Multi-tissue and multi-scale approach for nuclei segmentation in H&E stained images

Abstract

BackgroundAccurate nuclei detection and segmentation in histological images is essential for many clinical purposes. While manual annotations are time-consuming and operator-dependent, full automated segmentation remains a challenging task due to the high variability of cells intensity, size and morphology. Most of the proposed algorithms for the automated segmentation of nuclei were designed for specific organ or tissues.ResultsThe aim of this study was to develop and validate a fully multiscale method, named MANA (Multiscale Adaptive Nuclei Analysis), for nuclei segmentation in different tissues and magnifications. MANA was tested on a dataset of H&E stained tissue images with more than 59,000 annotated nuclei, taken from six organs (colon, liver, bone, prostate, adrenal gland and thyroid) and three magnifications (10×, 20×, 40×). Automatic results were compared with manual segmentations and three open-source software designed for nuclei detection. For each organ, MANA obtained always an F1-score higher than 0.91, with an average F1 of 0.9305 ± 0.0161. The average computational time was about 20 s independently of the number of nuclei to be detected (anyway, higher than 1000), indicating the efficiency of the proposed technique.ConclusionTo the best of our knowledge, MANA is the first fully automated multi-scale and multi-tissue algorithm for nuclei detection. Overall, the robustness and versatility of MANA allowed to achieve, on different organs and magnifications, performances in line or better than those of state-of-art algorithms optimized for single tissues.

Enhancing Multi-tissue and Multi-scale Cell Nuclei Segmentation with Deep Metric Learning

“…Furthermore, cell nuclei segmentation, being such as important task in cancer detection, is complicated because of cells which often touch and overlap, making the separation problem of the cell nuclei difficult [37,38]. Comparatively, many researchers demonstrated successful segmentation results of nuclei in H&E-stained images by manipulating colour spaces in the image [39] and then using certain thresholds [10]. However, most of these algorithms are applicable only for the separation of nuclei, which have an ellipse-like shape [40].…”

“…The most compelling advantage of deep learning is the ability to generalise and automatically learn problem-specific features directly from the original data. As such, deep convolutional neural networks (CNNs) demonstrated advancements in image recognition tasks and achieved state-of-the-art results in many medical imaging applications [8,10]. U-Net and DeepCell are examples of models that were designed specifically for cell or nuclei segmentation.…”

Identification of histological features to predict MUC2 expression in colon cancer tissues

“…Advanced nuclear segmentation techniques can be used in future to improve this analysis: CellProfiler, 16,17 QuPath, 18 Fiji 19 and MANA. 20 Deep learning approaches are ideally suited for this type of problems because of availability of large number of training and validation data. Doing so will enable the models to better identify features that are indicative of MUC2-high cells, and will thus make them more effective at predicting a tissue sample as MUC2-high or MUC2-low.…”

Automated Segmentation of Fluorescence Microscopy Images for 3D Cell Detection in human-derived Cardiospheres

“…Then, the identification of the external borders of the cardiospheres is performed by applying an object-based detection scheme to each image of the stack. The core technology of this step is an original object-based detection strategy that we previously developed and adapted to these images 26 , which is briefly described in the following.…”