top of page

What is COPD and How Can Artificial Intelligence Guide Its Diagnosis?

Writer's picture: Kaylyn KimKaylyn Kim

A leading cause of death worldwide, Chronic Obstructive Pulmonary Disease (COPD) is characterized by restricted airflow and breathing difficulties over time. With early intervention, COPD can be managed to slow disease progression. However, in low- and middle-income countries (LMICs), the underdiagnosis and misdiagnosis of COPD is a significant problem due to the inaccessibility of respiratory specialists and existing diagnostic methods like spirometry. This often leads to years of inadequate, costly treatment before a proper diagnosis is made. 


To address this issue, this study introduces SMART Screening, an innovative biomedical engineering solution for early COPD detection. The system leverages deep learning techniques, which is a branch of artificial intelligence that involves using filters to extract specific features or patterns from a dataset. In this case, we are analyzing cough sounds, which are unique biomarkers to identify the presence of COPD. SMART Screening employs convolutional neural networks (CNNs), which consist of various layers that each serve a specific function in the extraction process, to detect specific acoustic features from cough recordings and provide a reliable preliminary diagnosis. To enhance the model’s performance, the study implements data collected from West China Fourth Hospital and Chengdu International School, along with public databases like the Kaggle and Zenodo datasets. We trained and tested our model on around 1,800 audio recordings, and preliminary results indicate an accuracy of around 99% when tested on an unseen set of data with 188 recordings. This model is then integrated into a user-friendly mobile application to improve accessibility in low-resource settings. 


Designed to complement existing diagnostic tools, this technology has the potential to significantly improve early COPD detection in underserved populations. The next step will be to focus on how well this technology works on various demographics in the world, along with how it could be implemented for other respiratory diseases.


Bibliography

Alqudaihi, K., Aslam, N., Khan, I., Almuhaideb, A., Alsunaidi, S., Ibrahim, N., Alhaidari, F., Shaikh, F., Alsenbel, Y., Alalharith, D., Alharthi, H., Alghamdi, W., & Alshahrani, M. (2021). Cough sound detection and diagnosis using artificial intelligence techniques: Challenges and opportunities. IEEE Access, 9, 102327–102344. https://doi.org/10.1109/ACCESS.2021.3097559


Boers, E., Barrett, M., Su, J. G., Benjafield, A. V., Sinha, S., Kaye, L., Zar, H. J., Vuong, V., Tellez, D., Gondalia, R., Rice, M. B., Nunez, C. M., Wedzicha, J. A., & Malhotra, A. (2023). Global Burden of Chronic Obstructive Pulmonary Disease through 2050. JAMA Network Open, 6(12), e2346598. https://doi.org/10.1001/jamanetworkopen.2023.46598


Cleveland Clinic. (2022, May 17). What Is COPD? Symptoms, Treatment & More | Cleveland Clinic: Health Library. Cleveland Clinic. https://my.clevelandclinic.org/health/diseases/8709-chronic-obstructive-pulmonary-disease-copd


Elgendi, M., Menon, C., & Ghrabli, S. (2024). Identifying unique spectral fingerprints in cough sounds for diagnosing respiratory ailments. Scientific Reports, 14(1). https://doi.org/10.1038/s41598-023-50371-2


Kuluozturk, M., Kobat, M. A., Barua, P. D., Dogan, S., Tuncer, T., Tan, R.-S., Ciaccio, E. J., & Acharya, U. R. (2022). DKPNet41: Directed knight pattern network-based cough sound classification model for automatic disease diagnosis. Medical Engineering & Physics, 110, 103870. https://doi.org/10.1016/j.medengphy.2022.103870


Liu, Y., Carlson, S., Watson, K., Xu, F., & Greenlund, K. (2023). Morbidity and Mortality Weekly Report. In Centers for Disease Control and Prevention. https://www.cdc.gov/mmwr/volumes/72/wr/pdfs/mm7246-H.pdf


National Health Service. (2021, August 18). Spirometry. NHS. https://www.nhs.uk/conditions/spirometry/


Panagiotis Kapetanidis, Fotios Kalioras, Tsakonas, C., Pantelis Tzamalis, Kontogiannis, G., Karamanidou, T., Stavropoulos, T. G., & Sotiris Nikoletseas. (2024). Respiratory diseases diagnosis using audio analysis and artificial intelligence: A systematic review. Sensors, 24(4), 1173–1173. https://doi.org/10.3390/s24041173


Patel, K., Smith, D. J., Huntley, C. C., Channa, S. D., Pye, A., Dickens, A. P., Gale, N., & Turner, A. M. (2024). Exploring the Causes of COPD Misdiagnosis in Primary care: a Mixed Methods Study. PLOS ONE, 19(3), e0298432–e0298432. https://doi.org/10.1371/journal.pone.0298432


Plum, C., Stolbrink, M., Zurba, L., Bissell, K., Ozoh, B. O., & Mortimer, K. (2021). Availability of diagnostic services and essential medicines for non‐communicable respiratory diseases in African countries. The International Journal of Tuberculosis and Lung Disease, 25(2), 120–125. https://doi.org/10.5588/ijtld.20.0762


Smith, J., Ashurst, H., Jack, S., Woodcock, A., & Earis, J. (2006). The description of cough sounds by healthcare professionals. Cough, 2(1), 1. https://doi.org/10.1186/1745-9974-2-1


Spencer, P., & Krieger, B. (2013). The Differentiation of Chronic Obstructive Pulmonary Disease from Asthma: a Review of Current Diagnostic and Treatment Recommendations. The Open Nursing Journal, 7, 29–34. https://doi.org/10.2174/1874434601307010029


World Health Organization. (2024, November 6). Chronic Obstructive Pulmonary Disease (COPD). World Health Organisation. https://www.who.int/news-room/fact-sheets/detail/chronic-obstructive-pulmonary-disease-(copd)


Yao, L.-P., & Tao, R. (2021). Does Chronic Obstructive Pulmonary Disease Affect Workers’ Health? Frontiers in Public Health, 9. https://doi.org/10.3389/fpubh.2021.711629

1 view0 comments

Recent Posts

See All

Comments


"Once you learn how to die, you learn how to live" 
- Morrie Schwartz (from Tuesdays with Morrie)

bottom of page