Scholarlink

Background <p>Infectious mononucleosis (IM) is an acute self-limited disease caused mainly by Epstein-Barr virus (EBV) that is prone to be missed or misdiagnosed in febrile children due to a lack of obvious clinical symptoms. Establishing a model based on blood cell parameters for identifying patients with IM is important for the subsequent diagnosis and treatment of children with fever.</p> Methods <p>From January to December 2023, data of a total of 5511 febrile children (including 518 IM patients with primary EBV infection) were collected from Children's Hospital Zhejiang University School of Medicine. After assessing the changes in lymphoid clusters on the 3D-DIFF scatter plot and further analyzing the parameters of the scattergrams for quantifying the fluorescence signal changes in IM patients through the probability density function, eleven features were selected from among 41 blood cell parameters to construct the final Novel IM screening model (Novel IMs model).</p> Results <p>The Novel IMs model performed well in three sets, screening IM patients with an accuracy of more than 91.92%. Compared others, the Novel IMs model had the best identification accuracy, specificity, and sensitivity and reached an area under the receiver operating characteristic (ROC) curve (AUC) in identifying IM patients of 0.96.</p> Conclusion <p>This study is the first to develop a machine learning model to screening IM patients from febrile children using parameters obtained from blood cell scattergrams. This convenient and low-cost technique for initial screening of IM by hematology analysis may improve the screening coverage of the disease and further reduce misdiagnosis and underdiagnosis.</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

Novel screening model for infectious mononucleosis in febrile pediatric patients using a 3D-DIFF scattergram

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月7日

查看全文
  • Zhaoyang Peng,
  • Jingxian Zhang,
  • Lin Chen,
  • Kesheng Li,
  • Fengqing Cai,
  • Wenbo Zheng,
  • Lisu Huang,
  • Baohai Chen,
  • Jing Zou,
  • Chunda Wang,
  • Xiaoxia Gao,
  • Bin Huang,
  • Hongqiang Shen,
  • Gang Yu

摘要

Background

Infectious mononucleosis (IM) is an acute self-limited disease caused mainly by Epstein-Barr virus (EBV) that is prone to be missed or misdiagnosed in febrile children due to a lack of obvious clinical symptoms. Establishing a model based on blood cell parameters for identifying patients with IM is important for the subsequent diagnosis and treatment of children with fever.

Methods

From January to December 2023, data of a total of 5511 febrile children (including 518 IM patients with primary EBV infection) were collected from Children's Hospital Zhejiang University School of Medicine. After assessing the changes in lymphoid clusters on the 3D-DIFF scatter plot and further analyzing the parameters of the scattergrams for quantifying the fluorescence signal changes in IM patients through the probability density function, eleven features were selected from among 41 blood cell parameters to construct the final Novel IM screening model (Novel IMs model).

Results

The Novel IMs model performed well in three sets, screening IM patients with an accuracy of more than 91.92%. Compared others, the Novel IMs model had the best identification accuracy, specificity, and sensitivity and reached an area under the receiver operating characteristic (ROC) curve (AUC) in identifying IM patients of 0.96.

Conclusion

This study is the first to develop a machine learning model to screening IM patients from febrile children using parameters obtained from blood cell scattergrams. This convenient and low-cost technique for initial screening of IM by hematology analysis may improve the screening coverage of the disease and further reduce misdiagnosis and underdiagnosis.