The objective of this research project is to develop a reliable, valid and clinically meaningful computer-derived measure of asbestos-induced pleural and parenchymal lung disease using HRCT images. The standard chest x-ray has proven to be an insensitive and non-specific means of identifying asbestos-induced disease. Conventional and high resolution computed tomography (HRCT) are more sensitive and perhaps more specific than the chest x-ray. Visual interpretation of the CT scan is subjective and prone to the same problems observed with the chest x-ray and yields little quantitative information. The development of a reliable, valid, objective, non-invasive measure of asbestos-induced lung pleural and parenchymal fibrosis would be a considerable advance in the area of detection and diagnosis of occupational lung disease.
The specific aims proposed in this project will use computer-derived measures of volume, surface area, density, texture, and fractal dimension from the HRCT scan to investigate the physiologic and biologic consequences of asbestos-induced lung disease. These computer-derived measures may prove to be novel, and complement traditional measures applied to the evaluation of asbestos- induced lung disease. The research tools developed in this research proposal, once validated, could also be applied to other forms of interstitial lung disease, where research has also shown that interpretation of the chest x-ray lacks sensitivity, with 10-20% of patients with pathologically proven interstitial fibrosis having normal chest roentgenograms. From a large cohort (n=1,223) of sheet metal workers, 240 subjects will be randomly selected. Each subject will be evaluated by functional assessment, pulmonary function tests, exercise ergometry, chest x-ray, conventional and high resolution chest CT scan, and bronchoalveolar lavage on a yearly basis. The presence or absence of asbestos-related pleural and parenchymal fibrosis will be determined from chest x-ray. Using a computer workstation, regions of interest will be traced and the lung parenchyma will be reconstructed in 3D from the HRCT scan. Specifically designed programs are used to objectively quantify the volume and surface area of pleural fibrosis, and to analyze the density, texture and fractal dimension of parenchymal radiological abnormalities. The reliability and validity (clinical and biological) of these computer-derived measures will be determined. The contribution of these computer-determined measures of parenchymal and pleural disease to our understanding of the physiology and biology of asbestos-induced lung disease will also be determined.
