Arrow
use cases

Predict COPD

Use AI to identify COPD in patients using key EMR symptoms.

Problem

Chronic obstructive pulmonary disease (COPD) is a chronic inflammatory lung disease that causes obstruction of airflow from the lungs. Symptoms include shortness of breath, coughing, production of mucus (sputum), and wheezing. It is typically caused by long-term exposure to irritating gases or particulate matter, most often cigarette smoke. People with chronic obstructive pulmonary disease are at increased risk of developing heart disease, lung cancer, and several other conditions (1).

Why it matters

  • Approximately 10% of Canadians aged 35 and over are living with COPD (2).
  • Chronic obstructive pulmonary disease (COPD) is the third leading cause of death in the world. In 2019 it caused 3.23 million deaths (3).
  • Nearly 90% of COPD deaths in children under 70 years of age occur in low- and middle-income countries (3).

Air travels down the trachea and reaches the lungs through two large tubes (bronchi). Inside the lungs, these tubes divide many times, like the branches of a tree, into smaller tubes (bronchioles) that terminate in clusters of small air sacs (alveoli) (1). Your lungs rely on the natural elasticity of the bronchi and air sacs to force air out of your body. COPD causes them to lose their elasticity and over-expand, leaving some air trapped in the lungs when exhaling. As the disease worsens, it is more difficult to carry out normal daily activities, often due to shortness of breath. The disease can have considerable financial consequences due to limited productivity at work and at home and the cost of medical treatment. COPD patients often have other conditions, such as heart disease, osteoporosis, musculoskeletal disorders, lung cancer, depression, or anxiety (3).

Solution

Using structured numeric, categorical, hybrid, and unstructured text data enables predictive models to capture COPD symptoms and discriminate against diseases with similar symptoms (2).

This solution uses feature importance in medications, health conditions, risk factors, and patient age as a a set of key EMR symptoms to discover COPD.

Datasources

  • Diagnostic Imaging: Contains information about diagnostic images (for example, CT and MRI).
  • Health Services Laboratories: Captures data on hospital, community, and emergency medical services.
  • Social Determinants of Health (SDoH): Geo-centric data with details about the social and environmental influences on people’s health and outcomes.

Citations

  1. COPD - Symptoms and causes. (2020, 15 abril). Mayo Clinic.
  2. Zafari H, Langlois S, Zulkernine F, Kosowan L, Singer A. AI in predicting COPD in the Canadian population. Biosystems. 2022 Jan;211:104585. doi: 10.1016/j.biosystems.2021.104585. Epub 2021 Dec 2. PMID: 34864143.
  3. Chronic obstructive pulmonary disease (COPD). (2022, May 20). https://www.who.int/es/news-room/fact-sheets/detail/chronic-obstructive-pulmonary-disease-(copd)

Book a Free Consultation

Trusted by the world's top healthcare institutions