2. For normal sheep lungs, these bias values corresponded to 1.0% (0.0 to 2.6%) for Vtotal and 1.2% (0.0 to 3.6%) for Mtotal (percentage of the respective value obtained by whole-lung CT analysis). The corresponding values were 1.2% (0.0 to 4.2%) and 1.2% kinase inhibitor Gemcitabine (0.0 to 4.8%) for Vtotal and Mtotal, respectively, for opacified sheep lungs. For normal porcine lungs, the corresponding bias values were 0.9% (0.1 to 2.3%) for Vtotal and 0.5% (0.0 to 3.6%) for Mtotal. Finally, for porcine lungs with opacifications these bias values were 1.1% (0.1 to 5.6%) and 1.5% (0.1 to 8.5%) for Vtotal and Mtotal, respectively.Table 2Agreement between extrapolation from 10 CT slices and whole-lung analysisThe bias between methods for volume and mass of differently aerated lung compartments never exceeded 0.

5% and the respective limits of agreement were below 2.5% of Vtotal or Mtotal, respectively (Table (Table22).The accuracy of extrapolation for varying numbers (15 to 5) of CT slices is illustrated in Figure Figure2.2. For extrapolation from 10 or more CT slices, the bias values for Vtotal, Mtotal and%Mnon were very close to 0 and the limits of agreement were narrow. When less than 10 CT slices were used for extrapolation, bias values started to diverge from 0 and the 95% limits of agreement started to become considerably wider (Figure (Figure22).Variations in CT slice thickness between 5 and 10 mm did not have an effect on the accuracy of extrapolation (Table (Table33).Table 3Accuracy of extrapolation for CT sections with different slice thicknessResults on the accuracy of the extrapolation method for detecting intra-individual changes between two consecutive CTs are shown in Table Table44 and Figure Figure33.

Table 4Intra-individual changes between 58 pairs of consecutive CT scans in sheepFigure 3Accuracy of extrapolation for assessing changes of lung volume and masses between consecutive CTs. The accuracy of the extrapolation method for assessing changes (delta) of lung volume and masses between consecutive CTs was analyzed by the Bland-Altman …DiscussionOur study demonstrates the excellent accuracy of the extrapolation method for calculation of parameters characterizing the entire lung from only 10 reference CT slices. The extrapolation method was also very accurate in detecting changes between two consecutive CTs.

Our results confirm a previous report on the accuracy of such an extrapolation method in humans [18] and demonstrate that the extrapolation method is robust against variations in thoracic anatomy. The bias AV-951 between extrapolation and whole-lung CT analysis increased progressively when less than 10 reference CT slices were used for extrapolation.Quantification of lung aeration by CT has become an important research tool for studying normal and pathological aspects of lung aeration as well as the effects of mechanical ventilation on lung aeration.