26 December 2022 in BMJ Open Respiratory Research
Rationale: Spirometry and plethysmography are the gold standard pulmonary function tests (PFT) for diagnosis and management of lung disease. Due to the inaccessibility of plethysmography, spirometry is often used alone but this leads to missed or misdiagnoses as spirometry cannot identify restrictive disease without plethysmography. We aimed to develop a deep learning model to improve interpretation of spirometry alone. Methods: We built a multilayer perceptron model using full PFTs from 748 patients, interpreted according to international guidelines. Inputs included spirometry (forced vital capacity, forced expiratory volume in 1 s, forced mid-expiratory flow25–75), plethysmography (total lung capacity, residual volume) and biometrics (sex, age, height). The model was developed with 2582 PFTs from 477 patients, randomly divided into training (80%), validation (10%) and test (10%) sets, and refined using 1245 previously unseen PFTs from 271 patients, split 50/50 as validation (136 patients) and test (135 patients) sets. Only one test per patient was used for each of 10 experiments conducted for each input combination. The final model was compared with interpretation of 82 spirometry tests by 6 trained pulmonologists and a decision tree. Results: Accuracies from the first 477 patients were similar when inputs included biometrics+spirometry+plethysmography (95%±3%) vs biometrics+spirometry (90%±2%). Model refinement with the next 271 patients improved accuracies with biometrics+pirometry (95%±2%) but no change for biometrics+spirometry+plethysmography (95%±2%). The final model significantly outperformed (94.67%±2.63%, p<0.01 for both) interpretation of 82 spirometry tests by the decision tree (75.61%±0.00%) and pulmonologists (66.67%±14.63%). Conclusions: Deep learning improves the diagnostic acumen of spirometry and classifies lung physiology better than pulmonologists with accuracies comparable to full PFTs.
8 December 2022 in BMJ Open Respiratory Research
Background: The Global Lung Function Initiative (GLI-2012) focused on race/ethnicity as an important factor in determining reference values. This study evaluated the effects of changing from Canadian reference equations developed from an all-Caucasian cohort with European ancestry to the GLI-2012 on the interpretation of spirometry in a multiethnic population and aimed to identify the ethnic groups affected by discrepant interpretations. Methods: Clinically indicated spirometry in a multiethnic population (aged 20–80 years) collected from 2018 to 2021 was analysed. The predicted and lower limit of normal (LLN) values were calculated using three sets of reference equations: Canadian, GLI-race/ethnic-based (GLI-Race) and GLI-race/ethnic-neutral (GLI-Other). We compared the prevalence of concordance in the abnormal diagnoses (defined as <LLN) for forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), and FEV1/FVC among the three reference values, and evaluated whether race/ethnicity was associated with discordance. Results: Data from 406 participants were evaluated (non-Caucasian 43.6%). There was 85%–87% concordance for normal/abnormal FVC and FEV1 interpretations among the Canadian, GLI-Race and GLI-Other reference equations. In all ethnic groups, application of the Canadian references for interpretation led to a higher prevalence of abnormal (<LLN) FVC and FEV1compared with GLI-Race and GLI-Other. This trend was more prominent in Black, South-East Asian and Mixed/other ethnic groups when comparing the Canadian to the GLI-Race equations. In contrast, the discordance rates were similar among ethnic groups when compared with the GLI-Other reference equations. Interpretation of FEV1/FVC had a high rate of agreement among all equations. Conclusion: Interpretation using Canadian reference equations was associated with a higher prevalence of restrictive physiology compared with the GLI-2012 equations, particularly if the GLI-Race were used. These observations were mostly found in non-white Caucasian groups, highlighting the need to choose reference equations that reflect closely the ethnic mix of the population being evaluated in order to optimise patient management.
11 May 2023 in ERJ Open Research
This study demonstrates the inadequacy of the current technical standards of oscillometry that are based on the within-trial reproducibility of the lowest-frequency Rrs, and suggests the use of a simple variability measure encompassing both Rrs and Xrs https://bit.ly/3AYRid6.
30 November 2025 in BMJ Open Respiratory Research
Background: Hypersensitivity pneumonitis (HP) is an interstitial lung disease (ILD) caused by repeated exposure to inhaled antigens, leading to small airway and parenchymal inflammation. Diagnosis is based on a detailed clinical history, chest imaging and invasive tests such as bronchoalveolar lavage. Distinguishing HP from other ILDs is challenging. Respiratory oscillometry, a novel pulmonary function test (PFT), is highly sensitive to small airway abnormalities. Oscillometry measurement of reactance is strongly correlated with gender-age-physiology score, a prognostic tool used to predict mortality and disease severity in idiopathic pulmonary fibrosis (IPF). Objective: To determine if oscillometry and standard PFT patterns are different in HP and IPF. Methods: 39 HP (79.5% with fibrotic HP) were enrolled from October 2022 to December 2023 for oscillometry before clinically-indicated standard PFTs and compared with 39 age-matched and sex-matched patients with IPF who also had same day oscillometry and standard PFTs. The main oscillometry metrics of interest were R5-19 (the difference in resistance from 5 to 19 Hz, a metric of small airway function and ventilatory inhomogeneity that increases with worsening respiratory mechanics), X5 (reactance at 5 Hz) which primarily reflects respiratory elastance and AX (area of reactance), a summative measure of the respiratory system stiffness across a range of frequencies, that behaves similarly but in opposite direction to X5. Results: Patients with HP exhibited higher residual volume/total lung capacity (RV/TLC), lower per cent predicted (%) forced expiratory volume in 1 s (FEV 1 ) and % predicted forced vital capacity (FVC) than IPF (p0.40) is a feature of HP not observed in IPF. The strong correlations of RV/TLC with AX, X5 and R5-19 suggest that oscillometry can provide non-invasive markers of small airway obstruction in HP that can differentiate it from IPF.
13 August 2025 in IEEE Access
Intra-breath oscillometry (IBOsc) offers a high-resolution, non-invasive assessment of respiratory mechanics by tracking within-breath variations in respiratory impedance. Unlike traditional oscillometric approaches, which provide averaged impedance values over the entire respiratory cycle, IBOsc captures dynamic, nonlinear changes by analyzing single-frequency excitation signals. This study introduces a novel machine learning-based framework for automated pattern recognition in IBOsc data. The proposed pipeline incorporates artifact-tolerant preprocessing, impedance loop generation, feature engineering, and classification. Using carefully curated datasets from healthy individuals and patients with chronic obstructive pulmonary disease (COPD), interstitial lung disease (ILD), and obesity hypoventilation syndrome (OHS), the proposed framework successfully identifies clinically relevant patterns such as tidal expiratory flow limitation (tEFL). In the binary classification task distinguishing tEFL in healthy versus COPD patients, the best-performing model achieved an F1-score of 0.98 and an overall accuracy of 98.5% on a held-out test set of 204 samples. In the more complex three-class scenario involving healthy, COPD, and ILD patients, the model sustained strong performance, reaching a macro-averaged accuracy of 86.3% across 156 test samples, with a class-wise accuracy of 96.2% for tEFL detection. Beyond binary classification, the method proved effective in identifying both the presence and the resolution of tEFL patterns, which is a key clinical indicator for tracking therapeutic outcomes. The methodology demonstrated robustness across varying conditions and measurement setups, highlighting the potential of automated IBOsc analysis for enhancing clinical diagnostics and phenotyping of respiratory diseases.
21 October 2026 in BMJ Open Respiratory Research
Long-term survival after lung transplant is limited by chronic lung allograft dysfunction (CLAD). Oscillometry is a novel pulmonary function technique that complements spirometry. We hypothesise that oscillometry and spirometry at 3 months post-transplant are associated with CLAD. Double lung transplant recipients (n=280) were enrolled from 2017 to 2020 for oscillometry at the first post-transplant spirometry. We excluded re-transplantation and patients missing 3-month oscillometry-spirometry data. Patients who died within 3 years where a CLAD diagnosis was ascertained were included. We used Mann-Whitney U and Pearson’s χ2 tests for group comparisons, logistic regression to assess the associations of spirometry and oscillometry with CLAD, and receiver operating characteristic curves to identify the threshold values for CLAD. At 3 years post-transplant, 129 were alive and CLAD-free while 75 patients had CLAD. Preoperative and perioperative characteristics were similar except for higher body mass index in the CLAD group. At 3 months, CLAD patients had lower forced expiratory volume in 1 s (%FEV1) and FEV1/FVC (forced vital capacity), and worst lung mechanics with high R5 (resistance at 5 Hz), low X5 (reactance at 5 Hz) and high AX (area of reactance). %FEV1, FEV1/FVC, R5 and R5−19 (difference in resistance from 5 to 19 Hz), X5, X5 z-score and AX were independently associated with CLAD. A composite metric of %FEV1, FEV1/FVC, R5 z-score, R5−19 and AX provided the strongest association with CLAD. The 3-month composite spirometry-oscillometry score post-transplant provides risk stratification of CLAD. Patients at risk can be monitored closely, to enable earlier interventions and minimise progression to CLAD.