使用多机器学习模型预测急性胰腺炎的严重程度:构建一个在线预测平台.
Jie Cao1, Shike Long2,3, Huan Liu1
1Department of Gastroenterology, The Second Affiliated Hospital of Guilin Medical University, Guilin, Guangxi, China.
Frontiers in cellular and infection microbiology
|March 16, 2026
概括
早期评估急性胰腺炎 (AP) 严重程度至关重要. 一个使用机器学习的新网络工具准确地预测了严重的AP风险,帮助及时干预.
科学领域:
- 医疗信息学 医疗信息学
- 医疗保健中的机器学习
- 临床决策支持系统 临床决策支持系统
背景情况:
- 早期评估急性胰腺炎 (AP) 严重程度对于患者的治疗结果至关重要.
- 开发了一个基于网络的计算器,以估计AP进展到严重AP的可能性.
- 该工具旨在促进迅速的临床决策.
研究的目的:
- 开发和验证严重急性胰腺炎 (SAP) 的预测模型.
- 确定预测SAP开发的关键临床变量.
- 为SAP风险评估创建一个用户友好的临床应用程序.
主要方法:
- 对1289个AP患者记录进行了回顾性分析.
- 开发和比较10个机器学习模型,包括LightGBM,XGBoost和神经网络.
- 使用随机森林和LASSO进行特征选择;通过AUC和SHAP/PDP进行模型评估以获得可解释性.
主要成果:
- 轻GBM模型实现了最高的预测准确度,AUC为0.9726 (训练) 和0.9301 (测试).
- 确定了SAP的关键预测因素: (Ca),白细胞计数 (WBC),α-hydroxybutyrate脱酶 (α-HBDH) 和葡萄糖 (Glu).
- Ca显示出负相关性,而WBC,α-HBDH和Glu与SAP风险呈正相关性.
结论:
- 临床变量Ca,WBC,α-HBDH和Glu对SAP具有显著的预测能力.
- 为快速SAP风险评估开发了一个临床在线平台.
- 该平台支持及时干预和改善严重急性胰腺炎患者管理.
更多相关视频
07:15Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
7.6K
07:10Establishment of a Mouse Severe Acute Pancreatitis Model using Retrograde Injection of Sodium Taurocholate into the Biliopancreatic Duct
Published on: April 1, 2022
5.2K
相关概念视频
Acute Pancreatitis II: Clinical Manifestations and Management
1.0K
Acute pancreatitis presents a complex medical emergency characterized by rapid onset inflammation of the pancreas, demanding timely diagnosis and management to prevent complications. The condition primarily manifests through severe upper abdominal pain that often radiates to the back. This pain intensifies following the consumption of fatty foods. Accompanying symptoms such as nausea, vomiting, abdominal distention, fever, dyspnea, cyanosis, and jaundice can vary in intensity but significantly...
1.0K
Chronic Pancreatitis II: Collaborative Care
457
The management of chronic pancreatitis is multifaceted, involving a comprehensive approach that includes thorough assessment, diagnostic testing, and a variety of management strategies.
Assessment:
Assessment:
457
