Skip to Main Content

COVID-19 is an emerging, rapidly evolving situation.

What people with cancer should know:

Get the latest public health information from CDC:

Get the latest research information from NIH:

Principal Investigator
Cleveland Clinic
Position Title
Full Staff Member of Biostatistics, Department of Quantitative Health Sciences, Cleveland Clinic
About this CDAS Project
NLST (Learn more about this study)
Project ID
Initial CDAS Request Approval
Jan 30, 2020
Developing a prognostic prediction model using Delta Radiomic Features and Clinical Information for Lung Cancer Detection
We aim to develop a prognostic prediction model for lung cancer detection using NLT study population. Radiomic features from baseline and follow-up screens will be extracted and combined with patients’ clinical information. Patient level data pertaining to but not limited to patients’ demographics, level of education, co-morbidities, medication list, smoking history and insurance status. Both overall and size-specific analyses to predict lung cancer incidence will be performed.

1) A machine learning model combining delta radiomic features with clinical information for lung cancer detection will be developed. AUROC, accuracy, specification and sensitivity will be reported. We aim to achieve this objective by applying deep learning methods. We hypothesize that the features to be extracted should present:1) nodules’ several attributes (nodule malignancy, size, speculation and etc..) and other clustering futures. 2) lung masses or more complicated tissue. 3) emphysema characteristics.
2) Aggregate the above features and construct a patient-level prediction model for lung cancer detection. Patients’ clinical characteristics will be added in this phase.


Xiaofeng Wang, Cleveland Clinic
Xiaozhen Han, Cleveland Clinic
Xinge Ji, Cleveland Clinic
Yige Sun, Case Western Reserve University