Towards phase contrast lung CT in patients – a human chest phantom study

Lung diseases continue to pose a major burden of public health, with their incidence on the rise. To diagnose and stage lung diseases, high resolution computed tomography (HRCT) is virtually the sole applicable imaging modality in the clinics, reaching a spatial resolution of about 0.5 x 0.5 x 0.9 mm³. This limited resolution is not inherent to the technique itself, but rather stems from the necessity to minimize the x-ray exposure in patients. The smallest constituent in the human lung – the alveoli – have a size of roughly 0.1 mm, making them unresolvable in clinical CT scans. Consequently, tiny pathological changes like the onset of lung fibrosis or the structure of lung metastasis cannot be characterized in detail, sometimes leading to a non-conclusive diagnosis. In such cases, biopsies are typically taken from the lung, implying that a tiny piece of tissue is removed, which is an invasive method with a risk to cause internal bleeding or a pneumo-thorax.

Phase contrast CT (PCCT) is a rather novel x-ray imaging technique that can be performed at synchrotron sources, producing 3D images of the lung with high contrast, with much better resolution than clinical CT even at lower x-ray doses than those used in clinical CT scans. Despite these advantages, optimizing PCCT for a potential use in patient lung CT is a complex task. The team at the SYRMEP beamline of Elettra, along with colleagues from Heidelberg and Göttingen in Germany, as well as collaborators from the Cattinara University Hospital in Trieste, have devised a novel setup to scan heavy and bulky specimens, including an anthropomorphic human chest phantom (ANTHONY) prototype. The phantom can be equipped with fresh pig lungs obtained from a regular butcher. Since pig lungs closely resemble human lungs, this approach allows for obtaining images as close as possible to the conditions in real patients. Moreover, the phantom includes artificial ribs, absorbing shells and the capability to simulate breathing motion to make it ever more comparable to human lung CT imaging. Another important aspect is that the phantom can be transported and used in the same way in a clinical CT, enabling direct comparison of the resulting images.

With this new setup we were able to demonstrate that human lung CT will be possible at a resolution of .067 x 0.067 x 0.067 mm³ at the same dose level as used in clinical CT (see Figure 1). We believe that further optimization of the setup, such as using a novel detector together with the implementation of novel algorithms, will allow us to push the current limit even further. Based on these results, it is planned to implement a patient lung CT setup during the upcoming upgrade of Elettra and perform a first clinical trial with this novel imaging method when the synchrotron light will be back. Meanwhile, interested researchers will be able to use the phantom in the framework of EuroBioImaging initiative or Elettra Call for proposals.

Figure 1: a) The lung imaging setup at the Synchrotron Radiation for Medical Physics (SYRMEP) beamline of the Italian synchrotron consists of a positioning stage and rotary unit for heavy and bulky specimen (st), mounted on top is the actual chest phantom, equipped with a fresh porcine lung (ph) and a battery based vacuum pump (pu) to ensure steady inflation during the experiment. b) The current setup allows the depiction of the entire cross-section of the lung at ∼0.067 mm resolution, which enables three-dimensional structure analysis in great detail (displayed cube). c) a pig lung with an artificially generated pathological pattern of a solid tumor (*) and some surrounding sub-solid components (dashed area) scanned within the phantom in a clinical CT d) the same structures scanned at the SYRMEP beamline with the same x-ray dose. The dramatically higher resolution (0.067×0.067×0.067 mm³ versus 0.45×0.45×0.90 mm³) allows for superior characterization of the nodule's surface and substructures and reveals opacified terminal bronchioles in the dashed area – both important information for a clinical diagnosis.

This research was conducted by the following research team:

Christian Dullin^1,2,3,4, Willi Linus Wagner³, Marco Confalonieri⁵, and Giuliana Tromba^4
1 University Medical Center Goettingen, Department of Diagnostic and Interventional Radiology, Goettingen, Germany.
² Max Planck Institute for Multidisciplinary Sciences, Goettingen, Germany.
³ University Hospital Heidelberg, Department of Diagnostic and Interventional Radiology, Heidelberg, Germany.
⁴ Elettra - Sincrotrone Trieste S.C.p.A., Trieste, Italy.
⁵ Pulmonology Unit, University Hospital of Cattinara, Trieste, Italy.

Contact persons email: 

Local contact person email: