相关实验视频

Updated: Jan 11, 2026

Prospective, Randomized, and Controlled Study of a Human Umbilical Cord Mesenchymal Stem Cell Injection for Treating Diabetic Foot Ulcers

Prospective, Randomized, and Controlled Study of a Human Umbilical Cord Mesenchymal Stem Cell Injection for Treating Diabetic Foot Ulcers

Published on: March 3, 2023

可解释的机器学习模型用于预测糖尿病足患者的复发情况

Weijiao Mou¹, Waiping Shan¹, Shiyan Yu¹

¹Department of Endocrinology, School of Medicine, Chongqing University Central Hospital, Chongqing Emergency Medical Centre, Chongqing University, Chongqing, China.

BMJ open diabetes research & care

|November 13, 2025

概括

机器学习准确地预测了糖尿病足的复发. XGBoost模型确定了关键的风险因素,为改善患者管理和及时干预提供了潜力.

科学领域:

医疗信息学医疗信息学
计算生物学计算生物学
糖尿病研究研究糖尿病研究

背景情况:

糖尿病足 (DFU) 由于慢性疾病的进展和高复发率而构成重大挑战.
准确预测DFU复发对于及时干预和改善患者结果至关重要.
这项研究的重点是开发一个机器学习 (ML) 模型来预测3年DFU复发风险.

研究的目的:

开发和验证一种机器学习模型,用于预测糖尿病足的3年复发风险.
使用ML技术识别与DFU复发相关的关键风险因素.
增强糖尿病足患者的临床决策能力.

主要方法:

一组494名DFU患者被分为训练和测试组.
功能选择涉及LASSO,mRMR,费舍尔分数和RFE; 评估了七个ML算法.
模型性能使用AUROC进行评估,通过Platt缩放进行校准,通过SHAP分析进行可解释性.

主要成果:

该XGBoost模型实现了高预测性能,AUROC为0.924.
校准的XGBoost模型显示了良好的校准,Brier分数为0.096.
SHAP分析证实了已确定的风险因素的临床相关性,与现有知识保持一致.

结论:

开发的XGBoost模型显示出强大的预测准确性和DFU复发的临床相关性.
建议使用更大的样本大小进行进一步的多中心验证,以提高可概括性.
该模型显示了改善患者管理和DFU护理结果的前景.

关键词:

糖尿病患者的糖尿病脚.

更多相关视频

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

A Simplified Technique for Producing an Ischemic Wound Model

A Simplified Technique for Producing an Ischemic Wound Model

Published on: May 2, 2012

相关概念视频

Receiver Operating Characteristic Plot

Receiver Operating Characteristic Plot

A ROC (Receiver Operating Characteristic) plot is a graphical tool used to assess the performance of a binary classification model by illustrating the trade-off between sensitivity (true positive rate) and specificity (false positive rate). By plotting sensitivity against 1 - specificity across various threshold settings, the ROC curve shows how well the model distinguishes between classes, with a curve closer to the top-left corner indicating a more accurate model. The area under the ROC curve...