Objective . Diagnosing chronic obstructive pulmonary disease (COPD) using impulse oscillometry (IOS) is challenging due to the high level of clinical expertise it demands from doctors, which limits the clinical application of IOS in screening. The primary aim of this study is to develop a COPD diagnostic model based on machine learning algorithms using IOS test results. Approach . Feature selection was conducted to identify the optimal subset of features from the original feature set, which significantly enhanced the classifier’s performance. Additionally, secondary features area of reactance (AX) were derived from the original features based on clinical theory, further enhancing the performance of the classifier. The performance of the model was analyzed and validated using various classifiers and hyperparameter settings to identify the optimal classifier. We collected 528 clinical data examples from the China–Japan Friendship Hospital for training and validating the model. Main results . The proposed model achieved reasonably accurate diagnostic results in the clinical data (accuracy = 0.920, specificity = 0.941, precision = 0.875, recall = 0.875). Significance . The results of this study demonstrate that the proposed classifier model, feature selection method, and derived secondary feature AX provide significant auxiliary support in reducing the requirement for clinical experience in COPD diagnosis using IOS.

Small airway dysfunction (SAD) is an early functional abnormality associated with multiple chronic airway diseases. However, clinical assessment often relies on spirometry-based indices, which require forced maneuvers and are sensitive to subject effort, thereby increasing patient burden and complicating quality control. In contrast, Impulse Oscillometry (IOS) requires only tidal breathing, imposing minimal subject burden while providing respiratory impedance indices informative for SAD identification. This study proposes a dual-domain complementary deep learning framework based on IOS for SAD identification, leveraging within-breath impedance dynamics. Specifically, raw IOS time-series signals are transformed into time–frequency respiratory impedance maps (TFRIM) capturing impedance over frequency and within-breath time. A two-stream architecture is then used to jointly learn complementary features from TFRIM and the original time-series signals. To mitigate inter-subject baseline variability, we further introduce a demographics-driven adaptive feature modulation module for subject-specific calibration. The model jointly predicts multiple small-airway indices, with decision-level fusion applied during inference. Experimental validation on 2510 subjects using five-fold cross-validation demonstrates that the proposed framework achieves an accuracy of 81.39%, outperforming representative baselines. These results suggest the potential utility of combining within-breath IOS dynamics with subject-specific calibration for SAD identification, warranting further external validation before screening deployment.