Room: Room 202
Purpose: X-ray detector performance, in terms of the ability to produce high signal-to-noise ratio (SNR) images for a given amount of radiation, is quantified by the detective quantum efficiency (DQE). Current clinical detectors have low DQE (<30%) at high-frequencies, which pertaining more to small image features. The purpose of my research is to develop a novel x-ray detector design, called Apodized-Aperture Pixel (AAP), that has high DQE at high-frequencies (2.5x greater than current detector designs) for better visualization of small structures, such as microcalcifications in mammography, which are crucial for early detection of cancer.
Methods: The AAP design uses a micro-element sensor (0.005-0.025mm) and an anti-aliasing filter to produce desired clinical images of current pixel size (0.05-0.20mm). Conventional and AAP designs were modeled using cascaded system analysis to characterize and compare signal and noise. The modulation transfer function (MTF), Wiener noise power spectrum (NPS) and DQE were used to evaluate system performance with proof-of-concept experiments on cesium-iodide (CsI) and selenium (Se) clinical mammography detectors.
Results: MTF of the AAP design is 1.5x greater near the image cut-off frequency than conventional design for both CsI and Se converter layers. DQE of the AAP design was 2x greater with CsI and 2.4x greater with Se converter layers at high frequencies.
Conclusion: The AAP design achieves greater MTF and DQE at high-frequencies (near the image cut-off frequency) than conventional design. Higher MTF with the AAP design is due to use of micro-element sensor and percent improvement is independent of x-ray converter material type. Greater DQE with the AAP design is due to reduction of noise aliasing which depends on the x-ray interactions in the converter material. The AAP design can be used to acquire high SNR images for better visualization of fine detail that could improve cancer detection.