Evaluating boost diffusion sampling for improved abdominal imaging using prototype slice-specific shimming method

To determine whether prototype integrated slice-specific shimming (iShim) diffusion-weighted imaging (DWI) combined with Boost diffusion sampling improves image quality and diagnostic performance for focal liver lesions compared with conventional DWI and alternative iShim sampling schemes.

Eighty-five patients with focal liver lesions underwent 3T MRI with conventional DWI and iShim-DWI using Boost, flow-compensated (FC), monopolar (Mono), FC with five lobes (FC5), and stimulated echo acquisition mode (Steam) sampling at b-values of 0 and 1000 s/mm². Two radiologists independently scored six image-quality features. Signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and apparent diffusion coefficient (ADC) were measured. Interreader agreement, between-sequence differences, ADC correlations and agreement, and diagnostic performance for differentiating benign from malignant lesions were assessed.

Boost iShim-DWI yielded significantly better artifact, diagnostic-confidence, and total-image-quality scores than all other sequences (p < 0.05) and achieved the highest lesion SNR, liver SNR, and CNR (p < 0.05). Interreader agreement was excellent (ICC = 0.839-0.953). ADC values correlated strongly across sequences (rₛ = 0.812-0.961, p < 0.001). Boost iShim-DWI achieved the highest diagnostic performance (AUC = 0.942; 95% CI 0.897-0.988). Discussion: Boost iShim-DWI improves liver DWI image quality while preserving ADC comparability and may enhance differentiation of benign and malignant focal liver lesions.