Electronic bone suppression programs used in addition to CAD allow radiologists to detect more lung nodules than with chest X-rays alone.
Examination with dual-energy subtraction used with computer-aided detection (CAD) allows radiologists to detect more pulmonary nodules than do examinations with traditional chest X-rays alone, according to a study published in the American Journal of Roentgenology.
Researchers from the University of Bern in Switzerland undertook a retrospective study comparing the detection of lung nodules with dual-energy subtraction and bone suppression software alone and in combination with CAD, with the use of conventional X-rays alone. The researchers retrospectively selected and randomized 143 patients, 101 of whom had one to five nodules, which measured from 5 mm to 29 mm, and 42 who had no nodules.
Each case was read by three independent radiologists. Suspicious lesions were marked on the original chest X-rays, dual-energy subtraction images, and bone-suppressed images, before and after postprocessing with CAD. The observers’ marks and the CAD marks were compared with CT as the reference standard.
The results showed that traditional chest X-rays had the lowest reader sensitivity, while the bone suppression software/CAD combination had the highest sensitivity.
Chest X-ray alone detected only 72.67 true positive nodules with a sensitivity of 46.9 percent, while bone suppression with CAD detected 80 true-positive nodes with 51.6 percent sensitivity. Chest X-ray with CAD but before bone suppression detected 79.33 nodules with 51.2 percent sensitivity. Dual-energy subtraction and bone suppression also provided the same false-positive and true-positive rates. Up to 22.9 percent of the lesions were missed by the readers, detected only by CAD.
“Dual-energy subtraction and the electronic bone suppression program used in this study provided similar detection rates for pulmonary nodules,” the authors concluded. “Additionally, CAD alone or combined with bone suppression can significantly improve the sensitivity of human observers for pulmonary nodule detection.”