ASIR reconstruction boosts diagnostic power in HRCT of diffuse lung disease

High-resolution CT (HRCT) with thin slices and high-spatial-frequency algorithms is the gold standard for evaluating diffuse lung disease. Most high-resolution CT scans are reconstructed using filtered back projection (FBP), but a new algorithm -- adaptive statistical iterative reconstruction (ASIR) (GE Healthcare, Chalfont St. Giles, U.K.) -- may offer better image quality.

Iterative reconstruction techniques have been found to reduce noise and improve spatial resolution, but they haven't been systematically compared with FBP.

"We decided to do a research study comparing the visualization of subtle normal and abnormal findings on chest CT images of patients with diffuse lung disease reconstructed with conventional FBP and recently available ASIR techniques, the latter with and without high definition," presenter Priyanka Prakash told AuntMinnie.com.

A side-by-side direct comparison of 0.625-mm sections showed moderate-to-severe blurring of interlobular septa and small bronchi and bronchioles with FBP. ASIR images alone were deemed better than FBP, but were affected by some blurring of the interlobular septa and small bronchioles.

Not so for the high-definition version, ASIR-HD reconstruction, which found no blurring and no artifacts. Visualization of larger structures was generally not improved, however.

"ASIR-HD reconstruction results in superior resolution of subtle and tiny anatomical structures and lesions in diffuse lung disease," Prakash said.