A Novel Technique for Fluorescence Lifetime Tomography

Fluorescence lifetime imaging has emerged as a powerful tool for quantitatively assessing the molecular environment of live tissues in vivo. While fluorescence lifetime microscopy (FLIM) is a mature field, achieving effective 3D imaging in deep tissues has remained a significant challenge due to high scattering. In this study, we present a deep neural network-based approach, referred to as AUTO-FLI, which enables both 3D intensity and quantitative lifetime reconstructions at centimeters depth. This Deep Learning (DL)-based method incorporates an in silico framework to accurately generate fluorescence lifetime data for training and validation. The performance of this novel DL model is further validated with experimental data acquired on an anatomically accurate mouse-mimicking phantom. The results demonstrate that AUTO-FLI can provide precise 3D quantitative estimates of both intensity and lifetime distributions in highly scattering media. This method holds great promise for fluorescence lifetime-based molecular imaging at both the mesoscopic and macroscopic scales, with potential applications for pre-clinical translational research and fluorescence-guided surgery.