海洋学研究 ›› 2024, Vol. 42 ›› Issue (3): 64-74.DOI: 10.3969/j.issn.1001-909X.2024.03.005

• 研究综述 • 上一篇    下一篇

基于深度学习的生物组织病理图像分析在海洋监测中的发展潜力及案例分析

邸雅楠1,2(), 赵若轩2, 徐建洲1,2   

  1. 1.浙江大学 海南研究院,海南 三亚 572025
    2.浙江大学 海洋学院,浙江 舟山 316021
  • 收稿日期:2023-12-30 修回日期:2024-07-12 出版日期:2024-09-15 发布日期:2024-11-25
  • 作者简介:邸雅楠(1980—),女,甘肃省兰州市人,副教授,主要从事生态毒理学研究,E-mail:diyanan@zju.edu.cn
  • 基金资助:
    国家自然科学基金项目(42230404);国家自然科学基金项目(41976129)

Deep learning-based histopathological analysis and its potential application in marine monitoring: A review and case study

DI Ya’nan1,2(), ZHAO Ruoxuan2, XU Jianzhou1,2   

  1. 1. Hainan Institute, Zhejiang University, Sanya 572025, China
    2. Ocean College, Zhejiang University, Zhoushan 316021, China
  • Received:2023-12-30 Revised:2024-07-12 Online:2024-09-15 Published:2024-11-25

摘要:

生物组织病理指标可用于评价海洋生物健康,但在应用中存在效率低、成本高、主观性强等缺陷。将人工智能技术引入生物组织病理分析,可以发挥其高通量的图像分析优势,突破其在海洋生物健康评价和监测中的应用限制。该文通过对海洋生物组织健康评价指标、人工智能技术的图像分析应用以及利用人工智能开展组织病理图像处理的文献调研,提出基于深度学习的海洋动物组织病理图像分析思路,并以海洋贻贝作为模式生物进行技术开发。经过对贻贝鳃组织病理影像数据的训练、验证和预测等过程,确定Res-UNet深度学习模型可对贻贝在典型环境污染物胁迫下的病理损伤进行高效、准确定量,构建了一种能够自动化、高通量和弱主观性地分析海洋贻贝组织病理影像的工作流程,为海洋生物健康评价、海洋监测提供新思路与新技术。

关键词: 人工智能, 神经网络, 病理图像处理, 生物健康评价, 海洋模式生物, 海洋贻贝, 组织病理定量, 鳃丝面积

Abstract:

Histopathological indexes can be used to assess health of organism, but their application faces challenges such as low efficiency, high cost and strong subjectivity. Introducing artificial intelligence (AI) technology into histopathological analysis of biological tissues can leverage its high-throughput image analysis capabilities, overcoming the limitations in assessing and monitoring marine organism health. Based on our review on health assessment indicators of marine organisms, the application of AI in image analysis, and the use of AI for histopathological image processing, a deep learning-based histopathological image analysis approach was proposed using gill tissues of marine mussels as representative. Through a series of processes such as training, validation, and prediction of histopathology images, it was determined that the Res-UNet deep learning model can efficiently and accurately quantify histopathological damage in mussels’ gills. An automated, high-throughput, and less subjective workflow based on deep learning was finally established, offering new ideas and techniques for marine organism health assessment and marine monitoring.

Key words: artificial intelligence, neural networks, histopathological image processing, biological health assessment, marine model organisms, marine mussels, quantitative histopathology, gill filament area

中图分类号: