The potential for higher spatial resolution using ultrasound-modulation of Cerenkov luminescence

591 Objectives Cerenkov luminescence is an optical signal produced by many medically relevant radionuclides and can provide an alternative signal for imaging. Cerenkov luminescence is produced continuously along the path of a charged particle as it propagates through a dielectric medium such as tissue. Cerenkov luminescence imaging (CLI) offers potential advantages of reduced cost and complexity over radiation imaging but suffers from scattering-limited resolution at even moderate depths in biological tissues. The Frank-Tamm formula ( dN/dx 1 - [1/2n2] ) describes the number of photons N produced per distance x traveled by the charged particle with velocity =v/c. and is strongly dependent on refractive index, n. The objective of this work is to explore a strategy that involves modulating the Cerenkov signal of the object being imaged, with the goal of improving spatial resolution. We hypothesize that by modulating the refractive index via ultrasound, in a spatially precise manner, we can recover spatial resolution and circumvent the scattering limit. We will present both theoretical and experimental work supporting this idea and discuss challenges in detecting the weak signal. Methods Monte Carlo simulations were performed using GAMOS, a GEANT4 based software package.. The simulation consisted of a small (3 mm diameter) spherical volume of refractive index n=1.33 (water equivalent) containing 18F, 11C, or 90Y radionuclides. The spherical volume is intended to approximate the ultrasound focal region. The refractive index was modulated up to 1%. Our ultrasound modulation apparatus uses an SiPM-based optical detector which views a water tank. A focused ultrasound transducer is driven by an RF power amplifier connected to a function generator. The ultrasound transducer is attached to a motorized translation stage. The SiPM is connected to several amplifier stages which are digitized using a data acquisition system. The instrument is controlled by custom software, which provides a graphical interface to acquire and display data. Results Monte Carlo simulations confirm that Cerenkov luminescence is highly sensitive to refractive index and small variations in refractive index (increases on the order of 1%) yield larger fractional changes in light output of up to ~13%. Radionuclides with higher energy beta spectra, such as 90Y, have smaller fractional changes in light yield with changing refractive index. Lower energy radionuclides such as 18F, have larger fractional changes in light yield. The sensitivity of the experimental system was characterized by modulating the absorption coefficient of a volume of water illuminated with a non-Cerenkov light source (405-nm LED) located external to the water volume. The ultrasound transducer was scanned perpendicularly to the LED beam and the modulated optical signal was detected by the SiPM. The attached plot shows SNR vs position of the ultrasound transducer for various amplifier settings. The LED is brightest at the beam center, causing the greatest signal modulation, and this results in a Gaussian shaped SNR profile with FWHM of ~3 mm, as expected for the LED beam width and US focal volume. The noise term in the SNR is determined from a scan performed when either the LED or ultrasound sources are off. Conclusions Monte Carlo simulations have provided a theoretical basis for experimental work and confirm that Cerenkov luminescence is sensitive to refractive index, especially for radionuclides with beta emissions just above the Cerenkov threshold. We have demonstrated an instrument that is capable of detecting small modulation of optical signals. While Cerenkov luminescence is much weaker than the LED source we have tested to date, the modulation of the signal is strongly enhanced by index of refraction effects and should be within the detection limits of the setup.