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

doi:10.3969/j.issn.1001-909X.2024.03.005

摘要/Abstract

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:

P714

邸雅楠, 赵若轩, 徐建洲. 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[J]. 海洋学研究, 2024, 42(3): 64-74.

DI Ya’nan, ZHAO Ruoxuan, XU Jianzhou. Deep learning-based histopathological analysis and its potential application in marine monitoring: A review and case study[J]. Journal of Marine Sciences, 2024, 42(3): 64-74.

图/表 6

图1 人类及水生生物组织病理图像

Fig.1 Histopathological images of human and aquatic organisms

图2 本文研究团队前期建立的贻贝消化腺和鳃组织的定量指标[67-68]

Fig.2 The quantitative indicators for digestive gland and gill tissues of mussels established by our team[67-68]

图3 贻贝鳃组织病理图像标注示意图

Fig.3 Schematic annotation of gill histopathologic images of marine mussels

表1 基于不同模型的图分割性能评价

Tab.1 Performance evaluation of image segmentation based on different models

评价指标	模型架构			骨架网络
评价指标	UNet	LinkNet	FPN	VGG16-UNet	ResNet34-UNet	EfficientNetB0-UNet
准确率	0.898 5±0.086 25	0.838 4±0.041 68	0.744 9±0.116 4	0.894 4±0.083 56	0.947 8±0.061 92	0.914 4±0.093 33
IoU	0.745 5±0.097 53	0.718 9±0.077 96	0.713 3±0.117 9	0.787 5±0.099 64	0.852 6±0.113 3	0.793 3±0.120 2
Dice系数	0.850 7±0.065 17	0.808 8±0.090 58	0.827 4±0.079 94	0.877 7±0.064 56	0.915 9±0.078 76	0.885 8±0.081 20
HD95	23.24±24.68	25.74±23.54	49.78±35.62	18.66±27.22	10.92±27.84	21.00±34.67

图4 基于不同分析方法的贻贝鳃丝的病理变化

Fig.4 The histopathological alterations in gills of marine mussels based on different analyzing methods

图5 基于深度学习的海洋生物组织病理图像分析工作流程示意图

Fig.5 Deep learning-based workflow for histopathological image analysis of marine organisms

参考文献 70

[1]	BOESCH D F, PAUL J F. An overview of coastal environ-mental health indicators[J]. Human and Ecological Risk Assessment: An International Journal, 2001, 7(5): 1409-1417.
[2]	GODSON P S, VINCENT S G T, KRISHNAKUMAR S. Ecology and biodiversity of benthos[M]. [S.l.]: Elsevier, 2022.
[3]	郭海峡, 蔡榕硕, 谭红建. 长江口及邻近海域浮游植物生态系统气候变化综合风险评估[J]. 应用海洋学学报, 2023, 42(4):549-558.
	GUO H X, CAI R S, TAN H J. Assessment of climate change on comprehensive risks of phytoplankton ecosystem in Changjiang Estuary and adjacent waters[J]. Journal of Applied Oceanography, 2023, 42(4): 549-558.
[4]	YUAN S Y, LI Y, BAO F W, et al. Marine environmental monitoring with unmanned vehicle platforms: Present applications and future prospects[J]. Science of the Total Environment, 2023, 858: 159741.
[5]	CHEN S Y, QU M J, DING J W, et al. BaP-metals co-exposure induced tissue-specific antioxidant defense in marine mussels Mytilus coruscus[J]. Chemosphere, 2018, 205: 286-296.
[6]	GU Y Y, WEI Q, WANG L Y, et al. A comprehensive study of the effects of phthalates on marine mussels: Bioconcentration, enzymatic activities and metabolomics[J]. Marine Pollution Bulletin, 2021, 168: 112393.
[7]	VON MOOS N, BURKHARDT-HOLM P, KÖHLER A. Uptake and effects of microplastics on cells and tissue of the blue mussel Mytilus edulis L. after an experimental exposure[J]. Environmental Science & Technology, 2012, 46(20): 11327-11335.
[8]	REY-CAMPOS M, MOREIRA R, VALENZUELA-MUÑOZ V, et al. High individual variability in the transcriptomic response of Mediterranean mussels to Vibrio reveals the involvement of myticins in tissue injury[J]. Scientific Reports, 2019, 9:3569.
[9]	ROCHA T L, SABóIA-MORAIS S M T, BEBIANNO M J. Histopathological assessment and inflammatory response in the digestive gland of marine mussel Mytilus galloprovincialis exposed to cadmium-based quantum dots[J]. Aquatic Toxicology, 2016, 177: 306-315.
[10]	BRABY C E, SOMERO G N. Following the heart: Tem-perature and salinity effects on heart rate in native and invasive species of blue mussels (genus Mytilus)[J]. Journal of Experimental Biology, 2006, 209: 2554-2566.
[11]	SARÀ G, DE PIRRO M. Heart beat rate adaptations to varying salinity of two intertidal Mediterranean bivalves: The invasive Brachidontes pharaonis and the native Mytilaster minimus[J]. Italian Journal of Zoology, 2011, 78(2): 193-197.
[12]	ANANDRAJ A, MARSHALL D J, GREGORY M A, et al. Metal accumulation, filtration and O₂ uptake rates in the mussel Perna perna (Mollusca: Bivalvia) exposed to Hg²⁺, Cu²⁺ and Zn²⁺[J]. Comparative Biochemistry and Physiology Part C: Toxicology & Pharmacology, 2002, 132(3): 355-363.
[13]	SOLÉ M, PORTE C, BARCELO D, et al. Bivalves residue analysis for the assessment of coastal pollution in the Ebro Delta (NW Mediterranean)[J]. Marine Pollution Bulletin, 2000, 40(9): 746-753.
[14]	WARD J E, ZHAO S, HOLOHAN B A, et al. Selective ingestion and egestion of plastic particles by the blue mussel (Mytilus edulis) and eastern oyster (Crassostrea virginica): Implications for using bivalves as bioindicators of micro-plastic pollution[J]. Environmental Science & Technology, 2019, 53(15): 8776-8784.
[15]	SALEM AL-HOWITI N, OUANES BEN OTHMEN Z, BEN OTHMANE A, et al. Use of Tridacna maxima, a bivalve in the biomonitoring of the Saudi Arabian Red Sea coast[J]. Marine Pollution Bulletin, 2020, 150: 110766.
[16]	DITRIA E M, BUELOW C A, GONZALEZ-RIVERO M, et al. Artificial intelligence and automated monitoring for assisting conservation of marine ecosystems: A perspective[J]. Frontiers in Marine Science, 2022, 9: 918104.
[17]	SONG T, PANG C, HOU B Y, et al. A review of artificial intelligence in marine science[J]. Frontiers in Earth Science, 2023, 11: 1090185.
[18]	GONZÁLEZ-RIVERO M, BEIJBOM O, RODRIGUEZ-RAMIREZ A, et al. Monitoring of coral reefs using artificial intelligence: A feasible and cost-effective approach[J]. Remote Sensing, 2020, 12(3): 489.
[19]	SHEN D G, WU G R, SUK H I. Deep learning in medical image analysis[J]. Annual Review of Biomedical Engineering, 2017, 19: 221-248. DOI PMID
[20]	FUCHS T J, BUHMANN J M. Computational pathology: Challenges and promises for tissue analysis[J]. Compu-terized Medical Imaging and Graphics, 2011, 35(7/8): 515-530.
[21]	PROCTOR I E. What is histopathology and how to get the most out of your histopathologist[J]. British Journal of Hospital Medicine, 2015, 76(5): 66-68. DOI PMID
[22]	ADYANTHAYA S, JOSE M. Quality and safety aspects in histopathology laboratory[J]. Journal of Oral and Maxillo-facial Pathology, 2013, 17(3): 402-407.
[23]	FANTA E, RIOS F S, ROMÃO S, et al. Histopathology of the fish Corydoras paleatus contaminated with sublethal levels of organophosphorus in water and food[J]. Ecotoxi-cology and Environmental Safety, 2003, 54(2): 119-130.
[24]	HSIEH S L, HSIEH S, XU R Q, et al. Toxicological effects of polystyrene nanoplastics on marine organisms[J]. Envi-ronmental Technology & Innovation, 2023, 30: 103073.
[25]	GEYER H, SHEEHAN P, KOTZIAS D, et al. Prediction of ecotoxicological behaviour of chemicals: Relationship between physico-chemical properties and bioaccumulation of organic chemicals in the mussel Mytilus edulis[J]. Chemo-sphere, 1982, 11(11): 1121-1134.
[26]	LI J N, LUSHER A L, ROTCHELL J M, et al. Using mussel as a global bioindicator of coastal microplastic pollution[J]. Environmental Pollution, 2019, 244: 522-533. DOI PMID
[27]	SMITH J R, FONG P, AMBROSE R F. Dramatic declines in mussel bed community diversity: Response to climate change?[J]. Ecology, 2006, 87(5): 1153-1161. PMID
[28]	张翼飞, 曲梦杰, 丁家玮, 等. 多环芳烃对海洋贝类多生物水平毒性效应的研究进展[J]. 生态毒理学报, 2019, 14(1):18-29.
	ZHANG Y F, QU M J, DING J W, et al. Ecotoxicology: A review of multi-toxicity in marine bivalve induced by polycyclic aromatic hydrocarbons[J]. Asian Journal of Ecotoxicology, 2019, 14(1): 18-29.
[29]	AARAB N, MINIER C, LEMAIRE S, et al. Biochemical and histological responses in mussel (Mytilus edulis) exposed to North Sea oil and to a mixture of North Sea oil and alkylphenols[J]. Marine Environmental Research, 2004, 58(2/3/4/5): 437-441.
[30]	孟祥敬, 李斐, 王晓晴, 等. 石墨烯联合磷酸三苯酯(TPP)胁迫紫贻贝(Mytilus galloprovincialis)的生理生化响应[J]. 科学通报, 2020, 65(16):1599-1609.
	MENG X J, LI F, WANG X Q, et al. Physiological and biochemistry responses of graphene combined with triphenyl phosphate (TPP) to Mytilus galloprovincialis[J]. Chinese Science Bulletin, 2020, 65(16): 1599-1609.
[31]	MATOZZO V, ERCOLINI C, SERRACCA L, et al. Assessing the health status of farmed mussels (Mytilus galloprovincialis) through histological, microbiological and biomarker analyses[J]. Journal of Invertebrate Pathology, 2018, 153: 165-179. DOI PMID
[32]	WOLF J C, MAACK G. Evaluating the credibility of histo-pathology data in environmental endocrine toxicity studies[J]. Environmental Toxicology and Chemistry, 2017, 36(3): 601-611.
[33]	BASTI L, ENDO M, SEGAWA S, et al. Prevalence and intensity of pathologies induced by the toxic dinoflagellate, Heterocapsa circularisquama, in the Mediterranean mussel, Mytilus galloprovincialis[J]. Aquatic Toxicology, 2015, 163: 37-50.
[34]	CUEVAS N, ZORITA I, COSTA P M, et al. Development of histopathological indices in the digestive gland and gonad of mussels: Integration with contamination levels and effects of confounding factors[J]. Aquatic Toxicology, 2015, 162: 152-164. DOI PMID
[35]	COPPOLA F, BESSA A, HENRIQUES B, et al. Oxidative stress, metabolic and histopathological alterations in mussels exposed to remediated seawater by GO-PEI after contamination with mercury[J]. Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology, 2020, 243: 110674.
[36]	BERNET D, SCHMIDT H, MEIER W, et al. Histopathology in fish: Proposal for a protocol to assess aquatic pollution[J]. Journal of Fish Diseases, 1999, 22(1): 25-34.
[37]	CARELLA F, FEIST S W, BIGNELL J P, et al. Comparative pathology in bivalves: Aetiological agents and disease processes[J]. Journal of Invertebrate Pathology, 2015, 131: 107-120. DOI PMID
[38]	李太伟. 海洋贻贝组织病理指标的深入挖掘与应用[D]. 杭州: 浙江大学, 2021.
	LI T W. Deep investigation and application of histopatho-logical indices in marine mussels[D]. Hangzhou: Zhejiang University, 2021.
[39]	OECD. Test No. 487: In vitro mammalian cell micronucleus test[M]. 2023.
[40]	LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521: 436-444.
[41]	SARKER I H. Deep learning: A comprehensive overview on techniques, taxonomy, applications and research directions[J]. SN Computer Science, 2021, 2(6): 420.
[42]	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[43]	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA, 2016.
[44]	BI J R, ZHU Z L, MENG Q L. Transformer in computer vision[C]// 2021 IEEE International Conference on Computer Science, Electronic Information Engineering and Intelligent Control Technology (CEI). Fuzhou, China, 2021.
[45]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]// Proceedings of the International Conference on Neural Information Processing Systems (NIPS), Long Beach, USA, 2017.
[46]	HUSSAIN M. YOLO-v1 to YOLO-v8, the rise of YOLO and its complementary nature toward digital manufacturing and industrial defect detection[J]. Machines, 2023, 11(7): 677.
[47]	CIOCAN C M, CUBERO-LEON E, PUINEAN A M, et al. Effects of estrogen exposure in mussels, Mytilus edulis, at different stages of gametogenesis[J]. Environmental Pollution, 2010, 158(9): 2977-2984.
[48]	LEISER F, RANK S, SCHMIDT-KRAEPELIN M, et al. Medical informed machine learning: A scoping review and future research directions[J]. Artificial Intelligence in Medicine, 2023, 145: 102676.
[49]	YEARLEY A G, GOEDMAKERS C M W, PANAHI A, et al. FDA-approved machine learning algorithms in neuroradiology: A systematic review of the current evidence for approval[J]. Artificial Intelligence in Medicine, 2023, 143: 102607.
[50]	IIZUKA O, KANAVATI F, KATO K, et al. Deep learning models for histopathological classification of gastric and colonic epithelial tumours[J]. Scientific Reports, 2020, 10(1): 1504.
[51]	GERTYCH A, SWIDERSKA-CHADAJ Z, MA Z X, et al. Convolutional neural networks can accurately distinguish four histologic growth patterns of lung adenocarcinoma in digital slides[J]. Scientific Reports, 2019, 9: 1483. DOI PMID
[52]	FENG Y Q, ZHANG L, MO J. Deep manifold preserving autoencoder for classifying breast cancer histopathological images[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2020, 17(1): 91-101. DOI PMID
[53]	LITJENS G, KOOI T, BEJNORDI B E, et al. A survey on deep learning in medical image analysis[J]. Medical Image Analysis, 2017, 42: 60-88. DOI PMID
[54]	RONNEBERGER O, FISCHER P, BROX T. U-net: Convolutional networks for biomedical image segmentation[C]//Proceedings of the Medical Image Computing and Computer-Assisted Intervention. Munich, Germany, 2015.
[55]	ZHOU Z W, SIDDIQUEE M M R, TAJBAKHSH N, et al. UNet++: A nested U-net architecture for medical image segmentation[C]// Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: 4th International Workshop. Granada, Spain, 2018.
[56]	GRAHAM S, CHEN H, GAMPER J, et al. MILD-Net: Minimal information loss dilated network for gland instance segmentation in colon histology images[J]. Medical Image Analysis, 2019, 52: 199-211. DOI PMID
[57]	HERMSEN M, DE BEL T, DEN BOER M, et al. Deep learning-based histopathologic assessment of kidney tissue[J]. Journal of the American Society of Nephrology, 2019, 30(10): 1968-1979. DOI PMID
[58]	SONG Z G, ZOU S M, ZHOU W X, et al. Clinically applicable histopathological diagnosis system for gastric cancer detection using deep learning[J]. Nature Commu-nications, 2020, 11: 4294.
[59]	GRAHAM S, VU Q D, JAHANIFAR M, et al. One model is all you need: Multi-task learning enables simultaneous histology image segmentation and classification[J]. Medical Image Analysis, 2023, 83: 102685.
[60]	RIJTHOVEN M V, SWIDERSKA-CHADAJ Z, SEELIGER K, et al. You only look on lymphocytes once[C]// The Inter-national Conference on Medical Imaging with Deep Learning, Amsterdam, Netherlands, 2018.
[61]	RONG R C, SHENG H, JIN K W, et al. A deep learning approach for histology-based nucleus segmentation and tumor microenvironment characterization[J]. Modern Pathology, 2023, 36(8): 100196.
[62]	CRUZ-ROA A, GILMORE H, BASAVANHALLY A, et al. Accurate and reproducible invasive breast cancer detection in whole-slide images: A deep learning approach for quantifying tumor extent[J]. Scientific Reports, 2017, 7: 46450.
[63]	KARIMI D, NIR G, FAZLI L, et al. Deep learning-based gleason grading of prostate cancer from histopathology images-role of multiscale decision aggregation and data augmentation[J]. IEEE Journal of Biomedical and Health Informatics, 2020, 24(5): 1413-1426. DOI PMID
[64]	SIRINUKUNWATTANA K, PLUIM J P W, CHEN H, et al. Gland segmentation in colon histology images: The glas challenge contest[J]. Medical Image Analysis, 2017, 35: 489-502. DOI PMID
[65]	VAN DER OOST R, BEYER J, VERMEULEN N P E. Fish bioaccumulation and biomarkers in environmental risk assessment: A review[J]. Environmental Toxicology and Pharmacology, 2003, 13(2): 57-149. DOI PMID
[66]	YOUNIS E, ABDELWARITH D A, AL-ASGAH N, et al. Histological changes in the liver and intestine of Nile tilapia, Oreochromis niloticus, exposed to sublethal concentrations of cadmium[J]. Pakistan Journal of Zoology, 2013, 45: 833-841.
[67]	QU M J, XU J Z, YANG Y L, et al. Assessment of sulfamethoxazole toxicity to marine mussels (Mytilus galloprovincialis): Combine p38-MAPK signaling pathway modulation with histopathological alterations[J]. Ecoto-xicology and Environmental Safety, 2023, 249: 114365.
[68]	QU M J, DING J W, WANG Y, et al. Genetic impacts induced by BaP and Pb in Mytilus coruscus: Can RAPD be a validated tool in genotoxicity evaluation both in vivo and in vitro?[J]. Ecotoxicology and Environmental Safety, 2019, 169: 529-538.
[69]	BINELLI A, RIVA C, COGNI D, et al. Assessment of the genotoxic potential of benzo(α)pyrene and pp’-dichlorodi-phenyldichloroethylene in Zebra mussel (Dreissena polymor-pha)[J]. Mutation Research, 2008, 649(1/2):135-145.
[70]	DI Y N, SCHROEDER D C, HIGHFIELD A, et al. Tissue-specific expression of p53 and ras genes in response to the environmental genotoxicant benzo(α)pyrene in marine mussels[J]. Environmental Science & Technology, 2011, 45(20): 8974-8981.