Using Artificial Intelligence for the automatic interpretation of chest CT images
National Lung Cancer Screening Trial (NLST) provides a well-curated datasets of chest low-dose computed tomography (LDCT) images of approximately 26, 732 participants. It is considered as one of the larges medical imaging cohort available for academic research. NLST has shown significant (20%) reduction in lung cancer mortality using low dose computed tomography. Although, the primary focus of this trial was on detecting early stage lung cancer in heavy smokers, many other clinical details related to different diseases like diabetes, hypertension, cardiac diseases etc. have been archived for the participants. We would like to build an automatic decision support system to interpret these chest CT images to assess the overall health of an individual.
The goal of this project is to investigate different AI techniques first to automatically detect and segment different anatomical tissues and organs like lungs, heart, liver, fat, etc. from the chest CT. Once these organs are detected and segmented, we would like to build deep learning networks to compare these organs between different individuals to find different patterns which could correlate to the healthiness of the organ. Overall, we would like to build an automatic system, which, given a chest CT image, could detect the important organs and quantify the normality and abnormality of organs. Due to the large sample size of NLST, we need a significant number of cases related to different diseases and outcome, which would give the statistical power to these analyses.
1) Automatic tissue and organ detection and segmentation using different AI techniques.
2) Investigate different deep learning techniques comparing organs between different individuals to find tissue patterns of normal and diseased tissues.
3) Quantify the healthiness of the organ and investigate the interrelation with the other organs and quantify overall health of an individual.
Chintan Parmar
Roman Zeleznik
-
Opportunistic assessment of steatotic liver disease in lung cancer screening eligible individuals.
Weiss J, Bernatz S, Johnson J, Thiriveedhi V, Mak RH, Fedorov A, Lu MT, Aerts HJWL
J Intern Med. 2025 Jan 27 PUBMED