|
|
|
| Construction of a Machine Learning-Based Diagnostic Model Using Routine Blood Parameters for Identification of Mycoplasma Pneumoniae Infection |
| CHEN Yue, XIAO Chunhai |
| Department of Clinical Laboratory, Jinshan Branch of Shanghai Sixth People's Hospital, Shanghai 201500 |
|
|
|
|
Abstract 【Objective】 To explore the application of routine blood parameters in establishing machine learning screening models to construct and validate a nomogram diagnostic model for mycoplasma pneumoniae (MP) infection based on hematological indicators. 【Methods】 Clinical data from 4823 patients who underwent MP antigen testing and routine blood examinations at our hospital between January 2024 and June 2025 were retrospectively analyzed. Among them, 2545 MP-positive patients were assigned to the infection group, while 2278 MP-negative patients served as the control group. Four machine learning models, including Logistic regression, random forest, decision tree, and neural network, were established. Receiver operating characteristic (ROC) curves were used to compare the area under the curve (AUC) values of the four models. 【Results】 The AUC values of the Logistic regression, random forest, decision tree, and neural network models were 0.952, 0.942, 0.860, and 0.819, respectively. Multivariate analysis of the optimal Logistic regression model showed that variables with statistical significance (P<0.05) included age, red blood cell (RBC) count, red blood cell distribution width coefficient of variation (RDW-CV), mean platelet volume (MPV), mean corpuscular hemoglobin concentration (MCHC), and platelet (PLT) count. Based on the multivariate analysis results, a nomogram diagnostic model for MP infection using routine blood parameters was established along with clinical decision and clinical impact curves. 【Conclusion】 The Logistic regression and nomogram models that were constructed based on age, RBC count, RDW-CV, MPV, MCHC, and PLT count can effectively assist in the diagnosis of MP infection and demonstrate important clinical application value.
|
|
Received: 24 October 2025
|
|
|
|
|
|
[1] 吕智慧,张海涛,马常建,等.不同类型儿童肺炎支原体肺炎的临床表现和实验室检查特点[J].医学综述, 2023, 29(24):5865-5870.
[2] 张永威,刘玉华.利用列线图区分0～12岁社区获得性肺炎患儿肺炎支原体和病毒感染的探索性研究[J].中华妇幼临床医学杂志(电子版), 2025, 21(2):180-188.
[3] 李丰帆,吕星,张嘉梁,等.2023年湖南省儿童支原体肺炎耐药突变及治疗情况分析[J].中华预防医学杂志, 2024, 58(12):2038-2044.
[4] CHOI Y J, CHUNG E H, LEE E, et al. Clinical characteristics of macrolide-refractory mycoplasma pneumoniae pneumonia in Korean children: a multicenter retrospective study[J].Clin Med,2022,11(2):306-321.
[5] ZHOU Y J, WANG J, CHEN W J, et al. Impact of viral coinfection and macrolide-resistant mycoplasma infection in children with refractory Mycoplasma pneumoniae pneumonia[J].BMC Infect Dis,2020,20(1):633.
[6] 闫士忍,刘宏超,周剑锁,等.2017—2024年北京市某医院肺炎支原体抗体检测数据分析[J].中华预防医学杂志, 2025, 59(1):62-68.
[7] 孟艺.4063例0～18周岁体检健康人群血细胞参考值范围分析[J].浙江实用医学, 2012, 17(5):369-370.
[8] SAÏD M H, LAYANI M P, COLON S,et al. Mycoplasma pneumoniae-associated nephritis in children[J].Pediatr Nephrol,1999,13(1):39-44.
[9] 李姗姗.儿童重症肺炎支原体肺炎合并冷凝集素病1例[J].中国临床案例成果数据库, 2025, 7(1):E0951.
[10] 贾芳,周学章.血小板与免疫细胞、炎症反应和细菌相互作用的概况[J].现代免疫学,2020,40(3):252-256.
[11] GAERTNER F,MASSBERG S.Patrolling the vascular borders:platelets in immunity to infection and cancer[J].Nat Rev Immunol,2019,19(12):747-760.
[12] 刘燕玲,阚璇.肺炎支原体肺炎患儿血小板参数变化及意义[J].山东医药,2023,63(16):80-84. |
|
|
|
2026, Vol. 43 
