Nuclei Segmentation and Detection using Deep Convolutional Neural Networks

Automatic segmentation of microscopic images is an important task in medical image processing and analysis. Nuclei detection is an important example of this task. Imagine a faster research process for almost every deadly disease prevalent today. Diseases like cancer, heart diseases, COVID-19, rare disorders, etc. We have all seen or know some people who are fighting these diseases every day. Identifying the cells' nuclei is the starting point for most analyses because most of the human body's trillions of cells contain a nucleus full of DNA, the genetic code that programs each cell. The identification of nuclei helps researchers understand the underlying process. They can measure how different samples react to a certain drug. But, the manual process of identifying the cell's nuclei is a tedious task. This work proposes the use of deep learning techniques to perform segmentation and identify the cell's nuclei for which we chose a modified CNN architecture, U-Net. The results obtained show that U-Net can be a promising approach to efficiently segment and identify the cell's nuclei.