Objectives Standardizing magnetic resonance enterography (MRE)-defined transmural healing (TH) remains challenging in Crohn’s disease (CD) despite its prognostic superiority. We aimed to evaluate seven conventional MRE-defined TH criteria and develop a machine-learning-optimized model for improved TH assessment. Materials and methods In this double-center study, 467 active CD patients with 1263 MRE scans were stratified into three cohorts. Cohort 1 (n = 341) enabled retrospectively dual-metric comparison (attainment rate/prognostic protection) of seven MRE-defined TH criteria. Leveraging their strengths, we developed five machine-learning models for TH assessment to identify the optimal one. Semi-external validation was performed in prospective Ustekinumab (n = 92) and Upadacitinib (n = 34) cohorts. Results Among seven conventional TH criteria, magnetic resonance index of activity (MaRIA), C-score, and simplified MaRIA (sMaRIA) demonstrated higher attainment rates (23.75%/28.74%/41.64%) and lower disease progression rates (14.81%/18.37%/25.35%). Random forest (RF) model showed the most favorable overall performance across cohorts: Cohort 1 (AUC, 0.82 vs. 0.71–0.81), Ustekinumab (AUC, 0.83 vs. 0.73–0.81), and Upadacitinib (AUC, 0.80 vs. 0.58–0.80) cohorts. Dual-metric evaluation identified the RF model and C-score as clinically applicable tools. Notably, the RF model (namely SYSU-score) was more strongly associated with lower progression risk than C-score: in the Ustekinumab cohort, SYSU-score-defined TH showed lower disease progression risk (HR = 0.07, p < 0.001) than C-score (HR = 0.15, p < 0.001); Upadacitinib cohort validated SYSU-score as the only system significantly associated with lower progression risk (HR = 0.20, p < 0.05). Conclusions We established a validated machine-learning-derived TH criterion integrating the strengths of conventional MRE-defined systems. SYSU-score-defined TH status was associated with lower progression risk with robust prognostic discrimination, advancing standardized TH assessment for clinical implementation. Key Points <Emphasis Type="BoldItalic">Question</Emphasis> Among available MRI-based transmural healing (TH) criteria for Crohn’s disease (CD), which are the most clinically applicable, and can a new model improve standardized prognostic assessment? <Emphasis Type="BoldItalic">Findings</Emphasis> MaRIA, C-score, and sMaRIA showed relatively favorable clinical applicability, and the machine-learning-derived SYSU-score demonstrated robust prognostic performance in internal and semi-external validation. <Emphasis Type="BoldItalic">Clinical relevance</Emphasis> SYSU-score may support more standardized MRI-based assessment of TH and improve risk stratification for disease progression in patients with CD. Graphical Abstract

Development and validation of the SYSU-score for MRI-based transmural healing assessment in Crohn’s disease: a dual-center study

期刊论文
发表日期：

2026年6月19日

查看全文

Qingzhu Zheng,
Qiaochu Zhao,
Weitao He,
Lili Huang,
Xiaodi Shen,
Luyao Wu,
Yaoqi Ke,
Weikai Zheng,
Yangdi Wang,
Yujun Chen,
Ren Mao,
Zhenpeng Peng,
Shi-Ting Feng,
Ruonan Zhang,
Xuehua Li

摘要

Objectives

Standardizing magnetic resonance enterography (MRE)-defined transmural healing (TH) remains challenging in Crohn’s disease (CD) despite its prognostic superiority. We aimed to evaluate seven conventional MRE-defined TH criteria and develop a machine-learning-optimized model for improved TH assessment.

Materials and methods

In this double-center study, 467 active CD patients with 1263 MRE scans were stratified into three cohorts. Cohort 1 (n = 341) enabled retrospectively dual-metric comparison (attainment rate/prognostic protection) of seven MRE-defined TH criteria. Leveraging their strengths, we developed five machine-learning models for TH assessment to identify the optimal one. Semi-external validation was performed in prospective Ustekinumab (n = 92) and Upadacitinib (n = 34) cohorts.

Results

Among seven conventional TH criteria, magnetic resonance index of activity (MaRIA), C-score, and simplified MaRIA (sMaRIA) demonstrated higher attainment rates (23.75%/28.74%/41.64%) and lower disease progression rates (14.81%/18.37%/25.35%). Random forest (RF) model showed the most favorable overall performance across cohorts: Cohort 1 (AUC, 0.82 vs. 0.71–0.81), Ustekinumab (AUC, 0.83 vs. 0.73–0.81), and Upadacitinib (AUC, 0.80 vs. 0.58–0.80) cohorts. Dual-metric evaluation identified the RF model and C-score as clinically applicable tools. Notably, the RF model (namely SYSU-score) was more strongly associated with lower progression risk than C-score: in the Ustekinumab cohort, SYSU-score-defined TH showed lower disease progression risk (HR = 0.07, p < 0.001) than C-score (HR = 0.15, p < 0.001); Upadacitinib cohort validated SYSU-score as the only system significantly associated with lower progression risk (HR = 0.20, p < 0.05).

Conclusions

We established a validated machine-learning-derived TH criterion integrating the strengths of conventional MRE-defined systems. SYSU-score-defined TH status was associated with lower progression risk with robust prognostic discrimination, advancing standardized TH assessment for clinical implementation.

Key Points

Question Among available MRI-based transmural healing (TH) criteria for Crohn’s disease (CD), which are the most clinically applicable, and can a new model improve standardized prognostic assessment?

Findings MaRIA, C-score, and sMaRIA showed relatively favorable clinical applicability, and the machine-learning-derived SYSU-score demonstrated robust prognostic performance in internal and semi-external validation.

Clinical relevance SYSU-score may support more standardized MRI-based assessment of TH and improve risk stratification for disease progression in patients with CD.

Graphical Abstract