SESSION TITLE: Imaging Posters
SESSION TYPE: Original Investigation Poster
PRESENTED ON: Wednesday, October 30, 2013 at 01:30 PM - 02:30 PM
PURPOSE: Previously, a study suggested that total lung capacity could be ascertained using a single transverse CT slice taken at the level of the carina. So far, this study has not been repeated at any other institute. We hypothesized that the regression model derived in that study would give an accurate measure of total lung capacity when used at our institute.
METHODS: Charts from the pulmonary function test (PFT) facility at Unity Health System in Rochester, NY were enrolled in this study if a full PFT and high resolution noncontrast CT of the chest were performed within a one month period. Cross-sectional are of the lung was determined from the CT image using custom software which determined lung fields by Hounsfield units. Volume was extrapolated from cross sectional area using a previously published formula. Volume from CT image and from PFTs were analyzed with Pearson's correlation.
RESULTS: 46 charts met the above criteria. Our population was primarily Caucasian (95.65%), with a mean age of 65 years, mostly females (56.5%), with a mean body mass index of 31.5. The correlation coefficient between TLC and cross sectional area is 0.69 (p < 0.001). The regression model underestimated TLC by an average of 0.928 liters (p < 0.01).
CONCLUSIONS: The regression equation previously derived is consistently off when applied to our patient population. This error suggests that the model is not directly translatable from one institute to another; however, the strength of correlation suggests that a model designed to work with the specifics of an institute can be made. It remains to be seen if the observed difference is due to demographics or the technique in which images were obtained. The ability to accurately ascertain lung capacity from CT may prove a valuable diagnostic tool for the care of patients with lung disease and warrants further study.
CLINICAL IMPLICATIONS: There are many people who do not tolerate pulmonary function tests and this may give an objective, quantifiable and reproducible means to follow progression of their disease
DISCLOSURE: The following authors have nothing to disclose: Navitha Ramesh, Jazeela Fayyaz, Pratik Patel, Irtza Sharif
The technique uses open source software developed by the author that has only been used on research computers and never in a clinical setting. It has not been tested, approved, or even submitted for approval for clinical use.