Nuclei and glands instance segmentation in histology images: a narrative review

Examination of tissue biopsy and quantification the various characteristics cellular processes are clinical benchmarks in cancer diagnosis. Nuclei glands instance segmentation greatly assists high-throughput process accurate appraisal biopsy. It subsequently makes a significant improvement to computational pathology for diagnosis, treatment planning, survival analysis. Recent advancements field computer vision have automated manual, laborious, time-consuming histopathological analysis process. Automated image images cells tissues trace entirety ultrastructures, has been an active area research medical informatics decades. The developments computers, microscopy hardware, availability large-scale public datasets further fastened development this field. And realization that scientific diagnostic calls fresh ways undertake, captivated contemporary attention. In survey, 126 papers illustrating AI-based methods nuclei published last five years (2017–2022) deeply analyzed, limitations current approaches open challenges discussed. Moreover, potential future direction is presented, contribution state-of-the-art summarized. Further, generalized summary publicly available detailed insights on grand top-performing specific each challenge also provided. Besides, we intended give reader state existing pointers directions developing can be used practice enabling improved grading, prognosis, planning cancer. To best our knowledge, no previous work reviewed histology focusing segmentation.