Which Pixel to Annotate: A Label-Efficient Nuclei Segmentation Framework

Lou, Wei; Li, Haofeng; Li, Guanbin; Han, Xiaoguang; Wan, Xiang

doi:10.1109/tmi.2022.3221666

Cited by 19 publications

(4 citation statements)

References 88 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(b) Equispaced sampling denotes as Equispaced . (c) One-shot active learning methods include K -median Kolluru et al (2021), K -means Arthur and Vassilvitskii (2007), Farthest Point Sampling (FPS) Moenning and Dodg-son (2003), Representative Annotation (RA) Zheng et al (2019) and Consistency-based Patch Selection (CPS) Lou et al (2022). For sub-volume selection methods, we only compare CGS with random and equispaced selection due to the lack of related research.…”

Section: Resultsmentioning

confidence: 99%

Addressing Distribution Mismatch for Effective Semi-Supervised Neuron Segmentation

Zhang,

Zhai,

Guo

et al. 2024

Preprint

View full text Add to dashboard Cite

Semi-supervised learning holds promise for cost-effective neuron segmentation in Electron Microscopy (EM) volumes. This technique fully leverages extensive unlabeled data to regularize supervised training for robust predictions. However, diverse neuronal patterns and limited annotation budgets may lead to distribution mismatch between labeled and unlabeled data, hindering the generalization of semi-supervised models. To address this issue, we propose an improved pipeline for cost-effective neuron segmentation by integrating selective labeling and semi-supervised training. For selective labeling, we present an unsupervised heuristic tailored for labeled dataset selection in EM volumes. Guided by self-supervised learning on local patches, representative sub-volumes comprising spatially associated patches are greedily selected via a coverage-based criterion. Those sub-volumes can effectively reflect unlabeled data distribution within a limited budget. For semi-supervised training, we introduce spatial mixing into neuron segmentation and integrate it within a Siamese architecture. This enhancement allows us to reinforce cross-view consistency through intra- and inter-mixing of labeled and unlabeled datasets. The proposed strategies bridge the distribution gap and encourage the model to learn shared semantics across datasets, promoting more effective semi-supervised learning. Extensive experiments on public datasets consistently demonstrate the effectiveness of the proposed pipeline, providing a practical and efficient solution for large-scale neuron reconstruction. Codes and data will be available.

show abstract

Section: Resultsmentioning

confidence: 99%

Addressing Distribution Mismatch for Effective Semi-Supervised Neuron Segmentation

Zhang,

Zhai,

Guo

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…However, the handcrafted features limit the representation capabilities of nuclei entities. Recently, the nuclei classification models usually infer cell types based on the CNNs for nucleus segmentation (Zhang et al 2017;Basha et al 2018;Lou et al 2022Lou et al , 2023bMa et al 2023;Yu et al 2023) or nucleus centroid detection (Abousamra et al 2021;Huang et al 2023b). Graham et al (2019) propose a CNN of three branches, predicting nucleus types for the segmented nucleus instances.…”

Section: Related Workmentioning

confidence: 99%

Cell Graph Transformer for Nuclei Classification

Lou,

Li,

Wan

et al. 2024

AAAI

View full text Add to dashboard Cite

Nuclei classification is a critical step in computer-aided diagnosis with histopathology images. In the past, various methods have employed graph neural networks (GNN) to analyze cell graphs that model inter-cell relationships by considering nuclei as vertices. However, they are limited by the GNN mechanism that only passes messages among local nodes via fixed edges. To address the issue, we develop a cell graph transformer (CGT) that treats nodes and edges as input tokens to enable learnable adjacency and information exchange among all nodes. Nevertheless, training the transformer with a cell graph presents another challenge. Poorly initialized features can lead to noisy self-attention scores and inferior convergence, particularly when processing the cell graphs with numerous connections. Thus, we further propose a novel topology-aware pretraining method that leverages a graph convolutional network (GCN) to learn a feature extractor. The pre-trained features may suppress unreasonable correlations and hence ease the finetuning of CGT. Experimental results suggest that the proposed cell graph transformer with topology-aware pretraining significantly improves the nuclei classification results, and achieves the state-of-the-art performance. Code and models are available at https://github.com/lhaof/CGT

show abstract

“…To avoid distorting the image by sharpening the RGB image directly (because RGB lighting and color are mixed and it is difficult to apply convolution to the image without changing the color), it is necessary to first turn the RGB image into a YUV image by converting the three RGB channels into one channel representing luminance and two channels representing chrominance [54]. Assuming that the original image is represented as

{D_{rgb}}

.…”

Section: System Model Designmentioning

confidence: 99%

Two‐stage coarse‐to‐fine method for pathological images in medical decision‐making systems

He,

Zhu,

et al. 2023

IET Image Processing

View full text Add to dashboard Cite

Artificial intelligence decision systems play an important supporting role in the field of medical information. Medical image analysis is an important part of decision systems and an even more important part of medical diagnosis and treatment. The wealth of cellular information in histopathological images makes them a reliable means of diagnosing tumors. However, due to the large size, high resolution, and complex background structure of pathology images, deep learning methods still have various difficulties in the recognition of pathology images. Based on this, this study proposes a two‐stage continuous improvement‐based approach for pathology image recognition in medical decision systems. For pathology images with complex backgrounds, normalization and enhancement is performed to remove the effects of noise color, and light‐dark inconsistencies on the segmentation network. The continuous refinement PSP Net (CRPSPNet) is then designed for accurate recognition of the pathology images. CRPSPNet is divided into two stages: Pyramid Scene Parsing Network segmentation to obtain coarse segmentation results; and continuous refinement model refines the results of the first stage. Experiments using more than 1,000 osteosarcoma pathology images have shown that the method gives more accurate results with fewer computer resources and processing time than traditional optimization models. Its Intersection over Union achieves 0.76.

show abstract

Which Pixel to Annotate: A Label-Efficient Nuclei Segmentation Framework

Cited by 19 publications

References 88 publications

Addressing Distribution Mismatch for Effective Semi-Supervised Neuron Segmentation

Addressing Distribution Mismatch for Effective Semi-Supervised Neuron Segmentation

Cell Graph Transformer for Nuclei Classification

Two‐stage coarse‐to‐fine method for pathological images in medical decision‐making systems

Contact Info

Product

Resources

About