Training a mathematical model to estimate lung weight and volume using anthropometric and sociodemographic features
Principal Investigator
Name
Lygia Costa
Degrees
M.Sc.
Institution
Independent Researcher
Position Title
Medical Student
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-753
Initial CDAS Request Approval
Jan 27, 2021
Title
Training a mathematical model to estimate lung weight and volume using anthropometric and sociodemographic features
Summary
Our main goal is to elaborate a mathematical model of linear or non-linear combination between anthropometric, sociodemographic and/or CT measures that allow us to adjust the lung volume and weight to calculate a more reliable pulmonary involvement (PI). The idea is based on the studies that indicate that pathologies such as COVID-19 and COPD alter the lung volume and weight.
1. Estimate lung volume and weight using CT measures;
2. Assess which anthropometric and sociodemographic variables have the greatest influence on lung volume and weight;
3. Calculate the adjusted lung volume and weight using, at first, a linear combination between the most relevant anthropometric and sociodemographic variables, such as sex, age and height;
4. Estimate the PI using the CT-estimated lung volume and the adjusted lung volume;
5. Compare PI between control cases and compromised cases (COVID-19, COPD etc.) using statistical tests;
6. Assess correlation between PI and lung weight (CT-estimated and adjusted) using statistical correlation analysis;
7. Use coefficients found in step 2 as a mathematical model to adjust lung volume and weight in new data.
1. Estimate lung volume and weight using CT measures;
2. Assess which anthropometric and sociodemographic variables have the greatest influence on lung volume and weight;
3. Calculate the adjusted lung volume and weight using, at first, a linear combination between the most relevant anthropometric and sociodemographic variables, such as sex, age and height;
4. Estimate the PI using the CT-estimated lung volume and the adjusted lung volume;
5. Compare PI between control cases and compromised cases (COVID-19, COPD etc.) using statistical tests;
6. Assess correlation between PI and lung weight (CT-estimated and adjusted) using statistical correlation analysis;
7. Use coefficients found in step 2 as a mathematical model to adjust lung volume and weight in new data.
Aims
• Use anthropometric and sociodemographic measures to adjust the CT-estimated lung volume and the CT-estimated lung weight;
• Assess the PI in diseases such as COVID-19 and COPD using the adjusted lung volume;
• Assess whether the PI is more strongly associated with the adjusted lung weight than with the CT-estimated lung weight;
• Assess whether the PI estimation should consider the adjusted lung volume and weight in order to obtain a more reliable result.
Collaborators
Alysson Roncally Silva Carvalho, MD, PhD - Cardiovascular R&D Center, Faculty of Medicine, Centro Hospitalar Universitário do Porto, Porto University, Porto, Portugal