Algorithm can automatically detect abnormal lung volumes

Sunday, November 28 | 11:45 a.m.-11:55 a.m. | SSA19-07 | Room S403B

In this paper presentation, researchers will provide details on a computerized method for identifying abnormal lung volumes from interstitial lung disease (ILD).

Interstitial lung disease is a common pulmonary abnormality; however, correct diagnosis is difficult and the condition can be fatal. Lesion contrast is often low and the disease patterns are very complex in chest CT, said co-author Jiahui Wang, PhD, of the University of North Carolina at Chapel Hill.

As a result, the research group is developing a computerized system to aid radiologists in diagnosing and quantifying ILD. Making use of previous work that developed a texture analysis technique to segment lungs with severe ILD, the computer-aided detection (CAD) scheme automatically detects ILD inside the segmented lungs on thoracic CT, he said.

"The proposed methods of this study achieved high performance levels and can be used for computerized quantification of clinical indices, which would help radiologists in diagnosis of interstitial lung disease," Wang said.