CIA-Net: Robust Nuclei Instance Segmentation with Contour-Aware Information Aggregation

Abstract

Accurate segmenting nuclei instances is a crucial step in computer-aided image analysis to extract rich features for cellular estimation and following diagnosis as well as treatment. While it still remains challenging because the wide existence of nuclei clusters, along with the large morphological variances among different organs make nuclei instance segmentation susceptible to over-/under-segmentation. Additionally, the inevitably subjective annotating and mislabeling prevent the network learning from reliable samples and eventually reduce the generalization capability for robustly segmenting unseen organ nuclei. To address these issues, we propose a novel deep neural network, namely Contour-aware Informative Aggregation Network (CIA-Net) with multilevel information aggregation module between two task-specific decoders. Rather than independent decoders, it leverages the merit of spatial and texture dependencies between nuclei and contour by bi-directionally aggregating task-specific features. Furthermore, we proposed a novel smooth truncated loss that modulates losses to reduce the perturbation from outliers. Consequently, the network can focus on learning from reliable and informative samples, which inherently improves the generalization capability. Experiments on the 2018 MICCAI challenge of Multi-Organ-Nuclei-Segmentation validated the effectiveness of our proposed method, surpassing all the other 35 competitive teams by a significant margin.

Hover-Net: Simultaneous segmentation and classification of nuclei in multi-tissue histology images

“…In particular, we consistently obtain an improved performance over DIST, which justifies the use of our proposed horizontal and vertical maps as a regression target. We also report a better performance than the winners of the Computational Precision Medicine and MoNuSeg challenges [30], [29], that utlised the CPM-17 and Kumar datasets respectively. Therefore, HoVer-Net achieves state-of-the art performance for nuclear instance segmentation compared to all competing methods on multiple datasets that consist of a variety of different tissue types.…”

“…For this experiment, because we do not have the classification labels for all datasets, we perform instance segmentation without classification. This enables us to We compared our proposed model to recent segmentation approaches used in computer vision [21], [44], [32], medical imaging [22] and also to methods specifically tuned for the task of nuclear segmentation [25], [23], [31], [29], [30]. We also compared the performance of our model to two open source software applications: Cell Profiler [42] and QuPath [43].…”

“…[28] used the contour predictions as input into a further network for segmentation refinement. [29] proposed CIA-Net, that utilises a multi-level information aggregation module between two task-specific decoders, where each decoder segments either the nuclei or the contours. A Deep Residual Aggregation Network (DRAN) was proposed by [30] that uses a multi-scale strategy, incorporating both the nuclei and nuclear contours to accurately segment nuclei.…”

Nuclear Instance Segmentation Using a Proposal-Free Spatially Aware Deep Learning Framework

“…Results and comparative analysis Performance of the proposed model is compared against several deep learning based methods as reported in Table 1. Except the baseline method (CNN3) [7] which categories the image pixels into three classes using a CNN-based classifier, other methods in Table 1 (DR-Net [11], DCAN [1], BES-Net [12], and CIA-Net [15]) took a dense prediction approach and used encoder-decoder like CNN. As deduced from the results in Table 1, our proposed method based on SpaNet outperforms other state-of-the-art methods.…”

PFA-ScanNet: Pyramidal Feature Aggregation with Synergistic Learning for Breast Cancer Metastasis Analysis

“…This is caused by the mislabelled region and zigzag boundaries existed in WSI annotations, which have the tendency to overwhelm other informative regions in segmentation loss calculation and thus dominate the gradients. To solve the problem, we modify the binary cross-entropy loss into a truncated form [10] that can reduce the contribution of outliers with high confidence prediction. Our segmentation loss is shown as follows:…”