Quantifying the sensitivity limit of ICG imaging in the presence of tissue autofluorescence

Fluorescent imaging (FI) using indocyanine green (ICG) is a powerful tool in medical diagnostics and surgery. Although numerous studies have focused on optimizing injection protocols and suppressing excitation light leakage, tissue autofluorescence has not been widely recognized as a fundamental factor limiting sensitivity.

We aim to quantitatively determine the sensitivity limit for ICG detection in biological tissues, accounting for background signals from both scattered excitation light and tissue autofluorescence.

We combine experiments on tissue phantoms with varying ICG concentrations and Monte Carlo numerical simulation of light transport in media with different optical properties. Human skin autofluorescence was quantified in vivo using a nonfluorescent reference and a model medium with a known ICG concentration.

It was established that skin autofluorescence is the dominant source of background, exceeding the scattered light by 4 to 25 times in the imaging system used. The determined ultimate sensitivity for ICG detection in biological tissue is $8\times {10}^{-12}$ to $3\times {10}^{-11}M$ when accounting for the autofluorescence signal.

Tissue autofluorescence is a fundamental factor limiting the sensitivity of ICG FI in the near-infrared range. The developed approach will allow for future optimization of imaging equipment and protocols for ICG and other contrast agents.