Purpose: The ability to monitor the distribution and migration of cells in animal model is crucial to support the development of cancer therapy. To monitor single or few cells in live animal, we innovated an ultra-sensitive single-pixel bioluminescence tomography (SPBLT) by utilizing single-pixel imaging (SPI) technology and single-photon avalanche diode (SPAD). Simulation and experiment are proposed to validate the capability of SPBLT in detecting weak bioluminescence signal emitted from cells in vivo.
Methods: Our SPBLT comprises lens, digital micromirror device (DMD), and single pixel detector/SPAD. The DMD modulates the spatial patterns of bioluminescence image (BLI) generated from cells. The corresponding light intensity is collected by the SPAD. The BLIs are reconstructed through total variance (TV) algorithm with the information of DMD pattern and SPAD-collected intensity, and later used as the input for 3D BLT reconstruction to retrieve cell location in vivo. For simulation study, we generated artificial weak-signal BLIs, based on a phantom image, by adjusting average count of the images. The simulated BLIs were then modulated according to pre-defined DMD patterns. The Poisson and Gaussian models were incorporated to mimic system shot and dark noise, respectively. The peak signal-to-noise ratio (PSNR) was adopted to evaluate the BLI reconstruction. After the SPI-generated BLI is validated, the 3D BLT reconstruction will be performed in both simulation, phantom, and animal model setting, to validate SPBLT in target localization accuracy.
Results: Our simulations show that SPI can improve reconstructed image over 9.5 dB relative to conventional detection method, such as camera acquisition, especially at ultra-weak signal scenario. In addition, SPI is less susceptible against the detection noise.
Conclusion: We presented a novel in vivo 3D cell-tracking technique, SPBLT. The SPBLT is expected to provide investigators the new capability to understand how cell migrate in vivo and support the development of cancer therapy.
Funding Support, Disclosures, and Conflict of Interest: This work is in part supported by NIH R37 CA230341, R01 CA240811, R21 CA223403, and CPRIT RR200042.