 ( 10 ): 3199 . https://doi.org/10.1118/1.2047807 PMID: 16279073
30. Hattori K , Ikemoto Y , Takao W , Ohno S , Harimoto T , Kanazawa S , et al. Development of MRI phantom equivalent to human tissues for 3 .0- T MRI . Med Phys 2013 ; 40 ( 3 ): 032303 . https://doi.org/10.1118/1. 4790023 PMID: 23464335
31. Kuijper IT , Dahele M , Senan S , Verbakel WF . Volumetric modulated arc therapy versus conventional intensity modulated radiation therapy for stereotactic spine radiotherapy: a planning study and early clinical data . Radiother Oncol 2010 ; 94 ( 2 ): 224 ±8. https://doi.org/10.1016/j.radonc. 2009 . 12 .027 PMID: 20122745
32. Ryu S , Jin JY , Jin R , Rock J , Ajlouni M , Movsas B , et al. Partial volume tolerance of the spinal cord and complications of single-dose radiosurgery . Cancer 2007 ; 109 ( 3 ): 628 ± 36 . https://doi.org/10.1002/cncr. 22442 PMID: 17167762
33. Cai J , Sheng K , Sheehan JP , Benedict SH , Larner JM , Read PW . Evaluation of thoracic spinal cord motion using dynamic MRI . Radiother Oncol 2007 ; 84 ( 3 ): 279 ± 82 . https://doi.org/10.1016/j.radonc. 2007 . 06 .008 PMID: 17692979
34. International Commission on Radiation Units and Measurements , ICRU-62. Prescribing, recording and reporting photon beam therapy (supplement to ICRU 50) . ICRU, Bethesda, MD; 1999 .
35. Schreiber J , Anderson P , Hsu W. Use of computed tomography for assessing bone mineral density . Neurosurg Focus 2014 ; 37 ( 1 ):E4 https://doi.org/10.3171/ 2014 .5. FOCUS1483 PMID : 24981903
36. Batawil N , Sabiq S. Hounsfield unit for the diagnosis of bone mineral density disease: A proof of concept study . Radiography 2016 ; 22 ( 2 ):e93± e98