Research on user portrait and management direction construction of vertebral compression fracture based on machine learning
基于机器学习的椎体压缩骨折用户画像与管理方向构建研究

摘要

目的 运用机器学习技术构建椎体压缩骨折患者画像，并探索其在患者管理方向的应用，目的以此提升诊疗水平和管理效率。方法 收集我院2024年1月-2025年1月100例椎体压缩骨折患者的临床资料，涵盖基本信息、影像数据、治疗方式、康复情况等。采用特征工程提取关键特征，运用聚类算法进行患者群体划分，构建用户画像。对比画像构建前后患者管理效果，评估画像对治疗决策、康复指导等方面的影响。结果 成功划分4类患者群体，画像构建后治疗方案匹配度提升，康复时间缩短，患者满意度提高。结论 机器学习构建的椎体压缩骨折用户画像有助于精准化患者管理，改善治疗和康复效果，有一定的应用价值。

Abstract

Objective To develop a patient profile for vertebral compression fracture (VCF) patients using machine learning technology and explore its application in clinical management, aiming to enhance diagnostic accuracy and treatment efficiency.
Methods Clinical data from 100 VCF patients at our hospital between January 2024 and January 2025 were collected, including basic information, imaging data, treatment methods, and rehabilitation status. Key features were extracted through feature engineering, and cluster algorithms were applied to classify patient groups and construct user profiles. The effectiveness of these profiles was compared with pre-and post-treatment management outcomes to evaluate their impact on therapeutic decisions and rehabilitation guidance.
Results Four distinct patient groups were successfully classified. Post-profile construction showed improved treatment plan matching, shortened recovery time, and increased patient satisfaction.
Conclusion   Machine learning-based VCF user profiles facilitate precise patient management, improve treatment and rehabilitation outcomes, and demonstrate practical value in clinical practice.