Computational two-photon microscopy for deep tissue imaging

ABSTRACT Intravital microscopy has been widely employed to visualize the structures and functions of cells in their natural environment. However, the imaging depth of intravital imaging is constrained by tissue scattering. State-of-the- art multiphoton microscopy can only produce sharp images up to about 1 mm deep in biological tissues at cellular

resolution, which restricts numerous biological applications requiring access deep inside tissues. Deep tissue imaging faces two major challenges. First, the 3D refractive index of tissues, often referred to as the 'scattering potential,' undergoes dynamic changes in living organisms. Second, the total light power for illuminating

biological tissue is limited to prevent thermal damage. Therefore, optical imaging techniques that are capable of high-speed scattering correction and maximizing the light power at the focal plane are needed. In this proposal, our group aims to develop innovative two-photon microscopy systems with scattering correction

for deep tissue imaging. The fundamental concept behind our techniques is coupling the illumination light into high-transmission channels (referred to as 'open channels') within the scattering tissues while mitigating light in the low-transmission channels. To achieve this goal, we will employ various techniques, encompassing

advancements in optical hardware and computational software. The hardware innovation is to develop high- speed, high-efficiency light modulation, while the software innovation is to develop new deep-learning-based algorithms for optimizing correction patterns and reducing computational complexity. Our proof-of-concept

experiments demonstrated a remarkable 27-fold increase in fluorescent intensity for imaging through a bone sample after scattering correction. Our goal is to complete the entire correction process within 10 ms and achieve a 100-fold increase in fluorescent signal after correction. We will validate our techniques by imaging intact mouse

organs. We will assess our techniques by comparing the image signal-to-noise ratio and the resolution with the ground truth images obtained after tissue sectioning. The novel two-photon microscopy techniques with scattering correction will enable deep tissue imaging at previously unattainable depths while achieving cellular resolution. In the long term, we will collaborate with my

colleagues in the Alzheimer’s Disease Research Center and the Comprehensive Cancer Center at UC Davis to apply the microscopy to broad biological research. This technique will facilitate the study of cellular structures and activity within living organisms, such as neuronal activity deep in the brain, immune responses in the lymph

nodes, cancer metastasis, and hematopoietic stem cells in the bone marrow.