Project Summary
Positron Emission Tomography (PET) is a unique and powerful clinical imaging modality that can detect
molecular and functional information from the human body. However, the current scintillation-based detection
mechanism poses constraints in reducing the coincidence time resolution (CTR), which indicates how accurately
we can measure the point that positron was emitted from in the body, thus the signal-to-noise ratio (SNR). A
novel detection mechanism is proposed to significantly reduce CTR, down to 10-picosecond (ps), by directly
detecting the ionization charges, which occur before scintillation photon emission. With the 10-ps CTR, several
clinical advantages are expected, including real-time imaging with a much lower patient dose. To achieve less
than 10-ps CTR in PET, Dr. Jeong will use the advanced ultrafast optics technique of spectral encoding which
uses colors of light to measure the arrival timing of the annihilation photons created from the position. Previously,
a proof-of-concept experiment and simulation using the spectra encoding was performed at the SLAC national
accelerator laboratory. For this project, the same technique will be tailored to the annihilation photons, which
have higher energy and lower particle numbers, as described in Aim 1. Establish a sensitive free-space optical
setup that achieves < 10ps CTR, where spectral information acquired from the ultrafast laser probes will be
compared for the coincidence arrival timings of the annihilation photons. In Aim 2, in parallel, fiber-based
detection modules will be developed for stronger spatial overlap of the probe laser beam and the ionization track,
further enabling efficient beam delivery and high-packing fraction. In Aim 3, with the detector modules developed
in Aims 1 and 2, Dr. Jeong will demonstrate the dynamic imaging performance of the 10-ps CTR PET modules
toward functional brain imaging. Dr. Jeong will first perform simulations using the CTR values in the 10-ps range
to obtain expected image quality then collect dynamic data from radioactive sources with fixed positions between
the two modules, with milliseconds frames, relevant to human brain activity timescales. The SNR from the <10-
ps CTR detectors will be ~5 times greater than the SNR of data taken with the same time frames from two
detectors with 230-ps CTR. The proposed project is highly interdisciplinary, requiring forefront expertise in
medical imaging, ultrafast optics, fast data acquisition, and image reconstruction for PET. Through this project,
Dr. Jeong will be further trained in the areas of ultrafast optics and medical imaging, expanding from the previous
training experience to become an independent investigator. Experts in the above research areas will advise and
help Dr. Jeong to obtain new skill sets at Stanford, with excellent institutional support. Through this K99/R00
training plan which includes auditing classes, grant writing, career development, and student mentoring, Dr.
Jeong will be able to obtain critical skills to flourish as an independent investigator.