机器学习和可解释性研究预测精神分裂症的30天计划外再接收风险:一项回顾性研究
Yuting Tan1,2, Guiling Chen1,3, Shuge Wang1
1Department of Nursing, Tianyou Hospital Affiliated to Wuhan University of Science and Technology, Wuhan, Hubei, People's Republic of China.
Neuropsychiatric disease and treatment
|August 4, 2025
概括
机器学习准确地预测了精神分裂症患者30天的无计划再入院风险. 沙普利添加式扩展 (SHAP) 为早期干预和改善护理过渡提供了个性化的风险洞察力.
科学领域:
- 精神病学和心理健康 精神病学和心理健康
- 医疗信息学 医疗信息学
- 医疗保健中的机器学习
背景情况:
- 精神分裂症带来了重大挑战,包括高率的无计划再接收.
- 预测模型对于识别有风险的患者和优化护理至关重要.
- 电子医疗记录 (EMR) 为开发此类模型提供了丰富的数据.
研究的目的:
- 开发一种机器学习 (ML) 模型,用于预测精神分裂症患者的30天无计划再入院 (UPR).
- 使用夏普利添加式扩展 (SHAP) 进行模型解释性和个性化风险评估.
- 支持对精神分裂症护理的早期临床干预策略.
主要方法:
- 对1,123名精神分裂症患者的EMR进行了回顾性分析.
- 开发和比较五个ML模型:逻辑回归 (LR),决策树 (DT),随机森林 (RF),支向量机 (SVM) 和极端梯度增强 (XGB).
- 使用接收器操作特征曲线 (AUC) 下的面积和SHAP进行验证,以获得预测能力和可解释性.
主要成果:
- 30天的UPR率为30.54%.
- 确定了关键预测因素:体性并发症数量,疾病持续时间,最近住院时间长度,药物戒断史和性别.
- XGBoost模型实现了最高的AUC (0.830),证明了强大的预测性能.
结论:
- 极端梯度增强 (XGB) 模型有效预测精神分裂症患者的30天UPR.
- 整合SHAP提供了个性化的风险预测,有助于临床决策.
- 这种方法支持及时的出院规划和过渡期护理,可能减少再接收.
更多相关视频
12:18A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.6K
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
8.3K
相关概念视频
01:29Psychological and Sociocultural Causes of Schizophrenia
189
Schizophrenia, a complex psychiatric disorder, has been historically misunderstood. Early psychological theories attributed its origins to childhood trauma and unresponsive parenting. However, contemporary research largely rejects these notions, favoring the vulnerability-stress hypothesis. This model proposes that individuals with a genetic predisposition to schizophrenia may develop the disorder following exposure to significant environmental stressors. Notably, studies on high-risk...
189
01:52Regression Toward the Mean
6.5K
Regression toward the mean (“RTM”) is a phenomenon in which extremely high or low values—for example, and individual’s blood pressure at a particular moment—appear closer to a group’s average upon remeasuring. Although this statistical peculiarity is the result of random error and chance, it has been problematic across various medical, scientific, financial and psychological applications. In particular, RTM, if not taken into account, can interfere when...
6.5K