Digital radiography challenges CT-based lung cancer screening

Digital radiography could turn chest x-rays into a more cost-effective tool than CT for lung cancer screening, according to research presented Sunday.

Radiologists consider low-dose CT scanning a formidable ally for early cancer detection. But CT has its own challenges related to accuracy and false positive rates for small nodule detection. And it may not be as widely available as needed for the screening of large patient populations. A number of studies suggest that digital radiography, further enhanced by computer-aided detection and dose reduction refinements, could overcome CT's shortcomings.

A multicenter group of Chinese researchers evaluated the diagnostic performance of a real-time pulmonary nodule analysis software developed to aid DR softcopy reading. They found this technology effectively increased lung nodule detection rates and reduced inter-observer variations.

Li Fan, Ph.D., and colleagues at the Peking Union Medical College, the Beijing Friendship Hospital, the Zhejiang Hospital, and the Sino-American corporation EDDA Technology, gathered chest DR screening studies from a total of 200 patients. Researchers generated chest images from different DR equipment, though all with their corresponding CT match.

Two radiologists from each research institution read the studies. They marked small "actionable" nodules-those previously agreed upon as worthy of CAD analysis -in two phases, one using the nodule analysis software (EDDA's IQQA-Chest) and one without CAD.

According to preliminary results, the individual radiologist's detection rate without CAD averaged 50%. But the collective detection rate for small actionable nodules confirmed as true positives averaged 35%, underscoring an important inter-observer variation. With CAD, the individual and collective detection rates recorded were about 90%. In addition, the inter-observer variation decreased significantly.

"Digital radiography could be a more economic alternative to CT," Fan said.

Overlying bone structures, such as the ribs, present a challenge to DR in lung nodule detection. Two studies assessed the contribution of dual-energy subtraction in this area.

In one, Korean researchers led by Dr. S. K. Ahn used dual-energy subtraction to evaluate the accuracy of volume estimation of solitary pulmonary nodules. They found the technique eliminated rib shadows and allowed accurate lung nodule volume estimation through computerized volume measurement. In another study, Dr. Amy E. Musk and colleagues from the University of Maryland enrolled 139 patients who had been referred for a chest CT. They found the addition of dual energy subtraction helped reduce the recommendation rate for follow-up CT.

A study by Duke University researchers proposed the use of automated detection of lung nodules on DR using digital tomosynthesis. Dr. Nariman Majdi-Nasab and colleagues developed a CAD algorithm to identify nodules in five patients enrolled prospectively in an ongoing trial. The system performs image filtration, segmentation, and feature extraction while patients undergo tomographic imaging in 69 planes at 5 mm plane intervals.

The investigators found their combined CAD/tomographic scheme provided sensitivity and specificity readings that matched those of other CAD methods used to project chest radiographs. In addition, they currently investigate a technique to reduce false positives through the application of this procedure.

Although research indicates that digital radiography might be the cheapest solution compared to CT, the reality is that only a handful of institutions can offer the techniques proposed, said Dr. Cornelia Schaefer-Prokop, an associate professor of radiology at the Academic Medical Center in Amsterdam, the Netherlands.

A lung cancer screening program must consider four important aspects, including the availability of the diagnostic imaging modality used, radiation dose, diagnostic performance, and the overall costs involved in screening. Considering that low-dose CT is already a powerful and available tool, DR studies must provide convincing evidence of its superiority in these four areas, Schaefer-Prokop said.