Phantom FDA

Detailed Description

Database Description (link)

The anthropomorphic thoracic phantom (Kyotokagaku Incorporated, Tokyo, Japan) employed in this study is shown in Figure 1, along with the vasculature insert on which synthetic nodules were attached before CT imaging.  The phantom does not contain lung parenchyma so the space within the vascular structure is filled with air.

Fig 1:  Photograph of the exterior shell of the thoracic phantom (left) and the vasculature insert (right).

The synthetic lung nodules used in our data CT scans were manufactured by either Kyotokagaku Incorporated (Japan) and Computerized Imaging Reference Systems (CIRS, Norfolk, VA).  They consisted of objects varying in size (5, 8, 10, 12, 20, 40 mm), shape (spherical, elliptical, lobulated, spiculated), and density (-800, -630, -10, +100 HU).  Figure 2 shows examples of the various sizes and shapes of synthetic nodules used in our CT data collection.

Eight different layouts of nodules were specified by placing them in premarked positions within the phantom vasculature, where they were either attached to vessels or suspended in foam (non-attached configuration).  Care was taken to maintain constant positioning of the nodules when a particular layout was scanned multiple times or with different protocols.  For that purpose, vessels on which nodules were attached were color coded.  Table 1 tabulates the nodule configuration for the nodule layouts that are currently available at NBIA DICOM Radiology Portal in terms of nodule positioning, size, shape, and density. Figures 3-6 show a schematic diagram of the currently available layouts.  All tables and figures in this document will be updated as more data is posted.

Figure 2:  Photographs of the different types of synthetic nodules used in this study.  Each column shows example nodules in three sizes, with lobulated, elliptical, spiculated, spherical, and irregular nodules shown from left to right.  The three sizes shown here were manufactured to have the equivalent volumes of spherical nodules with diameters of 5, 10, and 20 mm (with the exception of the irregular shapes which have nominal diameters of about 5, 10, 12 mm).  Additional nodules used in this study span the size range between 5-60mm.

 Table 1.  Summary of currently available nodule layouts.  *SPH- spherical, ELL- elliptical, LOB- lobulated, SPI- spiculated.  ^Note: The 10mm, 100HU elliptical nodule in the right lung has a large hole in it.  A replacement was scanned as part of Nodule 6 which will be released by the end of 2014.

The phantom was scanned using a Philips 16-row scanner (Mx8000 IDT, Philips Healthcare, Andover, MA) and a Siemens 64-row scanner (Somatom Definition 64, Siemens Medical Solutions USA, Inc., Malvern, PA).  Scans were acquired with varying combinations of effective dose, pitch, and slice collimation, and were reconstructed with varying combinations of slice thicknesses and reconstruction kernels.  Ten exposures were acquired for each imaging protocol.  The phantom position was not changed during the 10 repeat exposures; however it was repositioned between different imaging protocols or different nodule layouts.  Table 2 summarizes the imaging protocols for the nodule layout.

NOTE:  Each study in the database contains 10 repeat scans for that particular acquisition protocol, multiplied by the number of reconstructions.   The study and series descriptions contain the following information:

Study description: 

Contains information on: the scanner vendor (currently Philips or Siemens), the exposure (in mAs), the pitch (currently either 1.2 or 0.9 according to the definition  , where Δd is the patient table travel in the horizontal direction and T is the detector width at the isocenter plane),  and slice collimation (in mm).

Series description:

  Contains information on reconstructed slice thickness (in mm), reconstructed slice increment (in mm), and reconstruction filter or kernel (currently either C for detail, or B for medium).

For example: there are 16 studies for Nodule Layout #1 (4 exposures x 2 slice collimations x 2 pitch settings).  Each study contains 30 series (10 repeat scans x 3 reconstructed slice thickness x 1 reconstruction kernel).

A key component of the CT lung phantom project is the ability to compare the estimated nodule size with the known true size or reference gold standard.  As part of our project, volume was used as a surrogate measure of size.   The true volume estimate of each synthetic nodule was derived from weight and density measures.  Both the CIRS-and Kyotokagaku nodules were accompanied by density measures.  Nodule weights were measured in our lab using a precision scale of 0.1 mg tolerance (Adventurer Pro AV 2646, Ohaus Corp, Pine Brook, NJ).  Three repeat weight measurements were made and these weights were averaged to produce a final estimated weight for each nodule.  Our estimates of the true volume of the synthetic nodules in each layout are summarized in Table 3 along with approximated xyz location (based on 0.8mm slice thickness) of nodule center in the CT scans. 

This phantom and the associated synthetic nodules designed in our lab have been used in a number of studies examining the accuracy and precision of volumetric measurements using CT

Table 3.  Approximate center location and estimated true volume of synthetic lung nodules in each nodule layout based on 0.75mm slice thickness, 0.4mm slice increment CT scans.  (SPH=spherical, ELL=elliptical, LOB= lobulated, SPI= spiculated, irr=irregular)

Please contact help@cancerimagingarchive.net who can direct you to Dr. Gavrielides with further scientific questions.

Appendix 1

 Figure 3.  Schematic diagram of Nodule Layout#1 in terms of nodule placement. Vessel branches within the anthropomorphic phantom were color coded for the purpose of mapping nodules to specific positions within the phantom’s vasculature structure in a reproducible manner.

Figure 4.  Schematic diagram of Nodule Layout#2 in terms of nodule placement.

Figure 5.  Schematic diagram of Nodule Layout#3 in terms of nodule placement.

Figure 6.  Schematic diagram of Nodule Layout#4 in terms of nodule placement.

Citations & Data Usage Policy 

Users must abide by the TCIA Data Usage Policy and Restrictions. Attribution should include references to the following citations:

Questions may be directed to help@cancerimagingarchive.net.

Other Publications Using This Data

TCIA maintains a list of publications which leverage our data.  If you have a manuscript you'd like to add please contact TCIA's Helpdesk.

1. Kalpathy-Cramer, J., et al, (2016) "Radiomics of Lung Nodules: A Multi-Institutional Study of Robustness and Agreement of Quantitative Imaging Features." Tomography 2(4)430-437. doi:  10.18383/j.tom.2016.00235
2. Peskin AP, Dima AA, Saiprasad G. An Automated Method for Locating Phantom modules in Anthropomorphic Thoracic Phantom CT Studies. The 2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition. 2012. (IPC conference link)

Version 2 (Current): Updated 2017/08/30

Version 1: Updated 2014/08/04