Journal of Zhejiang University

Bio-Design and Manufacturing 2026 Vol.9 No.4 P.796 - 808

OrganoidViT: a vision transformer-based deep learning model for segmenting viable and nonviable organoids in bright-field images

Author(s): Yijun Liu,Dingyuan Yu,Guoxiang Fu,Zhangjie Li,Chenyang Zhou,Jiaqi Xu,Ning Zhao,Lian Xuan,Xiaolin Wang
Affiliation(s): 1. School of Integrated Circuits (School of Information Science and Electronic Engineering), Shanghai Jiao Tong University, Shanghai 200240, China more
Corresponding email(s): xlwang83@sjtu.edu.cn
Key Words: Organoids, Artificial intelligence (AI), Deep learning, Vision transformer, Image segmentation

Share this article to： More <<< Previous Article \|

Yijun Liu. OrganoidViT: a vision transformer-based deep learning model for segmenting viable and nonviable organoids in bright-field images[J]. Journal of Zhejiang University Science D, 2026, 9(4): 796 - 808.

@article{title="OrganoidViT: a vision transformer-based deep learning model for segmenting viable and nonviable organoids in bright-field images",
author="Yijun Liu",
journal="Journal of Zhejiang University Science D",
volume="9",
number="4",
pages="796 - 808",
year="2026",
publisher="Zhejiang University Press & Springer",
doi="10.1631/bdm.2500438"
}

%0 Journal Article
%T OrganoidViT: a vision transformer-based deep learning model for segmenting viable and nonviable organoids in bright-field images
%A Yijun Liu
%J Journal of Zhejiang University SCIENCE D
%V 9
%N 4
%P 796 - 808
%@ 1869-1951
%D 2026
%I Zhejiang University Press & Springer
%DOI 10.1631/bdm.2500438

TY - JOUR
T1 - OrganoidViT: a vision transformer-based deep learning model for segmenting viable and nonviable organoids in bright-field images
A1 - Yijun Liu
J0 - Journal of Zhejiang University Science D
VL - 9
IS - 4
SP - 796
EP - 808
%@ 1869-1951
Y1 - 2026
PB - Zhejiang University Press & Springer
ER -
DOI - 10.1631/bdm.2500438

Abstract
Chinese Summary
Academic Network
Reviewer Comment

Abstract: The viability assessment of patient-derived tumor organoids is essential for preclinical drug screening, with microscopic imaging serving as a key method for evaluating drug effects. Traditional image analysis methods, such as manual evaluation and fluorescence staining, suffer from low efficiency, dye toxicity, and fluorescence degradation, making them unsuitable for high-throughput drug screening. Additionally, current artificial intelligence (AI)-based tools face challenges in precise segmentation for the accurate quantitative analysis of viable and nonviable organoids. To address these challenges, we introduce OrganoidViT, a novel deep learning model utilizing vision transformer technology for the precise segmentation of viable and nonviable colorectal cancer organoids in bright-field microscopy images, ensuring an accurate efficacy assay of different drugs. Trained on custom datasets of colorectal cancer organoids prepared using microdroplet technology, OrganoidViT facilitates high-throughput segmentation without fluorescence imaging, with experimental results indicating superior performance of OrganoidViT (accuracy of 99.7%) compared to manual annotation and traditional convolutional models. The capability of pixel-level segmentation enables accurate morphological assessments, including area, pellucidity, roundness, and grayscale entropy, which are critical for evaluating growth conditions and the effects of different therapeutic agents on colorectal cancer organoids. Thus, OrganoidViT shows promise for preclinical drug screening, enhancing both the efficiency and accuracy of organoid-based testing.

OrganoidViT：一种基于视觉 Transformer 的明场类器官活性与非活性分割深度学习模型

患者来源肿瘤类器官的活力评估是临床前药物筛选的关键环节，而显微成像是评价药物作用效果的核心手段。传统图像分析方法如人工评估、荧光染色等存在效率低下、染料具有细胞毒性以及荧光随着培养时间逐渐减弱等缺陷，难以满足高通量药物筛选的需求。此外，现有人工智能分析工具在实现类器官精准分割方面仍面临较大挑战。为应对上述挑战，本研究构建了 OrganoidViT，这是一种基于视觉 Transformer 的新型深度学习模型，可对明场显微图像中的活性与非活性结直肠癌类器官实现精准分割，为不同药物的疗效评估提供可靠支撑。OrganoidViT 模型基于微流控液滴技术构建的结直肠癌类器官专用数据集进行训练，可在无荧光标记条件下完成高通量图像分割。实验结果表明，相较于人工标注以及传统卷积神经网络模型，OrganoidViT 模型表现出更优异的分割性能，分割准确率高达 99.7%。OrganoidViT 的像素级精准分割能力可实现对类器官面积、透明度、圆度及灰度熵等关键形态学参数的精确量化分析，这对于评估类器官生长状态及不同治疗药物对结直肠癌类器官的影响具有重要意义。因此，OrganoidViT 在临床前药物筛选中展现出广阔的应用前景，能够显著提升类器官功能检测的效率与准确性。
关键词：类器官；人工智能；深度学习；视觉 transformer；图像分割