State-of-the-Art of Deep Learning Methods for Microscopic Image Segmentation: Applications to Cells, Nuclei, and Tissues

• DOI: 10.20944/preprints202409.2030.v1
• Publication Date: 2024-09-27T01:04:10Z
• Authors: Fatma Krikid; Hugo Rositi; Antoine Vacavant
• Keywords: 0301 basic medicine; microscopic image; Computer applications to medicine. Medical informatics; R858-859.7; deep learning; QA75.5-76.95; Review; 03 medical and health sciences; nucleus segmentation; tissue analysis; Electronic computers. Computer science; Photography; [INFO]Computer Science [cs]; TR1-1050; image segmentation; cell segmentation

Microscopic image segmentation (MIS) plays a pivotal role in various fields such as medical imaging and biology. With the advent of deep learning (DL), numerous methods have emerged for automating and improving the accuracy of this crucial image analysis task. This systematic literature review (SLR) aims to provide an exhaustive overview of the state-of-the-art DL methods employed for the segmentation of microscopic images. In this review, we analyze a diverse array of studies published in the last five years, highlighting their contributions, methodologies, datasets, and performance evaluations. We explore the evolution of DL techniques and their adaptation to specific segmentation challenges, from cell and nucleus segmentation to tissue analysis. This paper, through the integration of existing knowledge, provides valuable perspectives for researchers involved in the field of microscopic image segmentation.