Introduction:
Assess reproducibility of detection, staging, and grading of non-Hodgkin lymphoma (NHL) using whole-body diffusion-weighted imaging with background body signal suppression (WB-DWIBS).

Material and methods:
Thirty NHL patients underwent WB-DWIBS, divided into 2 groups according to staging and grading. Image analysis and apparent diffusion coefficient (ADC) measurement of the largest lymph node in each group were performed by 2 observers. Inter-observer agreement was performed.

Results:
Overall inter-observer agreement for detection of NHL was excellent (κ = 0.843; 92.05%) with excellent inter-observer agreement of nodal disease (cervical, thoracic and abdominal) (κ = 0.783, 0.769, and 0.856; 96.67%, 90.0%, and 93.3% respectively), extra-nodal disease (κ = 1; 100%), and splenic involvement (κ = 0.67; 83.3%). The overall inter-observer agreement of DWIBS in staging of NHL was excellent (κ = 0.90; 94.9%) with excellent inter-observer agreement for stage I (κ = 0.93; 96.4%), stage II (κ = 0.90; 94.8%), stage III (κ = 0.89; 94.6%), and stage IV (κ = 0.88; 94.0%). There was significant difference between ADC in stage I, II (0.77 ± 0.13, 0.85 ± 0.09 × 10^-3 mm²/s), and stage III, IV (0.63 ± 0.08, 0.64 ± 0.11 × 10^-3 mm²/s, p < 0.002, < 0.001). Interclass correlation showed almost perfect agreement for ADC measurement in staging and grading groups (r = 0.96 and r = 0.85, respectively, p < 0.001). There was significant difference between ADC in aggressive lymphoma (0.65 ± 0.1, 0.67 ± 0.13 × 10^-3 mm²/s) and indolent lymphoma (0.76 ± 0.14, 0.84 ± 0.09 × 10^-3 mm²/s, p < 0.028, < 0.001).

Conclusions:
DWIBS is reproducible for detection and staging of nodal and extra-nodal involvement in patients with NHL. ADC can quantitatively participate in the staging and grading of NHL.