Lung cancer is the leading cause of cancer death in the United States and worldwide. The National Lung Screening Trial (NLST) was a randomized multicenter study comparing low-dose helical computed tomography (CT) with chest radiography in the screening of older current and former heavy smokers for early detection of lung cancer. The NLST found a 20% reduction in lung cancer deaths in patients screened by low-dose helical CT compared to plain chest radiography with just three rounds of screening. However, there is much room for improvement in CT screening. Across all three rounds, when a positive screening result was found, 96.4% of the low-dose helical CT tests were false positive, a finding that identifies a suspiciously appearing nodule to be malignant, despite being benign. The vast majority of false positive results were most likely due to the detection of benign infectious or inflammatory nodules, but these results can cause weeks or months of anxious waiting until a confirmatory diagnosis is obtained by biopsy or follow-up imaging.

The current care paths for an initially positive CT screening result are designed to minimize the rate of negative, presumably unnecessary, invasive biopsies or surgery. A final decision on the appropriate intervention will be made after a series of follow-up imaging studies that depend on characterizing potentially malignant lesions based on size and morphology to rule out or confirm growth over time. This, however, requires several months to a year (the time interval to the next scheduled screening) to obtain an answer. This can be costly in terms of resources, time, the burden of anxious waiting on the patient, and potentially morbidity or mortality from delayed diagnosis and treatment.

In this project, we will address this shortcoming of existing lung cancer screening methods by developing a CT-based method of characterizing a mechanical property of pulmonary lesions, specifically tissue stiffness that should have a higher specificity than purely anatomic low-dose CT.

It is therefore the aim of the proposed study to decrease the false positive rate of CT screening by analyzing the mechanical properties of suspiciously appearing tissue during CT screening. We hypothesize that malignant pulmonary nodules are stiffer than benign nodules and that this difference in stiffness can be used to differentiate cancerous from benign nodules, which would help to decrease the false positive rates of CT screening. A measure of stiffness can be derived from high-resolution four-dimensional computed tomography (4D CT) using established mathematical formalisms and image processing algorithms. Unlike conventional 3D CT imaging that results in a static image of the scanned anatomy, 4D CT incorporates also the temporal changes of the anatomy caused by respiratory motion, yielding a CT "movie" that allows the evaluation of tumor motion and the calculation of the stiffness.

We are envisioning our method to be used as a confirmatory test in patients who have had an initial positive CT scan in order to more specifically identify lesions requiring biopsy without employing a "watch and wait" strategy involving follow-up CT imaging, which is the current path of care. This would not only decrease the burden of the prolonged waiting time for the patients, it would also be beneficial for the patient to begin treatment earlier. An early diagnosis and an early start of any appropriate type of treatment will ultimately result in decreased mortality rates.

If the proposed project should prove successful, this technique can be readily adapted in most diagnostic radiology departments as the majority of modern CT scanners are capable of 4D CT, and our image processing software can be easily implemented. The ease of adoption and clinical implementation of our technique would make our method immediately available to any high-risk lung cancer patients undergoing CT screening. The risks associated with our method are minimal, and the potential benefits or early detection of lung cancer may ultimately result in decreased mortality rates.