cardiovascular disease, and early signs of emphysema. High-resolution CT scans of the chest provide data sets that allow precise measurement of the size of lesions in the lung and allow tracking of changes in the size of these lesions, which is a very specific indication of malignancy. These CT data sets can also provide measures of occlusions or obstruction in cardiac vasculature (early indication of heart attack risk) and assessment of pulmonary function (indicative of emphysema).
CT imaging is a fast (a scan takes less than 10 seconds), non-invasive, and low-cost method for obtaining critical data that can be used to determine effective early intervention, preventing disease progression and greatly reducing the cost of treatment and improving the quality of life for the patient. To achieve the full potential of this technology will require new approaches to data analysis and quantitative feature extraction. For example, to deploy effective screening services, more precise methods for quantifying tumor size, determining the level of calcification in cardiac vasculature, and extracting measures of lung expiration capacity must be developed.
The critical contributions of optical technology to CT instrumentation (and other imaging platforms such as MRI and ultrasound) are often overlooked. Fundamentally, CT devices are optical instruments, employing photons chosen with wavelengths for which the body is partially transparent to precisely image the internal physiology of the patient. The x ray photon sources, the optics for focusing the x rays, and the detectors used to detect the x ray photons are designed employing many of the same techniques developed for the design of imaging instruments using visible light. Similarly, the mathematical methods for analyzing the raw transmitted x ray data, for reconstructing and visualizing three-dimensional models of internal anatomy, are almost identical to comparable techniques employed in other imaging modalities using visible or infrared light. Thus advances in detector technology, image reconstruction models, and techniques for quantitative feature extraction from large three-dimensional data sets will greatly enhance the performance of CT, MRI, and ultrasound imaging platforms.27
In general, advances in developing quantitative imaging data analysis procedures are hindered by the inability of the scientific community to access common data sets useful for comparing the performance of automated computerized methods to analyze the data. Establishment of the infrastructure to support public access to large data sets of image data and open-source software tools to extract clinically significant features from these data should be a national priority. Such an infrastructure is vital to accelerating advances in many different imaging modalities, including OCT, CT, MRI, ultrasound, x ray, diffuse optical imaging, and others.
27 Baer, T.M., J.L. Mulshine, and J.J. Jacobs. 2007. Biomedical Imaging Archive Network. Skeletal Radiology 36(9):799-801.