Objectives <p>This study aimed to develop and validate a fat-corrected virtual magnetic resonance elastography (FC-vMRE) framework based on diffusion-weighted imaging (DWI) to assess liver fibrosis in patients with metabolic dysfunction-associated steatotic liver disease (MASLD).</p> Materials and methods <p>A total of 463 MASLD patients underwent multi-<i>b</i>-values (0–1500 s/mm²) DWI acquisition, proton density fat fraction (PDFF), and conventional magnetic resonance elastography (MRE). Apparent diffusion coefficient (ADC) parameters were calculated with or without fat correction (based on PDFF). Using the training cohort (<i>n</i> = 361), ADC-to-MRE formulas were derived to compute virtual MRE (vMRE) and FC-vMRE. In the validation cohort (<i>n</i> = 102, with biopsy), the diagnostic performance of vMRE, FC-vMRE, and MRE for fibrosis staging was compared.</p> Results <p>ADC<sub>200-1200</sub> demonstrated the strongest correlation with MRE values (<i>R</i> = −0.706, <i>p</i> &lt; 0.001), yielding the formula: FC-vMRE = 6.50 − 3.13 × ADC<i>w</i>. Compared with vMRE, FC-vMRE showed superior agreement with MRE (bias: −0.001 kPa vs 0.469 kPa; intraclass correlation coefficient: 0.666 vs 0.381). In staging performance, FC‑vMRE significantly outperformed vMRE across all fibrosis stages. FC-vMRE showed promising diagnostic performance to MRE in detecting ≥ F2 fibrosis (AUC: 0.761 vs 0.848, <i>p</i> = 0.053), ≥ F3 fibrosis (AUC: 0.756 vs 0.818, <i>p</i> = 0.066), and cirrhosis (AUC: 0.838 vs 0.914, <i>p</i> = 0.065).</p> Conclusion <p>FC-vMRE provides a clinically promising alternative requiring only standard MRI equipment, effectively correcting for fat-related confounding in MASLD, and demonstrating encouraging diagnostic concordance with MRE and histology. This approach shows potential for broader clinical implementation.</p> Critical relevance statement <p>Fat-corrected virtual MRE using diffusion-weighted imaging provides promising fibrosis staging accuracy to conventional magnetic resonance elastography in metabolic dysfunction-associated steatotic liver disease patients without requiring specialized hardware.</p> Key Points <p><UnorderedList Mark="Bullet"> <ItemContent> <p>Fat infiltration confounds diffusion measurements in metabolic steatotic liver disease.</p> </ItemContent> <ItemContent> <p>Fat-corrected virtual MRE showed superior agreement with standard MRE.</p> </ItemContent> <ItemContent> <p>Hardware-free fibrosis assessment enables accessible liver stiffness measurement.</p> </ItemContent> </UnorderedList></p> Graphical Abstract <p></p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Development and validation of fat-corrected virtual MR elastography to assess fibrosis stage in metabolic dysfunction-associated steatotic liver disease

  • Jie Yuan,
  • Xinxin Xu,
  • Fuhua Yan,
  • Zhiwei Qin,
  • Huamei Yan,
  • Wenjie Yang,
  • Huimin Lin

摘要

Objectives

This study aimed to develop and validate a fat-corrected virtual magnetic resonance elastography (FC-vMRE) framework based on diffusion-weighted imaging (DWI) to assess liver fibrosis in patients with metabolic dysfunction-associated steatotic liver disease (MASLD).

Materials and methods

A total of 463 MASLD patients underwent multi-b-values (0–1500 s/mm²) DWI acquisition, proton density fat fraction (PDFF), and conventional magnetic resonance elastography (MRE). Apparent diffusion coefficient (ADC) parameters were calculated with or without fat correction (based on PDFF). Using the training cohort (n = 361), ADC-to-MRE formulas were derived to compute virtual MRE (vMRE) and FC-vMRE. In the validation cohort (n = 102, with biopsy), the diagnostic performance of vMRE, FC-vMRE, and MRE for fibrosis staging was compared.

Results

ADC200-1200 demonstrated the strongest correlation with MRE values (R = −0.706, p < 0.001), yielding the formula: FC-vMRE = 6.50 − 3.13 × ADCw. Compared with vMRE, FC-vMRE showed superior agreement with MRE (bias: −0.001 kPa vs 0.469 kPa; intraclass correlation coefficient: 0.666 vs 0.381). In staging performance, FC‑vMRE significantly outperformed vMRE across all fibrosis stages. FC-vMRE showed promising diagnostic performance to MRE in detecting ≥ F2 fibrosis (AUC: 0.761 vs 0.848, p = 0.053), ≥ F3 fibrosis (AUC: 0.756 vs 0.818, p = 0.066), and cirrhosis (AUC: 0.838 vs 0.914, p = 0.065).

Conclusion

FC-vMRE provides a clinically promising alternative requiring only standard MRI equipment, effectively correcting for fat-related confounding in MASLD, and demonstrating encouraging diagnostic concordance with MRE and histology. This approach shows potential for broader clinical implementation.

Critical relevance statement

Fat-corrected virtual MRE using diffusion-weighted imaging provides promising fibrosis staging accuracy to conventional magnetic resonance elastography in metabolic dysfunction-associated steatotic liver disease patients without requiring specialized hardware.

Key Points

Fat infiltration confounds diffusion measurements in metabolic steatotic liver disease.

Fat-corrected virtual MRE showed superior agreement with standard MRE.

Hardware-free fibrosis assessment enables accessible liver stiffness measurement.

Graphical Abstract