Objectives To develop and evaluate a novel multi-parameter MRI-based model (EFT1) for identifying high-risk metabolic dysfunction-associated steatohepatitis (MASH) to improve diagnostic accuracy and efficiency. Materials and methods A prospective study included 118 patients (55 male; 48 ± 13 years) with hepatic steatosis and metabolic risk factors. Among these, 80 patients were classified as having high-risk MASH. Magnetic resonance elastography (MRE), T1 mapping, chemical-shift encoded MRI for quantification of proton density fat fraction (PDFF) and R2* were performed, followed by liver biopsy. MRI parameters were analyzed and correlated with histological features. The EFT1 model was developed using logistic regression and full subset regression analysis on a training cohort (70%) and validated on a test cohort (30%). The performance of the model was compared with traditional scoring systems. Results Significant differences were observed in MRE, PDFF, and T1 between high-risk MASH and non-high-risk MASH groups. The EFT1 model, combining MRE, PDFF, and T1 showed strong diagnostic performance in both training (AUC 0.995, 95% CI 0.985–1.000) and test cohorts (AUC 0.995, 95% CI 0.979–1.000). At the optimal cut-off value of −0.431, the model achieved high sensitivity (98.2% training, 95.7% test) and specificity (96.3% training, 100% test). The EFT1 model outperformed traditional scoring systems (FIB-4, APRI, GPR) and showed comparable performance to the MAST score in identifying high-risk MASH. Conclusion The novel EFT1 model demonstrates reasonable performance in non-invasive identification of high-risk MASH patients compared to other models, achieving an appropriate balance between sensitivity and specificity. Clinical trial registration This study is registered with Chictr.org.cn (ChiCTR2400094017). Key Points <Emphasis Type="BoldItalic">Question</Emphasis> Current non-invasive scoring systems for high-risk MASH have limitations. A novel multi-parameter MRI-based model is proposed to improve diagnostic accuracy and efficiency. <Emphasis Type="BoldItalic">Findings</Emphasis> A novel multi-parameter MRI-based model (EFT1), incorporating MRE, PDFF, and T1 mapping, demonstrated higher accuracy in identifying high-risk MASH with an AUC of 0.995. <Emphasis Type="BoldItalic">Clinical relevance</Emphasis> The EFT1 model provides a highly accurate, non-invasive tool for early identification of high-risk MASH, facilitating timely intervention and personalized treatment strategies, potentially reducing disease progression and improving patient outcomes. Graphical Abstract

A novel multi-parameter MRI-based model for identification of high-risk metabolic dysfunction-associated steatohepatitis

期刊论文
发表日期：

2025年8月21日

查看全文

Wenxin Ma,
Xutong Huang,
Zhen Feng,
Wenli Tan,
Jinzhe Li,
Huamei Yan,
Yanxi Zheng,
Zhiwei Qin,
Fuhua Yan,
Huimin Lin,
Jie Yuan

摘要

Objectives

To develop and evaluate a novel multi-parameter MRI-based model (EFT1) for identifying high-risk metabolic dysfunction-associated steatohepatitis (MASH) to improve diagnostic accuracy and efficiency.

Materials and methods

A prospective study included 118 patients (55 male; 48 ± 13 years) with hepatic steatosis and metabolic risk factors. Among these, 80 patients were classified as having high-risk MASH. Magnetic resonance elastography (MRE), T1 mapping, chemical-shift encoded MRI for quantification of proton density fat fraction (PDFF) and R2* were performed, followed by liver biopsy. MRI parameters were analyzed and correlated with histological features. The EFT1 model was developed using logistic regression and full subset regression analysis on a training cohort (70%) and validated on a test cohort (30%). The performance of the model was compared with traditional scoring systems.

Results

Significant differences were observed in MRE, PDFF, and T1 between high-risk MASH and non-high-risk MASH groups. The EFT1 model, combining MRE, PDFF, and T1 showed strong diagnostic performance in both training (AUC 0.995, 95% CI 0.985–1.000) and test cohorts (AUC 0.995, 95% CI 0.979–1.000). At the optimal cut-off value of −0.431, the model achieved high sensitivity (98.2% training, 95.7% test) and specificity (96.3% training, 100% test). The EFT1 model outperformed traditional scoring systems (FIB-4, APRI, GPR) and showed comparable performance to the MAST score in identifying high-risk MASH.

Conclusion

The novel EFT1 model demonstrates reasonable performance in non-invasive identification of high-risk MASH patients compared to other models, achieving an appropriate balance between sensitivity and specificity.

Clinical trial registration

This study is registered with Chictr.org.cn (ChiCTR2400094017).

Key Points

Question Current non-invasive scoring systems for high-risk MASH have limitations. A novel multi-parameter MRI-based model is proposed to improve diagnostic accuracy and efficiency.

Findings A novel multi-parameter MRI-based model (EFT1), incorporating MRE, PDFF, and T1 mapping, demonstrated higher accuracy in identifying high-risk MASH with an AUC of 0.995.

Clinical relevance The EFT1 model provides a highly accurate, non-invasive tool for early identification of high-risk MASH, facilitating timely intervention and personalized treatment strategies, potentially reducing disease progression and improving patient outcomes.

Graphical Abstract