Patrick Hollmann MD, Abstract Co-Author: Nothing to Disclose

Sabine Dettmer, Abstract Co-Author: Nothing to Disclose

Olga Solyanik, Abstract Co-Author: Nothing to Disclose

Till Kaireit, Abstract Co-Author: Nothing to Disclose

Claudia de Wall, Abstract Co-Author: Nothing to Disclose

Frank K. Wacker MD, Abstract Co-Author: Research Grant, Siemens AG

Hoen-Oh Shin MD, Presenter: Nothing to Disclose

To determine whether density mappings of HU values in inspiration and expiration computed tomography (CT) datasets are more precise for detection of air trapping (AT) than a threshold-based method in expiration alone.

In a prospective study, forty-nine spirometrically controlled non-enhanced lung CT during inspiration and expiration were performed in lung transplant patients. Assessment of AT was performed using both, a standard threshold-based method in expiration (HU values from -910 HU to -850 HU) and density mappings of HU values in inspiration and expiration. To achieve a point-to-point correspondence of HU values in the density mappings, automatic segmentation of the lungs followed by non-rigid registration of the inspiration and expiration datasets were performed. Differences in HU-values ≤120 HU and ≤140 HU were measured using three threshold ranges (-910 HU to -800 HU, -910 HU to -750 HU, and -910 HU to -700 HU). Both methods were correlated with pulmonary function tests (PFT) using residual volume (RV) and the ratio of RV and total lung capacity (TLC) as correlating parameters.

Density mappings correlated significantly better with PFT than the threshold-based method. Cochran Q and McNemar testing revealed statistically significant differences of PFT, threshold-based method and different density mappings (p<0,05). Kappa analysis showed moderate agreement between RV and density mappings. Best result was reached for a threshold range of -910 HU to -750 HU for HU changes ≤140 HU (κ =0,471). RV and the threshold-based method revealed slight agreement (κ=0,052). Kappa statistics revealed substantial agreement between RV/TLC and density mappings. Best result was achieved for a threshold range of -910 HU to -700 HU for HU changes ≤140 HU (κ=0,675). Slight agreement was detected correlating the threshold-based method and RV/TLC (κ=0,08).

Density mappings of HU-values showed significantly better agreement to PFT than the threshold-based method for detection of air trapping. Differences in HU-values ≤140 HU and a threshold range of -910 HU to -700 HU showed the best agreement when compared to PFT.

The proposed method allows better assessment of AT for early treatment of bronchiolitis obliterans syndrome in lung transplant patients.

Hollmann, P, Dettmer, S, Solyanik, O, Kaireit, T, de Wall, C, Wacker, F, Shin, H, Multi Detector Row CT in Lung Transplants: Assessment of Air Trapping Using Density Mappings of Inspiration and Expiration Datasets.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12028320.html