Jung Jae Park MD, Presenter: Nothing to Disclose

Myung Jin Chung MD, Abstract Co-Author: Advisory Board, Samsung Advanced Institute of Technology Research Consultant, Samsung Mobile Display Company

Ji-Eun Kim MD, Abstract Co-Author: Nothing to Disclose

Chin A Yi MD, PhD, Abstract Co-Author: Nothing to Disclose

Man Pyo Chung, Abstract Co-Author: Nothing to Disclose

Kyung Soo Lee MD, PhD, Abstract Co-Author: Nothing to Disclose

Jiyoung Hwang MD, Abstract Co-Author: Nothing to Disclose

Kyoung Doo Song MD, Abstract Co-Author: Nothing to Disclose

Kyeong Eun Shin MD, Abstract Co-Author: Nothing to Disclose

To directly compare the radiologist’s performances for the evaluation of diffuse interstitial lung disease (DILD) on matching CT images reconstructed with both Adaptive Statistical Iterative Reconstruction (ASIR) and conventional filtered back projection (FBP)

Included were 67 patients with known DILD (31 men, 36 women; mean age, 60 ± 14 years). Patients underwent volume thin-section CT using 64-row MDCT with size-based adjustment of automatic exposure control. Mean DLP of exams was 211 ± 71 mGy·cm. Transverse images (1.25 mm thickness) were reconstructed using both FBP (high spatial-frequency algorithm) and ASIR (50% blended). Five radiologists independently assigned scores for the extent of lung abnormalities (GGO, ground glass opacity; RET, reticular opacity; CON, consolidation; HC, honeycomb opacity) evaluated on two subsets (FBP and ASIR), which were at 1-month interval. These semi-quantitative results from FBP and ASIR subsets were compared statistically. P value less than .05 was considered as significant.

Regarding quantitative scores for fibrotic score (RET plus HC) and the overall extent of lung parenchymal abnormalities, intra-observer agreement was high between the scores for FBP-reconstruction and ASIR readings (r = .88 – .97; P < .01, Spearman rank correlation coefficient). Inter-observer reliabilities were moderate in both subsets (r = .64 for FBP; r = .66 for ASIR, Interclass correlation coefficient). When we compared the images of ASIR and FBP subsets fin terms of the total extent of parenchymal abnormalities, four of five observers overestimated significantly the extent of GGO on ASIR than FBP (mean difference of extent (ΔE) = 3.1%, P < .01). The remaining one observer estimated the extent equally (ΔE = 0.7%, P = .13). Two of five observers underestimated significantly the extent of RET on ASIR than FBP (ΔE = -1.4%, P < .01). The remaining three observers estimated them equally (ΔE = 0.3%, P = .42 – .64). As for the extent of CON or HC, no significant difference was noted.

On ASIR images, the extents of specific disease patterns in DILD can be differently estimated from those assessed on conventional FBP images.

ASIR images may be unfamiliar to radiologists and may cause different results when evaluating disease extent, particularly in DILD imaging, as compared with FBP reconstructed images.

Park, J, Chung, M, Kim, J, Yi, C, Chung, M, Lee, K, Hwang, J, Song, K, Shin, K, ASIR Imaging for Diffuse Interstitial Lung Disease: Better or Worse for Low-contrast High Spatial Frequency Lung Abnormalities?.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9001622.html