Department of Gastroenterology, the First Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215006, China; Suzhou Clinical Centre of Digestive Diseases, Suzhou, Jiangsu 215006, China.
Department of Gastroenterology, the First Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215006, China; Department of General Surgery, Jintan Hospital Affiliated to Jiangsu University, Changzhou, Jiangsu 213299, China.
Int J Med Inform. 2024 Apr;184:105341. doi: 10.1016/j.ijmedinf.2024.105341. Epub 2024 Jan 20.
Aim to establish a multimodal model for predicting severe acute pancreatitis (SAP) using machine learning (ML) and deep learning (DL).
In this multicentre retrospective study, patients diagnosed with acute pancreatitis at admission were enrolled from January 2017 to December 2021. Clinical information within 24 h and CT scans within 72 h of admission were collected. First, we trained Model α based on clinical features selected by least absolute shrinkage and selection operator analysis. Second, radiomics features were extracted from 3D-CT scans and Model β was developed on the features after dimensionality reduction using principal component analysis. Third, Model γ was trained on 2D-CT images. Lastly, a multimodal model, namely PrismSAP, was constructed based on aforementioned features in the training set. The predictive accuracy of PrismSAP was verified in the validation and internal test sets and further validated in the external test set. Model performance was evaluated using area under the curve (AUC), accuracy, sensitivity, specificity, recall, precision and F1-score.
A total of 1,221 eligible patients were randomly split into a training set (n = 864), a validation set (n = 209) and an internal test set (n = 148). Data of 266 patients were for external testing. In the external test set, PrismSAP performed best with the highest AUC of 0.916 (0.873-0.960) among all models [Model α: 0.709 (0.618-0.800); Model β: 0.749 (0.675-0.824); Model γ: 0.687 (0.592-0.782); MCTSI: 0.778 (0.698-0.857); RANSON: 0.642 (0.559-0.725); BISAP: 0.751 (0.668-0.833); SABP: 0.710 (0.621-0.798)].
The proposed multimodal model outperformed any single-modality models and traditional scoring systems.
旨在利用机器学习（ML）和深度学习（DL）建立预测重症急性胰腺炎（SAP）的多模态模型。
在这项多中心回顾性研究中，纳入了 2017 年 1 月至 2021 年 12 月入院时被诊断为急性胰腺炎的患者。收集入院 24 小时内的临床信息和入院 72 小时内的 CT 扫描。首先，我们基于最小绝对收缩和选择算子分析选择的临床特征训练模型α。其次，从 3D-CT 扫描中提取放射组学特征，并使用主成分分析对降维后的特征进行模型β的开发。然后，在 2D-CT 图像上训练模型γ。最后，基于训练集中的上述特征构建多模态模型，即 PrismSAP。在验证集和内部测试集中验证 PrismSAP 的预测准确性，并在外部测试集中进一步验证。使用曲线下面积（AUC）、准确性、灵敏度、特异性、召回率、精度和 F1 评分评估模型性能。
共有 1221 名符合条件的患者被随机分为训练集（n=864）、验证集（n=209）和内部测试集（n=148）。266 名患者的数据用于外部测试。在外部测试集中，PrismSAP 表现最佳，所有模型中 AUC 最高为 0.916（0.873-0.960）[模型α：0.709（0.618-0.800）；模型β：0.749（0.675-0.824）；模型γ：0.687（0.592-0.782）；MCTSI：0.778（0.698-0.857）；RANSON：0.642（0.559-0.725）；BISAP：0.751（0.668-0.833）；SABP：0.710（0.621-0.798）]。
所提出的多模态模型优于任何单一模态模型和传统评分系统。
