Accurate segmentation of nuclear instances using a double-stage neural network

Abstract

Automatic nuclear instance segmentation is a crucial task in computational pathology as this information is required for extracting cell-based features in down-stream analysis. However, instance segmentation is a challenging task due to the nature of histology images which show large variations and irregularities in nuclei appearances and arrangements. Various deep learning-based methods have tried to tackle these challenges, however, most of these methods fail to segment the nuclei instances in crowded scenes accurately, or they are not fast enough for practical usage. In this paper, we propose a double-stage neural network for nuclear instance segmentation which leverages the power of an interactive model, NuClick, for accurate instance segmentation by replacing the user input with a nuclei detection module, YOLOv5. We optimized the proposed method to be lightweight and fast and show that it can achieve promising results when tested on the largest publicly available nuclei dataset.

Generating synthetic data in digital pathology through diffusion models: a multifaceted approach to evaluation

“…During the preparation of this work the authors used ChatGPT3.5 8 for grammar and spelling checks, as well as rephrasing to enhance readability. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.…”

“…The transition from conventional glass slides to high-resolution Whole-Slide Images (WSIs) has presented exciting opportunities for harnessing artificial intelligence, particularly deep learning, to revolutionize pathology, paving the way for computational pathology (1). Deep learning algorithms have demonstrated impressive capabilities in automating critical tasks such as cancer detection or grading (1) (3) (4) (5) (6) as well as cell segmentation (7) (8). However, the success of these deep learning algorithms heavily relies on the availability and quality of large-scale, annotated datasets (9), which can be challenging in the medical domain where data scarcity and privacy concerns are prominent (10).…”