MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network
Accurate nuclei segmentation is essential for extracting quantitative information from histology images to support disease diagnosis and treatment decisions. However, precise segmentation is challenging due to the presence of clustered nuclei, varied morphologies, and the need to capture global spat...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Alexandria Engineering Journal |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1110016824011669 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841553814983802880 |
|---|---|
| author | Xuening Wu Yiqing Shen Qing Zhao Yanlan Kang Wenqiang Zhang |
| author_facet | Xuening Wu Yiqing Shen Qing Zhao Yanlan Kang Wenqiang Zhang |
| author_sort | Xuening Wu |
| collection | DOAJ |
| description | Accurate nuclei segmentation is essential for extracting quantitative information from histology images to support disease diagnosis and treatment decisions. However, precise segmentation is challenging due to the presence of clustered nuclei, varied morphologies, and the need to capture global spatial correlations. While state-of-the-art Transformer-based models employ tri-decoder architectures to decouple the segmentation task into nuclei, edges, and cluster edges segmentation, their complexity and long inference times hinder clinical integration. To address this, we introduce MoE-NuSeg, a novel Mixture of Experts (MoE) network that consolidates the tri-decoder into a single decoder. MoE-NuSeg employs three specialized experts for nuclei segmentation, edge delineation, and cluster edge detection, thereby mirroring the functionality of tri-decoders while surpassing their performance and reducing parameters by sharing attention heads. We propose a two-stage training strategy: the first stage independently trains the three experts, and the second stage fine-tunes their interactions to dynamically allocate the contributions of each expert using a learnable attention-based gating network. Evaluations across three datasets demonstrate that MoE-NuSeg outperforms the state-of-the-art methods, achieving an average increase of 0.99% in Dice coefficient, 1.14% in IoU and 0.92% in F1 Score, while reducing parameters by 30.1% and FLOPs by 40.2%. The code is available at https://github.com/deep-geo/MoE-NuSeg. |
| format | Article |
| id | doaj-art-5932682b35c24d3ab205a54c17adc7b7 |
| institution | Kabale University |
| issn | 1110-0168 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Alexandria Engineering Journal |
| spelling | doaj-art-5932682b35c24d3ab205a54c17adc7b72025-01-09T06:13:23ZengElsevierAlexandria Engineering Journal1110-01682025-01-01110557566MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts networkXuening Wu0Yiqing Shen1Qing Zhao2Yanlan Kang3Wenqiang Zhang4Shanghai Engineering Research Center of AI & Robotics, Academy for Engineering & Technology, Fudan University, Shanghai 200433, ChinaDepartment of Computer Science, Johns Hopkins University, Baltimore 21218, USAShanghai Engineering Research Center of AI & Robotics, Academy for Engineering & Technology, Fudan University, Shanghai 200433, ChinaShanghai Engineering Research Center of AI & Robotics, Academy for Engineering & Technology, Fudan University, Shanghai 200433, ChinaEngineering Research Center of AI & Robotics, Ministry of Education, Academy for Engineering & Technology, School of Computer Science, Fudan University, Shanghai 200433, China; Corresponding author.Accurate nuclei segmentation is essential for extracting quantitative information from histology images to support disease diagnosis and treatment decisions. However, precise segmentation is challenging due to the presence of clustered nuclei, varied morphologies, and the need to capture global spatial correlations. While state-of-the-art Transformer-based models employ tri-decoder architectures to decouple the segmentation task into nuclei, edges, and cluster edges segmentation, their complexity and long inference times hinder clinical integration. To address this, we introduce MoE-NuSeg, a novel Mixture of Experts (MoE) network that consolidates the tri-decoder into a single decoder. MoE-NuSeg employs three specialized experts for nuclei segmentation, edge delineation, and cluster edge detection, thereby mirroring the functionality of tri-decoders while surpassing their performance and reducing parameters by sharing attention heads. We propose a two-stage training strategy: the first stage independently trains the three experts, and the second stage fine-tunes their interactions to dynamically allocate the contributions of each expert using a learnable attention-based gating network. Evaluations across three datasets demonstrate that MoE-NuSeg outperforms the state-of-the-art methods, achieving an average increase of 0.99% in Dice coefficient, 1.14% in IoU and 0.92% in F1 Score, while reducing parameters by 30.1% and FLOPs by 40.2%. The code is available at https://github.com/deep-geo/MoE-NuSeg.http://www.sciencedirect.com/science/article/pii/S1110016824011669Nuclei segmentationHistology imageSwin transformerMixture of Experts (moEs) |
| spellingShingle | Xuening Wu Yiqing Shen Qing Zhao Yanlan Kang Wenqiang Zhang MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network Alexandria Engineering Journal Nuclei segmentation Histology image Swin transformer Mixture of Experts (moEs) |
| title | MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network |
| title_full | MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network |
| title_fullStr | MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network |
| title_full_unstemmed | MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network |
| title_short | MoE-NuSeg: Enhancing nuclei segmentation in histology images with a two-stage Mixture of Experts network |
| title_sort | moe nuseg enhancing nuclei segmentation in histology images with a two stage mixture of experts network |
| topic | Nuclei segmentation Histology image Swin transformer Mixture of Experts (moEs) |
| url | http://www.sciencedirect.com/science/article/pii/S1110016824011669 |
| work_keys_str_mv | AT xueningwu moenusegenhancingnucleisegmentationinhistologyimageswithatwostagemixtureofexpertsnetwork AT yiqingshen moenusegenhancingnucleisegmentationinhistologyimageswithatwostagemixtureofexpertsnetwork AT qingzhao moenusegenhancingnucleisegmentationinhistologyimageswithatwostagemixtureofexpertsnetwork AT yanlankang moenusegenhancingnucleisegmentationinhistologyimageswithatwostagemixtureofexpertsnetwork AT wenqiangzhang moenusegenhancingnucleisegmentationinhistologyimageswithatwostagemixtureofexpertsnetwork |