Population of anatomically variable 4D XCAT adult phantoms for imaging research and optimization

Abstract

Purpose: The authors previously developed the 4D extended cardiac-torso (XCAT) phantom for multimodality imaging research. The XCAT consisted of highly detailed whole-body models for the standard male and female adult, including the cardiac and respiratory motions. In this work, the authors extend the XCAT beyond these reference anatomies by developing a series of anatomically variable 4D XCAT adult phantoms for imaging research, the first library of 4D computational phantoms.

Methods: The initial anatomy of each phantom was based on chest-abdomen-pelvis computed tomography data from normal patients obtained from the Duke University database. The major organs and structures for each phantom were segmented from the corresponding data and defined using nonuniform rational B-spline surfaces. To complete the body, the authors manually added on the head, arms, and legs using the original XCAT adult male and female anatomies. The structures were scaled to best match the age and anatomy of the patient. A multichannel large deformation diffeomorphic metric mapping algorithm was then used to calculate the transform from the template XCAT phantom (male or female) to the target patient model. The transform was applied to the template XCAT to fill in any unsegmented structures within the target phantom and to implement the 4D cardiac and respiratory models in the new anatomy. Each new phantom was refined by checking for anatomical accuracy via inspection of the models.

Results: Using these methods, the authors created a series of computerized phantoms with thousands of anatomical structures and modeling cardiac and respiratory motions. The database consists of 58 (35 male and 23 female) anatomically variable phantoms in total. Like the original XCAT, these phantoms can be combined with existing simulation packages to simulate realistic imaging data. Each new phantom contains parameterized models for the anatomy and the cardiac and respiratory motions and can, therefore, serve as a jumping point from which to create an unlimited number of 3D and 4D variations for imaging research.

Conclusions: A population of phantoms that includes a range of anatomical variations representative of the public at large is needed to more closely mimic a clinical study or trial. The series of anatomically variable phantoms developed in this work provide a valuable resource for investigating 3D and 4D imaging devices and the effects of anatomy and motion in imaging. Combined with Monte Carlo simulation programs, the phantoms also provide a valuable tool to investigate patient-specific dose and image quality, and optimization for adults undergoing imaging procedures.

Figure 1

(Top) 50th percentile male (left) and female (right) whole-body anatomies of the XCAT phantom. (Bottom) 4D cardiac and respiratory models of the XCAT.

Figure 2

BMI and ages of the patients used to develop the XCAT adult series.

Figure 3

Screenshot of the PeopleSize program. A user can select different regions of the body (middle menu) and display measurements for them at different percentiles (output on the right). The head is shown above as an example; the measurements for head circumference (50th and 90th percentile) are listed. A user can select the particular percentiles they want listed at the right in the program settings. They may also select the type of individuals they want. In this case, we selected US males, ages 18–64.

Figure 4

Procedure for calculating the MC-LDDMM transform. To calculate the transform, template and target images are required. The above steps are performed to create these images. These images are further broken down into a binary image for each organ that is used to calculate the transform.

Figure 5

The MC-LDDMM transform is used to transform the template XCAT phantom to define the detailed anatomy of the target phantom. The new XCAT model, based on the patient data, contains the same structures as the original XCAT.

Figure 6

Collage showing the 58 new adult anatomies. The head, arms, and legs are not shown to focus on the chest–abdomen–pelvis anatomy, which is based on the patient imaging data. The muscles and blood vessels are also not shown for presentation purposes.

Figure 7

New male phantom (BMI = 36) created by transforming the original male XCAT. The original XCAT is shown to the left while the new phantom is shown in the middle. The interior anatomy of the male phantom is shown with varying levels of detail. The new phantom contains the same number of structures as the original XCAT. The phantom also includes the models for the cardiac and respiratory motions. The new phantom is shown to the right at different phases of the cardiac and respiratory cycles. The dotted line in the respiratory pictures shows how the internal organs are moving.

Figure 8

Four adult male (top) and four adult female (bottom) XCAT phantoms with CT simulation results shown at the bottom.