Three-dimensional coded aperture imaging system proposed for early breast cancer detection using low tumor-to-background ratio

Coded aperture (CA) based image formation methods have been used in medical imaging for the past two decades. However, no commercially available Single Photon Emission Computed Tomography (SPECT) CA camera technology has been developed yet. This study investigated the lower tumor-to-background ratio (TBR) in 3D breast reconstruction with tumors. TBR is a potential factor for evaluating the malignancy of breast cancer because of its correlation with lesion size. The CA-based SPECT camera uses an iterative reconstruction algorithm called Maximum Likelihood Expectation Maximization (MLEM) with a mosaic Modified Uniformly Redundant Array (MURA) mask and anti-mask dual reconstruction. The background noise and intrinsic artifacts are suppressed by subtracting mask reconstructed images from the anti-mask images, thereby reducing the non-uniformity in the background. The reconstructed slices are then stacked to create a three-dimensional representation of the breast phantom. By employing a TBR of 5:1, mosaic MURA mask anti-mask subtraction has enhanced the quality of the reconstructed images and made it possible to detect breast lesions with a diameter of 4 mm. Furthermore, rather than using a sinogram, the proposed method can create a 3D breast phantom from a single projection. Simulation results suggest the feasibility of detecting 4 mm lesions at TBR = 5:1 using a single-projection mosaic MURA coded aperture system. Clinical validation is required to confirm translation to patient imaging.