Li Xuehua, Hu Cicong, Wang Haipeng, Lin Yuqin, Li Jiaqiang, Cui Enming, Zhuang Xiaozhao, Li Jianpeng, Lu Jiahang, Zhang Ruonan, Wang Yangdi, Peng Zhenpeng, Sun Canhui, Li Ziping, Chen Minhu, Shi Li, Mao Ren, Huang Bingsheng, Feng Shi-Ting
Department of Radiology, The First Affiliated Hospital, Sun Yat-Sen University, 58 Zhongshan II Road, Guangzhou 510080, People's Republic of China.
Department of Radiology, The First Affiliated Hospital of Wenzhou Medical University, Nanbaixiang, Ouhai District, Wenzhou 325000, People's Republic of China.
iScience. 2024 Sep 28;27(10):111022. doi: 10.1016/j.isci.2024.111022. eCollection 2024 Oct 18.
The fingerprint features of visceral adipose tissue (VAT) are intricately linked to bowel damage (BD) in patients with Crohn's disease (CD). We aimed to develop a VAT fingerprint index (VAT-FI) using radiomics and deep learning features extracted from computed tomography (CT) images of 1,135 CD patients across six hospitals (training cohort, = 600; testing cohort, = 535) for predicting BD, and to compare it with a subcutaneous adipose tissue (SAT)-FI. VAT-FI exhibited greater predictive accuracy than SAT-FI in both training (area under the receiver operating characteristic curve [AUC] = 0.822 vs. AUC = 0.745, = 0.019) and testing (AUC = 0.791 vs. AUC = 0.687, = 0.019) cohorts. Multivariate logistic regression analysis highlighted VAT-FI as the sole significant predictor (training cohort: hazard ratio [HR] = 1.684, = 0.012; testing cohort: HR = 2.649, < 0.001). Through Shapley additive explanation (SHAP) analysis, we further quantitatively elucidated the predictive relationship between VAT-FI and BD, highlighting potential connections such as Radio479 (wavelet-HLH-first-order standard deviation)-Frequency loose stools-BD severity. VAT-FI offers an accurate means for characterizing BD, minimizing the need for extensive clinical data.
克罗恩病（CD）患者内脏脂肪组织（VAT）的指纹特征与肠道损伤（BD）密切相关。我们旨在利用从六家医院的1135例CD患者的计算机断层扫描（CT）图像中提取的放射组学和深度学习特征，开发一种VAT指纹指数（VAT-FI）来预测BD，并将其与皮下脂肪组织（SAT）-FI进行比较。在训练队列（n = 600）和测试队列（n = 535）中，VAT-FI在预测BD方面均表现出比SAT-FI更高的准确性。在训练队列中，受试者工作特征曲线下面积（AUC）为0.822，而SAT-FI的AUC为0.745，P = 0.019；在测试队列中，VAT-FI的AUC为0.791，SAT-FI的AUC为0.687，P = 0.019。多变量逻辑回归分析表明，VAT-FI是唯一显著的预测因子（训练队列：风险比[HR] = 1.684，P = 0.012；测试队列：HR = 2.649，P < 0.001）。通过夏普利加性解释（SHAP）分析，我们进一步定量阐明了VAT-FI与BD之间的预测关系，突出了潜在的联系，如Radio479（小波-HLH-一阶标准差）-腹泻频率-BD严重程度。VAT-FI为BD的特征描述提供了一种准确的方法，最大限度地减少了对大量临床数据的需求。
