Comprehensive Review on Cytoplasm Nuclei Detection using Deep Learning

Nuclei Identification is one of the most difficult but also also extremely important tasks that are needed to be performed for the diagnosis and treatment for a lot of different maladies. The concept of nuclei detection is utilized in a variety of tests and other pathological processes that need to identity different cells in a histopathological landscape. The process of nuclei identification is largely done manually by the radiologists or microbiologists on the given slides. These slides can be hundreds in number and are meticulously marked painstakingly to achieve the cell detection for the recognition of any abnormalities or malignancies. The manual nature of the task allows for a large discrepancies in the form of human error that creep into the process. The manual marking is also labor intensive and can take up a large amount of time which can be critical for a variety of different diseases that require immediate diagnosis. Therefore, to improve this scenario and achieve an improvement in the nuclei detection this approach proposes an effective framework that utilizes data segmentation and feature extraction along with 3 different deep learning methodologies such as ResNet, U-Net, and CNN to achieve the cytoplasm nuclei detection. This approach will be well documented in the next version of this article