1.徐州医科大学法医学教研室,江苏 徐州 221000
2.司法鉴定科学研究院//上海市法医学重点实验室//司法部司法鉴定重点实验室//上海市司法鉴定专业技术服务平台,上海 200063
3.西安市公安局经济技术开发区分局刑事侦查大队,陕西 西安 710018
4.苏州市公安局,江苏 苏州 215000
于慧潇,硕士生,研究方向:法医病理学及人工智能算法,E-mail: yuhuixiao19552153521@163.com
纸质出版日期:2023-05-20,
收稿日期:2022-10-25,
扫 描 看 全 文
于慧潇,朱永正,赵天琦等.酶消化法结合人工智能技术在法医学溺死硅藻检验中的应用[J].中山大学学报(医学科学版),2023,44(03):430-438.
YU Hui-xiao,ZHU Yong-zheng,ZHAO Tian-qi,et al.Enzymatic Digestion Method Coupled with Artificial Intelligence Techniques in Forensic Drowning Diatom Detection[J].Journal of Sun Yat-sen University(Medical Sciences),2023,44(03):430-438.
于慧潇,朱永正,赵天琦等.酶消化法结合人工智能技术在法医学溺死硅藻检验中的应用[J].中山大学学报(医学科学版),2023,44(03):430-438. DOI: 10.13471/j.cnki.j.sun.yat-sen.univ(med.sci).2023.0309.
YU Hui-xiao,ZHU Yong-zheng,ZHAO Tian-qi,et al.Enzymatic Digestion Method Coupled with Artificial Intelligence Techniques in Forensic Drowning Diatom Detection[J].Journal of Sun Yat-sen University(Medical Sciences),2023,44(03):430-438. DOI: 10.13471/j.cnki.j.sun.yat-sen.univ(med.sci).2023.0309.
目的
2
人工智能(AI)全涂片自动化硅藻检验技术能比人类专家更快速、高效进行法医病理学溺死硅藻检验。然而,该技术仅与硅藻提取率较低的硝酸消化法联用过,本研究拟采用更加高效的蛋白酶K组织消解法(以下简称酶消化法)作为硅藻提取方法,探究该技术在其他硅藻提取方法中的泛化能力及可行性。
方法
2
收集6例溺死尸体的肺组织进行蛋白酶K消解并制成涂片,利用数字化图像矩阵切割方法将涂片进行数字化处理并据此建立硅藻-背景数据库,将数据集按照3:1:1的比例分为训练集、验证集和测试集,在ImageNet预训练基础上对卷积神经网络(CNN)模型进行训练和内部验证及外部测试。
结果
2
结果显示最佳模型外部测试的准确率达97.65%,且模型特征提取区域即为硅藻所在区域。实际应用中最佳的CNN模型对溺水尸体的硅藻检测精准率高达80 %以上。
结论
2
研究表明,基于CNN模型的AI自动化硅藻检验技术和酶消化法联用能高效识别硅藻,可以作为溺水鉴定中硅藻检测的辅助方法。
Objective
2
Artificial intelligence (AI) full smear automated diatom detection technology can perform forensic pathology drowning diatom detection more quickly and efficiently than human experts.However, this technique was only used in conjunction with the strong acid digestion method, which has a low extraction rate of diatoms. In this study, we propose to use the more efficient proteinase K tissue digestion method (hereinafter referred to as enzyme digestion method) as a diatom extraction method to investigate the generalization ability and feasibility of this technique in other diatom extraction methods.
Methods
2
Lung tissues from 6 drowned cadavers were collected for proteinase K ablation and made into smears, and the smears were digitized using the digital image matrix cutting method and a diatom and background database was established accordingly.The data set was divided into training set, validation set and test set in the ratio of 3:1:1, and the convolutional neural network (CNN) models were trained, internally validated, and externally tested on the basis of ImageNet pre-training.
Results
2
The results showed that the accuracy rate of the external test of the best model was 97.65 %, and the area where the model features were extracted was the area where the diatoms were located. The best CNN model in practice had a precision of more than 80 % for diatom detection of drowned corpses.
Conclusion
2
It is shown that the AI automated diatom detection technique based on CNN model and enzymatic digestion method in combination can efficiently identify diatoms and can be used as an auxiliary method for diatom detection in drowning identification.
法医病理学硅藻溺死蛋白酶K人工智能卷积神经网络
forensic pathologydiatomsdrowningproteinase Kartificial intelligenceconvolutional neural network
World health organization global report on drowning: preventing a leading killer[J]. WHO, 2014.
Lunetta P, Penttila A, Hallfors G. Scanning and transmission electron microscopical evidence of the capacity of diatoms to penetrate the alveolo-capillary barrier in drowning[J]. Int J Legal Med, 1998, 111(5): 229-237.
Piette MH, De-letter EA. Drowning: still a difficult autopsy diagnosis[J]. Forensic Sci Int, 2006, 163(1/2): 1-9.
Ming M, Meng X, Wang E. Evaluation of four digestive methods for extracting diatoms[J]. Forensic Sci Int, 2007, 170(1): 29-34.
Pedraza A, Bueno G, Deniz O, et al. Lights and pitfalls of convolutional neural networks for diatom identification[C]. SPIE, 2018, 10679: 88-96.
Pedraza A, Bueno G, Deniz O, et al. Automated diatom classification (part B): a deep learning approach[J]. Appl Sci, 2017, 7(5): 460.
Salido J, Sánchez C, Ruiz-santaquiteria J, et al. A low-cost automated digital microscopy platform for automatic identification of diatoms[J]. Appl Sci, 2020, 10(17): 6033.
Krause LMK, Koc J, Rosenhahn B, et al. Fully convolutional neural network for detection and counting of diatoms on coatings after short-term field exposure[J]. Environ Sci Technol, 2020, 54(16): 10022-10030.
Zhou Y, Zhang J, Huang J, et al. Digital whole-slide image analysis for automated diatom test in forensic cases of drowning using a convolutional neural network algorithm[J]. Forensic Sci Int, 2019, 302: 109922.
Zhang J, Zhou Y, Vieira DN, et al. An efficient method for building a database of diatom populations for drowning site inference using a deep learning algorithm[J]. Int J Legal Med, 2021, 135(3): 817-827.
Zhou Y, Cao Y, Huang J, et al. Research advances in forensic diatom testing[J]. Forensic Sci Res, 2020, 5(2): 98-105.
周圆圆, 曹永杰, 杨越, 等. 基于人工智能硅藻自动化识别系统的实际案例应用[J]. 法医学杂志, 2020, 36(2): 239-242.
Zhou YY, Cao YJ, Yang Y, et al. Application of artificial intelligence automatic diatom identification system in practical cases[J]. J Forensic Med, 2020, 36(2): 239-242.
Ludes B, Coste M, North N, et al. Diatom analysis in victim' s tissues as an indicator of the site of drowning[J]. Int J Legal Med, 1999, 112(3): 163-166.
Coelho S, Ramos P, Ribeiro C, et al. Contribution to the determination of the place of death by drowning- A study of diatoms' biodiversity in douro river estuary[J]. J Forensic Leg Med, 2016, 41: 58-64.
Carballeira R, Vieira DN, Febrero-bande M, et al. A valid method to determine the site of drowning[J]. Int J Legal Med, 2018, 132(2): 487-497.
Szegedy C, Ioffe S, Vanhoucke V, et al. Inception-v4, inception-resnet and the impact of residual connections on learning[C]. AAAI, 2017: 4278-4284.
朱永正, 张吉, 程奇, 等. 4种深度学习图像分类算法在人工智能硅藻检验中的比较[J]. 法医学杂志, 2022, 38(1): 31-39.
Zhu YZ, Zhang J, Cheng Q, et al. Comparison among four deep learning image classification algorithms in AI-based diatom test[J]. J Forensic Med, 2022, 38(1): 31-39.
Deng J, Dong W, Socher R, et al. ImageNet: a large-scale hierarchical image database[C]. IEEE, 2009: 248-255. http://dx.doi.org/10.1109/CVPR.20 09.520 6848http://dx.doi.org/10.1109/CVPR.2009.5206848
Zeiler MD. Adadelta: an adaptive learning rate method[C]. ArXiv, 2012. https://arxiv.org/abs/1212.5701https://arxiv.org/abs/1212.5701
Simonyan K, Vedaldi A, Zisserman A. Deep inside convolutional networks: visualising image classification models and saliency maps[J]. CoRR, 2014. https://arxiv.org/abs/1312.6034https://arxiv.org/abs/1312.6034
Zeiler MD, Fergus R. Visualizing and understanding convolutional networks[M]. Springer, 2014: 818-833. https://doi.org/10.1007/978-3-319-10590-1_53https://doi.org/10.1007/978-3-319-10590-1_53
Bach S, Binder A, Montavon G, et al. On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation[J]. PloS One, 2015, 10(7): e0130140.
Selvaraju RR, Cogswell M, Das A, et al. Grad-cam: visual explanations from deep networks via gradient-based localization [C]. IEEE, 2017: 618-626.
余海星, 张吉, 朱永正, 等. 1991-2020年法医学溺死领域中英文文献计量学分析[J]. 法医学杂志, 2022, 38(1): 98-109.
Yu HS, Zhang J, Zhu YC, et al. Bibliometrics analysis in English and Chinese literature on drowning in forensic medicine from 1991 to 2020[J]. J Forensic Med, 2022, 38(1): 98-109.
Jacomy M, Venturini T, Heymann S, et al. Forceatlas2, a continuous graph layout algorithm for handy network visualization designed for the gephi software [J]. PloS one, 2014, 9(6): e98679.
Yu W, Xue Y, Knoops R, et al. Automated diatom searching in the digital scanning electron microscopy images of drowning cases using the deep neural networks[J]. Int J Legal Med, 2021, 135(2): 497-508.
陈吉, 刘晓荣, 杨佳雯, 等. 基于YOLOv3的硅藻电子显微镜图像识别模型的构建及应用[J]. 法医学杂志, 2022, 38(1): 46-52.
Chen J, Liu XR, Yang JW, et al. Construction and application of YOLOv3-based diatom identification model of scanning electron microscope images[J]. J Forensic Med, 2022, 38(1): 46-52.
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构