There will be an estimated 159,000 deaths attributed to lung cancer along with 224,000 new lung cancer cases in 2014 in the United States. The high mortality rate is in part reflective of the lack of physical symptoms in the early stages and the previous lack of lung cancer screening. Since the Center for Medicare and Medicaid Services (CMS) has approved lung cancer screening using low dose computed tomography (CT) examination earlier this year, it remains to be seen if lung cancer screening reduces the morbidity and mortality associated with lung cancer. Chest CT exam has been proven to be a sensitive, non-invasive modality for early detection of lung cancer that may ultimately improve lung cancer mortality. However, the high sensitivity of CT exam often results in the detection of a large number of indeterminate nodules, which limits the efficacy of lung cancer screening. Often, follow-up scans and/or invasive biopsies are needed to determine the nature of these nodules, which may result in biopsy complications, exposure to ionizing radiation, patient anxiety, and cost. In light of the CMS decision to approve lung cancer screening with low dose CT, novel approaches to accurately and non-invasively assess the nature of indeterminate nodules could significantly improve lung cancer screening and treatment. We propose to investigate the local vascular patterns (or macro-vasculature) and image texture in a region of interest (ROI) surrounding a nodule depicted on CT images and assess whether macro-vascular and texture features can discriminate between benign and malignant nodules and predict tumor growth. We will develop software to extract and quantify image features in ROIs surrounding a nodule. We will investigate the association between the macro-vascular and texture features and the verified status of the nodules (benign/malignant) to test if: (1) macro-vascular and texture variables can discriminate benign and malignant nodules and (2) macro-vascular and texture variables can predict a nodule's growth rate (doubling- time). Our approach is dramatically different from the typical approaches that focus on quantifying the micro- vascular network within the tumor and may prove to add another tool for evaluating suspicious nodules. If successful, the macro-vascular or texture features could aid in improving the assessment of indeterminate nodules and quantifying response to lung cancer treatment. This may significantly improve the efficacy of CT based lung cancer screening by maintaining its high sensitivity while reducing false positive findings.

Public Health Relevance

Lung cancer is the leading cause of cancer deaths in the United States with an estimate of 159,000 deaths in 2014 in the United States alone. Chest CT exam is an effective tool to screen for early lung cancer detection, but there are challenges associated with screening. One challenge is the detection of a large number of suspicious nodules whose diagnosis is indeterminate. We are proposing to study the local vasculature surrounding a nodule to develop an image biomarker that can aid in discriminating benign and malignant nodules. If successful, our biomarker may also be able to identify responders and nonresponders to treatment.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Exploratory/Developmental Grants (R21)
Project #
1R21CA197493-01A1
Application #
9109919
Study Section
Special Emphasis Panel (ZRG1)
Program Officer
Henderson, Lori A
Project Start
2016-04-01
Project End
2018-03-31
Budget Start
2016-04-01
Budget End
2017-03-31
Support Year
1
Fiscal Year
2016
Total Cost
Indirect Cost
Name
University of Pittsburgh
Department
Radiation-Diagnostic/Oncology
Type
Schools of Medicine
DUNS #
004514360
City
Pittsburgh
State
PA
Country
United States
Zip Code
15213
Wang, Lei; Zhu, Jianbing; Sheng, Mao et al. (2018) Simultaneous segmentation and bias field estimation using local fitted images. Pattern Recognit 74:145-155
Wang, Xiaohua; Leader, Joseph K; Wang, Renwei et al. (2017) Vasculature surrounding a nodule: A novel lung cancer biomarker. Lung Cancer 114:38-43
Wang, Lei; Chang, Yan; Wang, Hui et al. (2017) An active contour model based on local fitted images for image segmentation. Inf Sci (N Y) 418-419:61-73
Wilson, David O; Pu, Jiantao (2016) The bell tolls for indeterminant lung nodules: computer-aided nodule assessment and risk yield (CANARY) has the wrong tune. J Thorac Dis 8:E836-7