Correcting Cherenkov Images for Large-Scale Tissue-Optical Property Attenuation Using SFDI and Patterned Light Reflectance for Quantitative Dosimetry

摘要

Tissue optical properties attenuate a substantial percentage of the optical light being detected during real-time Cherenkovacquisition, which distorts the signal linearity previously observed with absorbed dose in homogeneous media. This hindersprogression toward establishing quantitative dosimetry using Cherenkov imaging in vivo. By spectrally weighting effectiveattenuation ( _( )) maps generated by multi-wavelength Spatial Frequency Domain Imaging (SFDI), it became possible tomore successfully correct clinical Cherenkov images for areolar attenuation (6% difference, as compared to the treatmentplan) compared to selecting one wavelength channel in a previous study (41% difference). Additionally, using a reflectedlight-based patient positioning system, we were able to characterize and correct for gross tissue optical properties in patientimages, namely for large-scale surface and subsurface attenuation. While the use of wide-field SFDI enabled pixel-bypixelcorrections, the benefit of using an integrated, light-based system for reflectance-based corrections negates the useof an external imaging system, which substantially smooths workflow.