Fig 1.
MRI sequences of a patient with stage T1b NSCLC in the left upper lobe.
A. T2 HASTE sequence (non-gated). B. ADC-map. C. DWI (calculated b-value of 1400). D. Fusion of T2 and DWI, color represents diffusion restriction.
Table 1.
Patient characteristics, measured tumor volumes for all 16 tumors in 15 patients.
Fig 2.
Example of PET-CT vs. MR including GTV contours in a patient with a large adenocarcinoma in the left hilum with concomitant atelectasis of the upper lobe.
A. Axial slice of the contrast enhanced planning CT. B. FDG-PET-scan. C. Anatomically registered FDG-PET/CT. The turquois contour (colorwash) indicates the FDG-PET-based delineation of the GTV (stGTV), the corresponding DWI-based GTV (dwGTV) is overlaid as a purple line. D. T2-TSE axial. E. DWI-sequence (calculated b-value of 1400). F. Anatomically registered T2/DWI. The purple contour (colorwash) indicates the DWI-based delineation of the GTV (dwGTV), the corresponding FDG-PET-based GTV (stGTV) is shown as a turqois line.
Fig 3.
Scatter plots and Bland-Altman analysis of tumor volume A. Linear regression analysis with fitted line. p = 0.005. B. Bland-Altman analysis showing upper and lower limits of agreement (mean volume difference ± 2 standard deviations) between stGTV (x-axis) and dwGTV (y-axis). The p-value indicates a significant dependence of measured volume differences from the average tumor size, as tested with linear regression analysis. C. Bland-Altman analysis showing upper and lower limits of agreement (mean volume difference ± 2 standard deviations) between stGTV (x-axis) and dwGTV (y-axis) in logarithmic scale. In logarithmic form, no significant dependence of measured volume differences from the average tumor size was observed (tested with linear regression analysis).