Comparison of the performances between anti-scatter grid and slot scanning technique for digital chest radiography: effect of anatomical background

To investigate and compare the nodule detection in digital chest imaging between anti-scatter grid and slot scanning methods. Anthropomorphic chest phantom was imaged with a flat-panel based digital radiography system. The system was operated in both the slot scanning and full-field modes with and without anti-scatter grid. Imaging technique was 120 kVp and 1 to 16 mAs for both modes. 10-mm in diameter computer-simulated nodules with a nominal peak contrast ratio of 5% were inserted at hilum and mediastinum locations by applying SPR values. 4-AFC experiment was conducted to measure the ratio of correct observations as a function of the exposure level for various imaging conditions and locations. These images were displayed randomly on a review workstation and reviewed by three observers. The average ratios of correct observations were computed across over the readers. The statistical significance of the differences in fractions between imaging techniques was computed by the Student t-test. Nodule detection was not significantly improved by raising the exposure level in the hilum and mediastinum regions. Slot scan without grid and with grid received the highest and next highest fractions of correct response, followed in order by full-field without and with grid for the hilum region, and full-field with and without grid for the mediastinum region. Statistical significant difference was found for most comparisons between slot scan with or without grid and full-field with or without grid.