Spectral CT may help identify cases of COVID-19 in its early stages
Spectral CT may identify COVID-19 in earlier stages
October 29, 2020
by John R. Fischer
, Senior Reporter
Spectral CT may enable clinicians to identify cases of COVID-19 at earlier stages, according to researchers in France.
Dr. Beatrice Daoud at Antony's Private Hospital in France and her colleagues say that using dual-layer detector–based spectral CT to perform electron density imaging (ED imaging) improves visualization of ground-glass opacities (GGOs), the main lung lesions found on chest CTs for cases of COVID-19 pneumonia, compared to conventional CT which lacks the sensitivity needed for detection.
"Our standard chest CT protocol allowed us to evaluate the different spectral results for the assessment of COVID-19 lung lesions," Dr. Béatrice DAOUD, M.D., senior radiologist in the radiology department at Antony's Private Hospital, told HCB News. "Electron density map appeared to be the best to visualize ground-glass opacities (GGOs) and to enhance subtle lung lesions that initially were barely visible on conventional CT. Subsequently, these could be better identified on conventional CT. Through this process we were able to identify patients that were requiring close monitoring."
Spectral data improves contrast enhancements, reduces artifacts, and relays information about tissue composition. Dual-layer CT extracts value from low- and high-energy X-rays in a single acquisition, making spectral information available for all scans. ED image displays the electron density of each voxel relative to the electron density of water (in %), while the dual layer detector technology offers noise suppression that provides better quality to the ED images and enables better visualization of GGOs.
The researchers conducted their retrospective study on four patients with RT-PCR-confirmed COVID-19 who underwent initial and follow-up chest CTs. They used Philips’ IQon Spectral CT to perform dual-layer spectral CT scans and evaluated if they could enhance subtle lung lesions that initially are barely visible on conventional CT, using ED imaging. Two thoracic radiologists reviewed the scans with Philips’ IntelliSpace Portal workstation and its built-in Spectral CT viewer software. Iterative reconstruction methods were used to reconstruct the spectral scans, which were compared to the initial conventional CT images and then with the follow-up conventional CT images.
Pulmonary lesions in all four patients stood out more on the ED images than on the initial conventional CT images and were confirmed on the follow-up conventional CT images. Lesion extent on the ED images appeared greater than on the initial conventional images and was more similar to that on the follow-up scans. For example, three lesions that were less than one centimeter were barely visible on initial scans but were seen on ED images and identified on conventional follow-up CT. ED images also showed more promising findings by enhancing GGO contrast compared to the normal lung.
The study was conducted at a time when France faced a shortage of RT-PCR tests. Daoud says that not only did spectral CT scanning help with triage but supported patients that required further lung lesion evaluation. She adds that ED mapping could lead to early detection and better triage and hopes the study will create wider recognition and availability for dual layer spectral CT technology.
"The time required to validate the benefits of a more advanced technology before it can become the new standard is likely one of the most important challenges to its wider availability," said Daoud. "Our ongoing study will hopefully further demonstrate the benefits of spectral CT in COVID-19 diagnosis, and hopefully contribute to it's wider use and adoption."
Daoud and her colleagues are currently running a multicenter scientific study to evaluate if ED map can better quantify the extent of COVID-19 lung lesions than conventional CT, and consequently improve patient management.
The findings were published in the American Journal of Roentgenology.