Scholarlink

<p>Alpha-thalassemia is a widespread genetic disorder, and accurately distinguishing between alpha-plus (α⁺) and alpha-zero (α⁰) types is critical for effective screening and management. This study developed and evaluated machine learning models to classify α⁺ and α⁰ carriers based on hematological parameters. A dataset of 956 cases was analyzed, including variables such as red blood cell (RBC) count, hemoglobin (Hb) level, and RBC indices. Feature selection identified the most predictive markers, and five machine learning models were trained and compared. The stacking ensemble model demonstrated the best performance, achieving 94% accuracy and a high F1-score. Key predictors included RBC count, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). Correlation analysis revealed strong interrelationships among RBC indices, while platelet (PLT) and white blood cell (WBC) parameters had moderate associations. These findings suggest that machine learning, particularly ensemble methods, can enhance the detection of alpha-thalassemia carriers. The development of models based on both data-driven and clinical features provides a flexible framework for screening and could support more personalized approaches in future research.</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 improves detection of alpha thalassemia carriers compared to clinical features

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月21日

查看全文
  • Elmira Mohammadi,
  • Mohsen Rastegar,
  • Amir Jamshidnezhad,
  • Amirabbas Azizi

摘要

Alpha-thalassemia is a widespread genetic disorder, and accurately distinguishing between alpha-plus (α⁺) and alpha-zero (α⁰) types is critical for effective screening and management. This study developed and evaluated machine learning models to classify α⁺ and α⁰ carriers based on hematological parameters. A dataset of 956 cases was analyzed, including variables such as red blood cell (RBC) count, hemoglobin (Hb) level, and RBC indices. Feature selection identified the most predictive markers, and five machine learning models were trained and compared. The stacking ensemble model demonstrated the best performance, achieving 94% accuracy and a high F1-score. Key predictors included RBC count, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin concentration (MCHC). Correlation analysis revealed strong interrelationships among RBC indices, while platelet (PLT) and white blood cell (WBC) parameters had moderate associations. These findings suggest that machine learning, particularly ensemble methods, can enhance the detection of alpha-thalassemia carriers. The development of models based on both data-driven and clinical features provides a flexible framework for screening and could support more personalized approaches in future research.