Application of integrated slice-specific dynamic shimming diffusion weighted imaging (DWI) in upper abdomen

To compare image quality between integrated slice-specific dynamic shimming diffusion-weighted imaging (iShim-DWI) and conventional single-shot echo-planar (SS-EPI) DWI in the upper abdomen, and to evaluate performance differences within iShim-DWI using high (800 s/mm²) versus ultra-high (1600 s/mm²) b-values.

A cohort of 162 consecutive participants was included. Among them, 110 underwent both iShim-DWI (b = 800 s/mm²) and SS-EPI-DWI (b = 800 s/mm²), while the remaining 52 received iShim-DWI at both b = 800 s/mm² and b = 1600 s/mm². Qualitative assessment focused on image quality features including fat suppression, distortion, artifacts, overall image quality and lesion confidence score. Lesion detection efficiency was compared. Quantitative evaluation included signal intensity (SI), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and ADC-related parameters of normal upper abdominal organs and lesions.

iShim-DWI received significantly higher scores in artifact reduction overall image quality and lesion confidence score compared with SS-EPI-DWI (all P < 0.0001). Fat suppression did not differ significantly between the two sequences (P = 0.0703). Five additional lesions were detected by iShim-DWI that were not visible on SS-EPI-DWI. iShim-DWI yielded significantly higher SNR and CNR for both normal organs (P < 0.001) and lesions (P = 0.0023; 0.0024), along with higher ΔADC and lower ADC variability (SD) compared with SS-EPI-DWI (both P < 0.05). Within iShim-DWI, the b = 800 s/mm² protocol provided significantly higher lesion SNR (P = 0.0231) and CNR (P = 0.0438) than b = 1600 s/mm², while lesion confidence scores were similarly high (P > 0.9999).

iShim-DWI provides substantially improved image quality, more stable ADC measurements, and better lesion detection and visualization in the upper abdomen compared with SS-EPI-DWI. In our cohort, iShim-DWI at b = 800 s/mm² achieves a favorable balance between lesion contrast and SNR, and may serve as a preferred routine clinical option. Further validation in larger cohorts is warranted.