Author(s): Tyler O'Loughlin
Mentor(s): Vern Hart
Institution UVU
Coherent diffraction imaging (CDI) is a powerful technique for reconstructing phase shifts occurring in biological samples by capturing the interference between scattered light. These diffraction patterns contain spectral information, where shorter wavelengths (higher spatial frequencies) yield images with enhanced spatial resolution, improving detail and accuracy. However, higher spatial frequencies are located further from the central axis, rendering them fainter and more challenging to measure. To address this, our group developed a gyroscopic CDI system featuring a rotating platform driven by stepper motors. This platform positions the sample between a broadband light source and a beam profiler, which are mounted on opposite sides, enabling precise rotation for enhanced data collection. This innovative system maintains normal incidence between scattered light and the detector, enabling accurate measurement of faint short-wavelength signals without compromising resolution. Our reconstruction approach mirrors that of computed tomography (CT) scanning, where individual 2D slices of scattered light data are sequentially reconstructed and combined to form a detailed 3D model. This slice-based methodology enhances spatial accuracy and reduces noise and artifacts, providing a more precise volumetric representation of complex samples. However, the unique data collection geometry necessitates a tailored reconstruction algorithm. Thus, we propose modifying the 'filtered back projection' (FBP) algorithm to accommodate the slice-based gyroscopic CDI system. The adaptation involves calibrating for rotational and geometric variations, ensuring seamless integration of 2D slices into a coherent 3D model. This methodology lays a foundation for high-resolution 3D imaging of biological samples, with potential applications in cellular imaging, tissue analysis, and advanced medical diagnostics.