APSR2026-565

Yes
Late-Breaking
Submitted
Oral Presentation
AO03-005
 
 
https://storage.unitedwebnetwork.com/files/1285/1220879-46903-APSR2026-565_439770.pptx
Format: PPTX Only
 
Format: PDF Only
07. COPD
6

No more than 15 authors can be listed.
Please ensure the authors are listed in the right order.
If there are multiple affiliations, please use the "+" button next to "Select Country/Region" to add additional affiliation fields.

junjie Xia * The Third Hospital of Mianyang Respiratory and Critical Care Medicine Department Mianyang China
yue lan - The Third Hospital of Mianyang Respiratory and Critical Care Medicine Department Mianyang China
xin huang - The Third Hospital of Mianyang Respiratory and Critical Care Medicine Department Mianyang China
ling Huang - The Third Hospital of Mianyang Respiratory and Critical Care Medicine Department Mianyang China
yu Qiu - The Third Hospital of Mianyang Respiratory and Critical Care Medicine Department Mianyang China
Wenjun Li - The Third Hospital of Mianyang Respiratory and Critical Care Medicine Department Mianyang China
 
 
 
 
 
 
 
 
 
Presenting Author Details
junjie
Xia
xiajun1922@163.com
 
31-40
China
Original Research
Hematological Biomarkers and Machine Learning for Early Identification of Pre-COPD in High-Altitude Populations
Maximum 25 words
Abstract Content
Pre-chronic obstructive pulmonary disease (pre-COPD) represents a critical stage for early intervention before the development of persistent airflow limitation. However, its identification largely relies on pulmonary function testing, which remains limited in accessibility, particularly in high-altitude regions. Hematological biomarkers may provide a practical alternative for early detection.
We conducted a cross-sectional study in Hongyuan County, a high-altitude region of Sichuan Province, China, including 1,256 adults (211 pre-COPD and 1,044 healthy controls). Pre-COPD was defined as post-bronchodilator FEV₁/FVC ≥0.70 with FEV₁ <80% predicted and/or small airway dysfunction. A total of 48 demographic and laboratory variables were collected. An ensemble feature selection strategy combining Multisurf and mutual information was applied to identify key predictors. Seven machine learning models, including logistic regression, decision tree, random forest, support vector machine, extreme gradient boosting, light gradient boosting machine, and artificial neural network (ANN), were developed and evaluated using 10-fold cross-validation. Model performance was assessed using multiple metrics, including AUC, accuracy, precision, recall, and false-positive rate.
Ten variables were selected as optimal predictors, including red blood cell count, total protein, neutrophil count, direct bilirubin, albumin, hemoglobin, age group, total cholesterol, lymphocyte-to-platelet ratio, and cystatin C. Among all models, the ANN model demonstrated the best overall performance, with superior accuracy and the lowest false-positive rate compared to other models (P < 0.05). Several biomarkers showed biological relevance to pre-COPD, reflecting hypoxia adaptation, systemic inflammation, oxidative stress, and metabolic alterations. Notably, the lymphocyte-to-platelet ratio may indicate low-grade systemic inflammation associated with early airflow limitation.
A machine learning model based on routine hematological biomarkers can effectively identify individuals at high risk of pre-COPD in high-altitude populations. This approach provides a feasible and accessible screening tool in resource-limited settings and may facilitate early intervention strategies to prevent COPD progression.
 
 
 
 
 
No
No
 
JPG or JPEG or PNG format ONLY
 
maximum 3 keywords and no abbreviations, separated by commas
 
294