Scholarlink

Background <p>Parkinson’s Disease (PD) diagnosis lacks effective non-invasive markers, complicating early detection and timely intervention. Machine learning (ML) combined with clinical blood biomarkers may provide a feasible approach for early diagnosis and monitoring.</p> Aim <p>This study aims to construct and validate a non-invasive diagnostic model for PD using machine learning and routine clinical blood biomarkers, and identify key biomarkers linked to disease severity.</p> Methods <p>A total of 920 participants (428 PD and 492 non-PD) from two medical centers were included as training and validation sets. Biomarker selection was performed via least absolute shrinkage and selection operator (LASSO) and stepwise regression. Five machine learning models—logistic regression (LR), support vector machine (SVM), decision tree (DT), Naive Bayes (NB) and K-Nearest Neighbor (KNN)—were constructed and compared. The optimal model was interpreted using Shapley values (SHAP), and correlation with PD severity (Hoehn-Yahr stage) was assessed.</p> Results <p>The SVM model demonstrated the best external validation performance (AUC = 0.916, recall = 0.949, F1-score = 0.843). SHAP analysis revealed superoxide dismutase (SOD) contributed the most to the model prediction, followed by gender and uric acid (UA). Furthermore, albumin (ALB) and SOD showed significant negative correlations with PD severity.</p> Conclusion <p>The SVM-based diagnostic model effectively differentiates PD from controls using readily obtainable clinical biomarkers, offering promising clinical utility for PD screening, diagnosis, and progression monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning-based non-invasive Parkinson’s disease diagnostic model using clinical blood biomarkers

  • 期刊论文

  • 发表日期：

    2025年10月6日

查看全文
  • Jiaqi Han,
  • Mengge Sun,
  • Ji Yang,
  • Yu An

摘要

Background

Parkinson’s Disease (PD) diagnosis lacks effective non-invasive markers, complicating early detection and timely intervention. Machine learning (ML) combined with clinical blood biomarkers may provide a feasible approach for early diagnosis and monitoring.

Aim

This study aims to construct and validate a non-invasive diagnostic model for PD using machine learning and routine clinical blood biomarkers, and identify key biomarkers linked to disease severity.

Methods

A total of 920 participants (428 PD and 492 non-PD) from two medical centers were included as training and validation sets. Biomarker selection was performed via least absolute shrinkage and selection operator (LASSO) and stepwise regression. Five machine learning models—logistic regression (LR), support vector machine (SVM), decision tree (DT), Naive Bayes (NB) and K-Nearest Neighbor (KNN)—were constructed and compared. The optimal model was interpreted using Shapley values (SHAP), and correlation with PD severity (Hoehn-Yahr stage) was assessed.

Results

The SVM model demonstrated the best external validation performance (AUC = 0.916, recall = 0.949, F1-score = 0.843). SHAP analysis revealed superoxide dismutase (SOD) contributed the most to the model prediction, followed by gender and uric acid (UA). Furthermore, albumin (ALB) and SOD showed significant negative correlations with PD severity.

Conclusion

The SVM-based diagnostic model effectively differentiates PD from controls using readily obtainable clinical biomarkers, offering promising clinical utility for PD screening, diagnosis, and progression monitoring.