Biological samples that are transparent and thin are difficult to observe wi
th a normal bright field microscope. Phase contrast and fluorescence micr
oscopes are commonly used, but they have limitations in terms of image c
larity and sample viability. AI-assisted 3D label-free quantitative microsco
py overcomes these limitations by providing high-quality images and qua
ntitative cellular information without the need for fluorescence dyes. By u
sing structured light illumination and reconstruction algorithms, this appr
oach enables the visualization of organelle structures and offers valuable s
tatistical data for studying cell activity and drug effects on cancer cells.
In contrast to expensive and specialized equipment required for commerc
ial phase contrast microscopy, our system is modular and integrates illumi
nation control, image acquisition, and phase reconstruction using thin-fil
m transistor (TFT) panels. It can be easily incorporated into existing invert
ed microscopes, providing high-quality images without the need for costl
y setups. Automation of the system, including light source control and im
age capture/storage, saves time and allows for flexible adjustments of the
light source position and incident angles. The AI algorithms enhance imag
e acquisition efficiency, sample applicability, and image quality, making th
e system suitable for capturing dynamic cellular information, particularly f
or rapidly changing cell behaviors like division or fusion, and enabling lon
g-term observations.