Fractal Dimension in Assessing Pulmonary Ground Glass Opacities
Research type
Research Study
Full title
The Role of Fractal Dimension Analysis in Neoplastic and Non-Neoplastic Pulmonary Ground Glass Opacity
IRAS ID
161498
Contact name
Gordon W Cowell
Contact email
Sponsor organisation
NHS Greater Glasgow & Clyde
Duration of Study in the UK
1 years, 0 months, 1 days
Research summary
Ground-glass opacity (GGO), a frequently encountered sign of lung disease on computed tomography (CT) scans, can represent cancer or benign disease. For radiologists evaluating such scans, assessment and risk stratification of a GGO can be challenging, particularly if incidental. Thus any method to reliably classify GGOs would be of great value in guiding patient management and future imaging, particularly as regards lung cancer.
Fractal dimension (FD) analysis is a calculation of image texture complexity, an objective method which measures the textural complexity of an area of GGO. Cancerous lesions tend to have a higher FD than benign lesions which otherwise may look identical to the naked eye. Several software programs exist for fractal analysis. We will use a publicly available imaging processing program called ImageJ with a plugin called FracLac to define a specific area, called a region of interest (ROI), within a GGO, and calculate the FD.
We plan to perform a retrospective study measuring FD on CT and PET-CT scans previously acquired in patients with known lung cancer and a non-cancerous cause of GGOs called NSIP (a form of lung inflammation and fibrosis). The FD from a representative GGO for each CT scan will be recorded, with subsequent statistical analysis.
In addition, for GGO cancer cases with available PET-CT scans, the FD will be correlated with the maximum standardised uptake value (SUVmax) on PET-CT, a measure of tumour metabolic activity. The purpose of this comparison is to test whether the FD could provide similar biological information about the metabolic activity of a lung cancer, using the data already contained within the simpler CT scan.
Although FD measurement is an established technique, to the best of our knowledge differentiation of cancer and non cancer GGO by FD calculation has not been previously reported.
REC name
South Central - Hampshire B Research Ethics Committee
REC reference
15/SC/0131
Date of REC Opinion
27 Feb 2015
REC opinion
Favourable Opinion