Journal of Stem Cell Research & Therapy

Filtered Back Projection (FBP) in Testing Reactor (TREAT) and Medical Imaging Sources

Joint Event on International Conference on STEM CELL AND ORTHOPEDICS & International Conference on PSYCHOSOMATIC AND LABORATORY MEDICINE

July 19-20, 2023 | Rome, Italy

Harish Aryal

Marymount University, USA

Scientific Tracks Abstracts: J Stem Cell Res Ther

Abstract:

Filtered back projection is a technique that unfolds the source information from the image. It has two distinct steps: a) Back projection of each event that provides information about the source location. Since these projections are summed up together, there is a significant blurring at the regions where these projections overlap b) Filtering by means of an appropriate filter in the frequency domain, we may reconstruct the image with higher resolution. Several image reconstruction techniques have been used such as maximum likelihood algorithms, and algebraic reconstruction techniques (ART). However, these indirect methods are computationally intensive. Filtered back projection is preferable when computational time is limited, and it has powerful applications in imaging when used with the right choice of filter. This research presents multiple illustrations of the Filtered Back Projection (FBP) technique. These demonstrations include a) FBP for a Hodoscope Plane neutron Flux obtained from the Transient Reactor Test (TREAT), b) FBP of Step Function, and c) FBP of a cosine source. FBP technique was successfully implemented in real-time imaging of fuel slumping behavior during transient tests at the TREAT reactor. Transient Reactor Test (TREAT) reactor is the principal facility in the US for safety testing of reactor fuel. It was in operation from 1959 to 1994 and has been restarted in 2018 after a 25-year hiatus. A key instrument used in TREAT is the hodoscope. It records the motion of fissionable material in the test capsule as the fuel fails during accident scenarios. It consists of multi-channel front and back collimators. During the experiment, fast neutrons produced in the test capsule are collimated providing a 2-dimensional “map” of fuel location, and are time stamped in the hodoscope detection system. Hodoscope system measurements need improvements to yield higherresolution images and the purpose of this research was to support imaging and was a part of the restart program and was successful. The research involves an MCNP model with a central fuel element replaced by dummy slotted fuel for sample positioning for flux information at the hodoscope plane and the corresponding image was successfully reconstructed using FBP and Hann filter with minimum blurring. It was found that the right choice of cut-off frequency removes noises that cause blurring, and the cut-off frequency determines how the filter affects both image noise and resolution. In other words, the higher the cut-off frequency, the better the spatial resolution, and therefore much more detailed image can be obtained. In conclusion, the Hann filter with a cut-off frequency close to 1 cycle/cm was recommended for image reconstruction. These imaging modalities have broader impacts on nuclear medicine for developing diagnostic and therapeutic procedures. For instance, to treat tumor cells a radioactive drug is typically administered in targeted organs along with gamma cameras to acquire two-dimensional projections of the activity. These projections data are then filtered and back-projected to get the image of the targeted organ. Proper selection of filter and cut-off frequencies are crucial in imaging will be discussed in detail in the preceding sections. In addition, detailed analytical FBP solutions for Cosine and Step functions are also presented to provide insights into its mathematical implications. In summary, this research provides both the analytical and computational analysis of FBP used in imaging that has applications both in nuclear engineering as well as in medical studies. In the future, we are optimistic to create a holistic model that provides high quality images with optimal resolution and minimal blurring.

Biography :

Harish Aryal is a Tenure-Track Assistant Professor in the College of Sciences and Humanities, School of Science, Mathematics, and Engineering at Marymount University since Spring 2023 prior to MU he was an Assistant Professor at the University of Texas-Permian Basin from 2020-2022. He holds Ph.D. in Nuclear Engineering (2020), two masters’ in mechanical (2016) and Nuclear Engineering (2014), and Bachelor’s in Nuclear Engineering with a Physics minor (2012), all degrees from ABET-accredited universities from the States with High Honours. His expertise involves the study of heat/mass transfer, and fluid mechanics processes in nuclear-mechanical systems. Key sub-areas: phase change, single/multi-phase fluids, HTGR, MSR, LMFBR, LWR using RELAP, CFD, MCNP, Scale, multi-physics, PRA, Mathematica, SAPHIRE, OpenBUGS, Fusion 360. Has grant experience with NSF, DoEd, DOE, NRC, and DOD (and has experience in modeling and simulation of reactor systems, radiation shielding, medical imaging, nuclear security, and non-proliferation. Currently, his lab involves creating renewable energy-Engr projects tailored to historically underrepresented cohorts at MU. Another avenue includes the use of radioisotopes in nuclear medicine image reconstruction/applications to support real-time imaging.