This project focuses on developing a computer vision solution using Deep Learning for the analysis of embryonic tissues. Enhanced tissue engineering workflows through computer vision approach. Provided efficient method for biologists to assess and improve tissue quality. The solution is implemented using PyTorch and consists of a two-phase model: Segmentation and Profiling.

Developed dynamic UNet Architecture from scratch: Designed to process high-resolution microscopy images, enabling precise segmentation of embryonic tissues.

Automated evaluation and quality enhancement of lab-cultivated embryonic tissues

Demonstrated AI's potential in medical image analysis

Integrated deep learning techniques in microscopy image processing

Developed custom CNN model for tissue property extraction from variable-resolution images

Its Built to accommodate variable image resolutions, allowing for the extraction of multiple tissue properties from segmented images.