Innovative X-ray-based imaging technologies for rapid and accurate diagnosis of thoracic disease in critically ill patients
My most recent research at Toronto General Hospital explored the value of dual-energy (DE) imaging and digital tomosynthesis (DT) as solutions to the limitations of conventional radiographic thoracic imaging (chest radiography). Both DE and DT have been around for a few decades, but recent advancements in digital detectors have made this technology increasingly promising in clinical use.
To this end, I conducted a study in collaboration with Ali Ursani, BEng, Fatima Ursani and Narinder Paul, MD, from Toronto General Hospital. Dr. Paul, who is the Site Chief, Toronto General Hospital – Joint Department of Medical Imaging, was the principal investigator in the study. In addition, four imaging experts from Carestream Health – Samuel Richard, PhD, Xiaohui Wang, PhD, Nathan Packard, PhD, and Levon Vogelsang PhD – were key participants. Additional support for patient recruitment consisted of the research coordinator and the team of technologists trained on the system with DE and DT functionalities.
Among the study’s many objectives were these:
- Understand the gap between the performance of state-of-the-art chest radiograph (CXR) and computed tomography and the current needs of imaging facilities and patients.
- Explore the concepts of dual-energy imaging and digital tomosynthesis as viable solutions to address these needs.
- Examine the potential benefits that might be offered by a combined DE and DT system.
![Carestream DE and DT imaging]()
Background: the traditional CXR is by far the most common approach to chest imaging. And, for over a century, its value in the diagnostic process has been indisputable. Yet, this modality has its drawbacks as well. Most fundamentally, chest radiographs are limited by the fact that they are flat, two-dimensional projections of complex, three-dimensional anatomical structures. Depending on what the radiologist is looking for, the anatomical structure can either be of interest or create a confounding background in the image. For instance, if the structure of interest is a lung nodule, the overlapping anatomy composed of ribs, lungs and vessels can impede the detection of pathologies such as lung nodules.
This results in obscuring anatomical noise. There are two broad techniques for reduction of background anatomical noise. The first technique is tissue discrimination which refers to enhancement of tissues via techniques like contrast injection or tissue subtraction as is achieved using DE imaging. Secondly there is spatial discrimination which refers to the separation of overlapping structures. Techniques to achieve spatial discrimination are Computed Tomography (CT) and DT.
The study: first, dual-energy imaging was meticulously evaluated for its performance compared to traditional chest X-rays. With DE, two radiographs, at two different kilo-voltage levels, are taken of the same target area. The two images are then combined using different weighting factors depending on whether bone contrast or soft tissue has to be isolated. Factors that affect the outcome of DE imaging include energy selection (the choice of high and low energy values), dose allocation (the ratio of dose between high and low energies), and filtration (which increases the energy separation between the high- and low-energy images).
In performing phantom, patient and cadaver imaging, we found that the dual energy technique consistently reduced anatomical clutter and provided improved conspicuity of abnormalities such as lung nodules. Also it was found that differential filtration (whereby filtration is changed between the high and low energies) is a more dose efficient technique as compared to fixed filtration (same filtration is maintained between high and low energies).
Next, we looked at the relative advantages of digital tomosynthesis. DT uses spatial discrimination: multiple X-rays of a single target are taken from a limited number of different angles. These cross-sectional images are then used to reconstruct 3D images. Factors determining the efficacy of DT include the angular range of image acquisition, the use of detector binning, and the amount of scan time used to acquire the projections.
While it does have some limitations when compared to Computed Tomography, DT nevertheless reduced overlying clutter and improved feature conspicuity – which was demonstrated in phantom, patient and cadaver imaging.
We conclude that combining the tissue discrimination approach of dual-energy imaging with the spatial discrimination capabilities of digital tomosynthesis in a single, integrated system would be highly beneficial. Furthermore, interesting quantitative imaging applications can be explored using the above-mentioned technologies. #HealthIT, #Radiology, #digitalradiography, #imaginginformatics #X-ray, #Xray
Shailaja Sajja, MS, is a scientific researcher in the Joint Department of Medical Imaging, Toronto General Hospital. The details and outcomes of this study were presented at RSNA 2015 and received a Certificate of Merit.
The post Study: Dual-Energy Imaging and Digital Tomosynthesis appeared first on Everything Rad.