1
|
Xu WJ, Shang WY, Feng JM, Song XY, Li LY, Xie XP, Wang YM, Liang BM. Machine learning for accurate detection of small airway dysfunction-related respiratory changes: an observational study. Respir Res 2024; 25:286. [PMID: 39048993 PMCID: PMC11270925 DOI: 10.1186/s12931-024-02911-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/06/2024] [Accepted: 07/08/2024] [Indexed: 07/27/2024] Open
Abstract
BACKGROUND The use of machine learning(ML) methods would improve the diagnosis of small airway dysfunction(SAD) in subjects with chronic respiratory symptoms and preserved pulmonary function(PPF). This paper evaluated the performance of several ML algorithms associated with the impulse oscillometry(IOS) analysis to aid in the diagnostic of respiratory changes in SAD. We also find out the best configuration for this task. METHODS IOS and spirometry were measured in 280 subjects, including a healthy control group (n = 78), a group with normal spirometry (n = 158) and a group with abnormal spirometry (n = 44). Various supervised machine learning (ML) algorithms and feature selection strategies were examined, such as Support Vector Machines (SVM), Random Forests (RF), Adaptive Boosting (ADABOOST), Navie Bayesian (BAYES), and K-Nearest Neighbors (KNN). RESULTS The first experiment of this study demonstrated that the best oscillometric parameter (BOP) was R5, with an AUC value of 0.642, when comparing a healthy control group(CG) with patients in the group without lung volume-defined SAD(PPFN). The AUC value of BOP in the control group was 0.769 compared with patients with spirometry defined SAD(PPFA) in the PPF population. In the second experiment, the ML technique was used. In CGvsPPFN, RF and ADABOOST had the best diagnostic results (AUC = 0.914, 0.915), with significantly higher accuracy compared to BOP (p < 0.01). In CGvsPPFA, RF and ADABOOST had the best diagnostic results (AUC = 0.951, 0.971) and significantly higher diagnostic accuracy (p < 0.01). In the third, fourth and fifth experiments, different feature selection techniques allowed us to find the best IOS parameters (R5, (R5-R20)/R5 and Fres). The results demonstrate that the performance of ADABOOST remained essentially unaltered following the application of the feature selector, whereas the diagnostic accuracy of the remaining four classifiers (RF, SVM, BAYES, and KNN) is marginally enhanced. CONCLUSIONS IOS combined with ML algorithms provide a new method for diagnosing SAD in subjects with chronic respiratory symptoms and PPF. The present study's findings provide evidence that this combination may help in the early diagnosis of respiratory changes in these patients.
Collapse
Affiliation(s)
- Wen-Jing Xu
- Department of Respiratory and Critical Care Medicine, West China School of Medicine and West China Hospital, Sichuan University, No. 37 Guoxue Alley, Chengdu, 610041, China
| | - Wen-Yi Shang
- Machine Intelligence Laboratory, College of Computer Science, Sichuan University, Chengdu, 610065, China
| | - Jia-Ming Feng
- West China School of Medicine, West China Hospital, Sichuan University, Chengdu, Sichuan, P. R. China
| | - Xin-Yue Song
- Department of Respiratory and Critical Care Medicine, West China School of Medicine and West China Hospital, Sichuan University, No. 37 Guoxue Alley, Chengdu, 610041, China
| | - Liang-Yuan Li
- Department of Respiratory and Critical Care Medicine, West China School of Medicine and West China Hospital, Sichuan University, No. 37 Guoxue Alley, Chengdu, 610041, China
| | - Xin-Peng Xie
- College of Electrical Engineering and Automation, Sichuan University, Chengdu, 610065, China
| | - Yan-Mei Wang
- Institute of Traditional Chinese Medicine of Sichuan Academy of Chinese Medicine Sciences(Sichuan Second Hospital of T.C.M), Chengdu, 610000, China
| | - Bin-Miao Liang
- Department of Respiratory and Critical Care Medicine, West China School of Medicine and West China Hospital, Sichuan University, No. 37 Guoxue Alley, Chengdu, 610041, China.
| |
Collapse
|
2
|
Huang YCT, Henriquez L, Chen H, Henriquez C. Development and evaluation of a computerized algorithm for the interpretation of pulmonary function tests. PLoS One 2024; 19:e0297519. [PMID: 38285673 PMCID: PMC10824436 DOI: 10.1371/journal.pone.0297519] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/19/2023] [Accepted: 01/07/2024] [Indexed: 01/31/2024] Open
Abstract
Pulmonary function tests (PFTs) are usually interpreted by clinicians using rule-based strategies and pattern recognition. The interpretation, however, has variabilities due to patient and interpreter errors. Most PFTs have recognizable patterns that can be categorized into specific physiological defects. In this study, we developed a computerized algorithm using the python package (pdfplumber) and validated against clinicians' interpretation. We downloaded PFT reports in the electronic medical record system that were in PDF format. We digitized the flow volume loop (FVL) and extracted numeric values from the reports. The algorithm used FEV1/FVC<0.7 for obstruction, TLC<80%pred for restriction and <80% or >120%pred for abnormal DLCO. The algorithm also used a small airway disease index (SADI) to quantify late expiratory flattening of the FVL to assess small airway dysfunction. We devised keywords for the python Natural Language Processing (NLP) package (spaCy) to identify obstruction, restriction, abnormal DLCO and small airway dysfunction in the reports. The algorithm was compared to clinicians' interpretation in 6,889 PFTs done between March 1st, 2018, and September 30th, 2020. The agreement rates (Cohen's kappa) for obstruction, restriction and abnormal DLCO were 94.4% (0.868), 99.0% (0.979) and 87.9% (0.750) respectively. In 4,711 PFTs with FEV1/FVC≥0.7, the algorithm identified 190 tests with SADI < lower limit of normal (LLN), suggesting small airway dysfunction. Of these, the clinicians (67.9%) also flagged 129 tests. When SADI was ≥ LLN, no clinician's reports indicated small airway dysfunction. Our results showed the computerized algorithm agreed with clinicians' interpretation in approximately 90% of the tests and provided a sensitive objective measure for assessing small airway dysfunction. The algorithm can improve efficiency and consistency and decrease human errors in PFT interpretation. The computerized algorithm works directly on PFT reports in PDF format and can be adapted to incorporate a different interpretation strategy and platform.
Collapse
Affiliation(s)
- Yuh-Chin T. Huang
- Department of Medicine, Duke University Medical Center, Durham, NC, United States of America
| | - Luke Henriquez
- Department of Cognitive Science, Case Western University, Cleveland, OH, United States of America
| | - Hengji Chen
- Department of Biomedical Engineering, Pratt School of Engineering, Duke University, Durham, NC, United States of America
| | - Craig Henriquez
- Department of Biomedical Engineering, Pratt School of Engineering, Duke University, Durham, NC, United States of America
| |
Collapse
|
3
|
Waląg D, Soliński M, Kołtowski Ł, Górska K, Korczyński P, Kuźnar-Kamińska B, Grabicki M, Basza M, Łepek M. Deep learning algorithm for visual quality assessment of the spirograms. Physiol Meas 2023; 44:085004. [PMID: 37552997 DOI: 10.1088/1361-6579/acee41] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/27/2022] [Accepted: 08/08/2023] [Indexed: 08/10/2023]
Abstract
Objective. The quality of spirometry manoeuvres is crucial for correctly interpreting the values of spirometry parameters. A fundamental guideline for proper quality assessment is the American Thoracic Society and European Respiratory Society (ATS/ERS) Standards for spirometry, updated in 2019, which describe several start-of-test and end-of-test criteria which can be assessed automatically. However, the spirometry standards also require a visual evaluation of the spirometry curve to determine the spirograms' acceptability or usability. In this study, we present an automatic algorithm based on a convolutional neural network (CNN) for quality assessment of the spirometry curves as an alternative to manual verification performed by specialists.Approach. The algorithm for automatic assessment of spirometry measurements was created using a set of randomly selected 1998 spirograms which met all quantitative criteria defined by ATS/ERS Standards. Each spirogram was annotated as 'confirm' (remaining acceptable or usable status) or 'reject' (change the status to unacceptable) by four pulmonologists, separately for FEV1 and FVC parameters. The database was split into a training (80%) and test set (20%) for developing the CNN classification algorithm. The algorithm was optimised using a cross-validation method.Main results. The accuracy, sensitivity and specificity obtained for the algorithm were 92.6%, 93.1% and 90.0% for FEV1 and 94.1%, 95.6% and 88.3% for FVC, respectively.Significance.The algorithm provides an opportunity to significantly improve the quality of spirometry tests, especially during unsupervised spirometry. It can also serve as an additional tool in clinical trials to quickly assess the quality of a large group of tests.
Collapse
Affiliation(s)
- Damian Waląg
- Faculty of Physics, Warsaw University of Technology, Koszykowa St. 75, 00-662, Warsaw, Poland
| | - Mateusz Soliński
- School of Biomedical Engineering & Imaging Sciences, Faculty of Life Sciences & Medicine, King's College London, Strand, London WC2R 2LS, United Kingdom
| | - Łukasz Kołtowski
- 1st Department of Cardiology, Medical University of Warsaw, Stefana Banacha St. 1a, 02-097, Warsaw, Poland
| | - Katarzyna Górska
- Department of Internal Medicine, Pulmonary Diseases and Allergy, Medical University of Warsaw, Stefana Banacha St. 1A, 02-097, Warsaw, Poland
| | - Piotr Korczyński
- Department of Internal Medicine, Pulmonary Diseases and Allergy, Medical University of Warsaw, Stefana Banacha St. 1A, 02-097, Warsaw, Poland
| | - Barbara Kuźnar-Kamińska
- Department of Pulmonology, Allergology and Respiratory Oncology, Poznan University of Medical Sciences, Szamarzewskiego St. 82, 61-001, Poznan, Poland
| | - Marcin Grabicki
- Department of Pulmonology, Allergology and Respiratory Oncology, Poznan University of Medical Sciences, Szamarzewskiego St. 82, 61-001, Poznan, Poland
| | - Mikołaj Basza
- Medical University of Silesia, Poniatowskiego St. 15, 40-055, Katowice, Poland
| | - Michał Łepek
- Faculty of Physics, Warsaw University of Technology, Koszykowa St. 75, 00-662, Warsaw, Poland
| |
Collapse
|
4
|
Hill D, Torop M, Masoomi A, Castaldi PJ, Silverman EK, Bodduluri S, Bhatt SP, Yun T, McLean CY, Hormozdiari F, Dy J, Cho MH, Hobbs BD. Deep Learning Utilizing Suboptimal Spirometry Data to Improve Lung Function and Mortality Prediction in the UK Biobank. MEDRXIV : THE PREPRINT SERVER FOR HEALTH SCIENCES 2023:2023.04.28.23289178. [PMID: 37162978 PMCID: PMC10168495 DOI: 10.1101/2023.04.28.23289178] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/11/2023]
Abstract
Background Spirometry measures lung function by selecting the best of multiple efforts meeting pre-specified quality control (QC), and reporting two key metrics: forced expiratory volume in 1 second (FEV1) and forced vital capacity (FVC). We hypothesize that discarded submaximal and QC-failing data meaningfully contribute to the prediction of airflow obstruction and all-cause mortality. Methods We evaluated volume-time spirometry data from the UK Biobank. We identified "best" spirometry efforts as those passing QC with the maximum FVC. "Discarded" efforts were either submaximal or failed QC. To create a combined representation of lung function we implemented a contrastive learning approach, Spirogram-based Contrastive Learning Framework (Spiro-CLF), which utilized all recorded volume-time curves per participant and applied different transformations (e.g. flow-volume, flow-time). In a held-out 20% testing subset we applied the Spiro-CLF representation of a participant's overall lung function to 1) binary predictions of FEV1/FVC < 0.7 and FEV1 Percent Predicted (FEV1PP) < 80%, indicative of airflow obstruction, and 2) Cox regression for all-cause mortality. Findings We included 940,705 volume-time curves from 352,684 UK Biobank participants with 2-3 spirometry efforts per individual (66.7% with 3 efforts) and at least one QC-passing spirometry effort. Of all spirometry efforts, 24.1% failed QC and 37.5% were submaximal. Spiro-CLF prediction of FEV1/FVC < 0.7 utilizing discarded spirometry efforts had an Area under the Receiver Operating Characteristics (AUROC) of 0.981 (0.863 for FEV1PP prediction). Incorporating discarded spirometry efforts in all-cause mortality prediction was associated with a concordance index (c-index) of 0.654, which exceeded the c-indices from FEV1 (0.590), FVC (0.559), or FEV1/FVC (0.599) from each participant's single best effort. Interpretation A contrastive learning model using raw spirometry curves can accurately predict lung function using submaximal and QC-failing efforts. This model also has superior prediction of all-cause mortality compared to standard lung function measurements. Funding MHC is supported by NIH R01HL137927, R01HL135142, HL147148, and HL089856.BDH is supported by NIH K08HL136928, U01 HL089856, and an Alpha-1 Foundation Research Grant.DH is supported by NIH 2T32HL007427-41EKS is supported by NIH R01 HL152728, R01 HL147148, U01 HL089856, R01 HL133135, P01 HL132825, and P01 HL114501.PJC is supported by NIH R01HL124233 and R01HL147326.SPB is supported by NIH R01HL151421 and UH3HL155806.TY, FH, and CYM are employees of Google LLC.
Collapse
Affiliation(s)
- Davin Hill
- Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, USA
- Channing Division of Network Medicine, Brigham and Women’s Hospital, Boston, MA, USA
| | - Max Torop
- Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, USA
| | - Aria Masoomi
- Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, USA
| | - Peter J. Castaldi
- Channing Division of Network Medicine, Brigham and Women’s Hospital, Boston, MA, USA
- Division of General Medicine and Primary Care, Brigham and Women’s Hospital, Boston, MA, USA
- Harvard Medical School, Boston, MA, USA
| | - Edwin K. Silverman
- Channing Division of Network Medicine, Brigham and Women’s Hospital, Boston, MA, USA
- Harvard Medical School, Boston, MA, USA
- Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital, Boston, MA, USA
| | - Sandeep Bodduluri
- Division of Pulmonary, Allergy and Critical Care Medicine, University of Alabama at Birmingham, Birmingham, AL, USA
| | - Surya P. Bhatt
- Division of Pulmonary, Allergy and Critical Care Medicine, University of Alabama at Birmingham, Birmingham, AL, USA
| | | | | | | | - Jennifer Dy
- Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, USA
| | - Michael H. Cho
- Channing Division of Network Medicine, Brigham and Women’s Hospital, Boston, MA, USA
- Harvard Medical School, Boston, MA, USA
- Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital, Boston, MA, USA
| | - Brian D. Hobbs
- Channing Division of Network Medicine, Brigham and Women’s Hospital, Boston, MA, USA
- Harvard Medical School, Boston, MA, USA
- Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital, Boston, MA, USA
| |
Collapse
|
5
|
Huang H, Huang X, Liao J, Li Y, Su Y, Meng Y, Zhan Y. Sex-specific non-linear associations between body mass index and impaired pulmonary ventilation function in a community-based population: Longgang COPD study. Front Pharmacol 2023; 14:1103573. [PMID: 36969844 PMCID: PMC10034327 DOI: 10.3389/fphar.2023.1103573] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/20/2022] [Accepted: 03/01/2023] [Indexed: 03/11/2023] Open
Abstract
Aim: To investigate the prevalence of pulmonary airflow limitation and its association with body mass index (BMI) in a community-based population in Shenzhen, China.Methods: Study participants were recruited from Nanlian Community in Shenzhen, China, and spirometry was performed to assess lung function including forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), FEV1/FVC ratio, and FEV1 divided by predicted value. Pulmonary airflow limitation was determined by the Chinese Guideline of Pulmonary Function Examination. Multivariable logistic regression models were used to examine the association between BMI and pulmonary airflow limitation. Age, sex, educational attainment, occupation, and current cigarette smoking were used as potential confounders.Results: Of the 1206 participants, 612 (50.7%) were men and 594 (49.3%) were women with the average age being 53.7 years old. After adjusting for age, sex, educational attainment, occupation, and current cigarette smoking, higher BMI was associated with lower odds (odds ratio: 0.98, 95% confidence interval: 0.97, 0.99) of pulmonary airflow limitation by assuming a linear relationship. Further investigation of the interaction terms, we found that the magnitudes of the associations differed in men and women. A U-shaped relationship was observed in women, while the association was almost linear in men.Conclusion: The relationship between BMI and pulmonary airflow limitation was U-shaped in women and linear in men.
Collapse
Affiliation(s)
- Hao Huang
- Department of Epidemiology, School of Public Health (Shenzhen), Sun Yat-Sen University, Shenzhen, China
- Nanlian Community Health Service Center, Shenzhen Longgang Central Hospital, Shenzhen, China
| | - Xueliang Huang
- Nanlian Community Health Service Center, Shenzhen Longgang Central Hospital, Shenzhen, China
| | - Jiaman Liao
- Nanlian Community Health Service Center, Shenzhen Longgang Central Hospital, Shenzhen, China
| | - Yushao Li
- Nanlian Community Health Service Center, Shenzhen Longgang Central Hospital, Shenzhen, China
| | - Yaoting Su
- Nanlian Community Health Service Center, Shenzhen Longgang Central Hospital, Shenzhen, China
| | - Yaxian Meng
- Department of Epidemiology, School of Public Health (Shenzhen), Sun Yat-Sen University, Shenzhen, China
| | - Yiqiang Zhan
- Department of Epidemiology, School of Public Health (Shenzhen), Sun Yat-Sen University, Shenzhen, China
- *Correspondence: Yiqiang Zhan,
| |
Collapse
|