DeepSeg: A Decision Support System for Enhanced Segmentation of Human Tissue Image

Cell Microscopy is considered a challenging field in the healthcare system. A subpart of it, the numeric determination of tissue types and RBC, WBC cells in human tissue constitutes a major part of research to clarify the probability of being infected with pathogens. Precise identification of these cell structures by humanized techniques is scientifically near-impossible and out of reach for high-time consumption. Based on the fact that under various light conditions, the qualitative nature of the image changes along with its property to depict the data in its true form for differentiating into its respective classes. Keeping the challenges in focus, this paper proposes an optimum autonomous cell segmentation method, DeepSeg, based on a neural network approach. An important step for the detection, monitoring, and analysis of the human tissue image is executed by the process of image segmentation. Moreover, feature extraction for forming image segments to understand the data in a much better way becomes easier with the application of DeepSeg. Visual and graphical illustrations of the result received confirm the validity and correctness of the proposed procedure. Further, the demanding situations and the problems faced at some point in the identification procedure and the strategies to tackle them have additionally been mentioned. This research work would prove to be a decision support mechanism in the domain of advanced pathology in medical science.