Fused high b-value diffusion weighted and T2-weighted MR images in staging of pediatric Hodgkin's lymphoma: A pilot study

Purpose: To evaluate the value of fused high b-value diffusion weighted and T2-weighted MRI compared to T1-weighted imaging, T2-weighted imaging and DWI for staging pediatric Hodgkin's lymphoma. Method: 21 consecutive pediatric patients who underwent whole-body MRI at Hodgkin's lymphoma staging were included. Fused, colorized DWI/T2-weighted images were created. Image sets consisting of (a) T1-weighted, T2-weighted and DWI images and (b) T1-weighted, T2-weighted, DWI and DWI/T2-weighted fused images were reviewed by a radiologist using a cross-over design with blinding and randomization. Scoring was performed using a standardized form, based on detection, characterization and reading time, using a FDG-PET/CT based reference standard. Test characteristics, test agreement to a FDG-PET/CT based reference standard and reading times were calculated. Results: Agreement for whole-body MRI without fused images and FDG-PET/CT was very good for nodal staging (kappa = 0.86, 95% CI 0.78-0.93) and extra-nodal staging (kappa = 0.90, 95% CI 0.71-1.09). Agreement improved with the addition of the fused DWI/T2-weighted images (kappa = 0.92 95% CI 0.87-0.97 (nodal staging), kappa = 0.92 95% CI 0.76-1.08 (extra-nodal staging). Sensitivity and specificity for staging nodal disease were 99 % and 95% respectively for whole-body MRI including DWI/T2-weighted fused images (versus 88 % and 97 % without fused images) and 100 % and 99 % for extra-nodal disease (83 % and 100 % without fused images). Disease stage for MRI including fused DWI/T2-weighted images agreed with the reference standard in 18 out of 21 patients. Conclusions: The addition of DWI/T2-weighted fusion images to T1-weighted, T2-weighted and DWI whole-body MRI might shorten reading times and might improve the diagnostic performance of whole-body MRI in staging pediatric Hodgkin's lymphoma.