• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

一种基于YOLOv5框架的改进型硅藻自动检测方法及其在法医硅藻检验中分类识别的初步研究

An improved automated diatom detection method based on YOLOv5 framework and its preliminary study for taxonomy recognition in the forensic diatom test.

作者信息

Yu Weimin, Xiang Qingqing, Hu Yingchao, Du Yukun, Kang Xiaodong, Zheng Dongyun, Shi He, Xu Quyi, Li Zhigang, Niu Yong, Liu Chao, Zhao Jian

机构信息

Jiangsu JITRI Sioux Technologies Co., Ltd., Suzhou, China.

School of Forensic Medicine, Kunming Medical University, Kunming, China.

出版信息

Front Microbiol. 2022 Aug 19;13:963059. doi: 10.3389/fmicb.2022.963059. eCollection 2022.

DOI:10.3389/fmicb.2022.963059
PMID:36060761
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC9437702/
Abstract

The diatom test is a forensic technique that can provide supportive evidence in the diagnosis of drowning but requires the laborious observation and counting of diatoms using a microscopy with too much effort, and therefore it is promising to introduce artificial intelligence (AI) to make the test process automatic. In this article, we propose an artificial intelligence solution based on the YOLOv5 framework for the automatic detection and recognition of the diatom genera. To evaluate the performance of this AI solution in different scenarios, we collected five lab-grown diatom genera and samples of some organic tissues from drowning cases to investigate the potential upper/lower limits of the capability in detecting the diatoms and recognizing their genera. Based on the study of the article, a recall score of 0.95 together with the corresponding precision score of 0.9 were achieved on the samples of the five lab-grown diatom genera cross-validation, and the accuracy of the evaluation in the cases of kidney and liver is above 0.85 based on the precision and recall scores, which demonstrate the effectiveness of the AI solution to be used in drowning forensic routine.

摘要

硅藻检验是一种法医技术，可在溺水诊断中提供辅助证据，但需要通过显微镜费力地观察和计数硅藻，因此引入人工智能（AI）使检验过程自动化很有前景。在本文中，我们提出了一种基于YOLOv5框架的人工智能解决方案，用于硅藻属的自动检测和识别。为了评估该人工智能解决方案在不同场景下的性能，我们收集了五种实验室培养的硅藻属以及一些溺水案例中的有机组织样本，以研究检测硅藻及其属的能力的潜在上限/下限。基于本文的研究，在对五种实验室培养的硅藻属样本进行交叉验证时，召回率达到了0.95，相应的精确率为0.9，并且基于精确率和召回率，肾脏和肝脏案例中的评估准确率高于0.85，这证明了该人工智能解决方案在溺水法医常规工作中使用的有效性。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/f0c8527417b4/fmicb-13-963059-g0009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/28f3e7a7594a/fmicb-13-963059-g0001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/0ea23d136bb0/fmicb-13-963059-g0002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/037eea58433e/fmicb-13-963059-g0003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/0e503692da7e/fmicb-13-963059-g0004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/97bcfd766dd4/fmicb-13-963059-g0005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/0d4b1b54fc3c/fmicb-13-963059-g0006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/c341b1912c29/fmicb-13-963059-g0007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/03c96f360f07/fmicb-13-963059-g0008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/f0c8527417b4/fmicb-13-963059-g0009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/28f3e7a7594a/fmicb-13-963059-g0001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/0ea23d136bb0/fmicb-13-963059-g0002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/037eea58433e/fmicb-13-963059-g0003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/0e503692da7e/fmicb-13-963059-g0004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/97bcfd766dd4/fmicb-13-963059-g0005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/0d4b1b54fc3c/fmicb-13-963059-g0006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/c341b1912c29/fmicb-13-963059-g0007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/03c96f360f07/fmicb-13-963059-g0008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/5c37/9437702/f0c8527417b4/fmicb-13-963059-g0009.jpg

相似文献

1
An improved automated diatom detection method based on YOLOv5 framework and its preliminary study for taxonomy recognition in the forensic diatom test.一种基于YOLOv5框架的改进型硅藻自动检测方法及其在法医硅藻检验中分类识别的初步研究
Front Microbiol. 2022 Aug 19;13:963059. doi: 10.3389/fmicb.2022.963059. eCollection 2022.
2
The Diagnostic Value of Quantitative Assessment of Diatom Test for Drowning: An Analysis of 128 Water-related Death Cases using Microwave Digestion-Vacuum Filtration-Automated Scanning Electron Microscopy.硅藻检验定量评估对溺水的诊断价值：应用微波消解-真空过滤-自动扫描电子显微镜对128例与水相关死亡病例的分析
J Forensic Sci. 2017 Nov;62(6):1638-1642. doi: 10.1111/1556-4029.13455. Epub 2017 Feb 15.
3
Evaluation of Inspection Efficiency of Diatom Artificial Intelligence Search System Based on Scanning Electron Microscope.基于扫描电子显微镜的硅藻人工智能搜索系统检测效率评估。
Fa Yi Xue Za Zhi. 2022 Feb 25;38(1):40-45. doi: 10.12116/j.issn.1004-5619.2021.410719.
4
A comparative assessment of time-consuming and laborious diatom analysis：Brief experimentation with suggestion of automatic identification.耗时费力的硅藻分析的比较评估：自动识别建议的简要试验
Forensic Sci Int. 2024 Feb;355:111939. doi: 10.1016/j.forsciint.2024.111939. Epub 2024 Jan 17.
5
Research Progress of Automatic Diatom Test by Artificial Intelligence.人工智能自动鉴定硅藻的研究进展。
Fa Yi Xue Za Zhi. 2022 Feb 25;38(1):14-19. doi: 10.12116/j.issn.1004-5619.2021.410404.
6
Automated diatom searching in the digital scanning electron microscopy images of drowning cases using the deep neural networks.利用深度神经网络对溺亡案件的数字扫描电子显微镜图像进行自动化硅藻搜索。
Int J Legal Med. 2021 Mar;135(2):497-508. doi: 10.1007/s00414-020-02392-z. Epub 2020 Aug 13.
7
Digital whole-slide image analysis for automated diatom test in forensic cases of drowning using a convolutional neural network algorithm.基于卷积神经网络算法的数字全玻片图像分析在溺死法医案例中的自动化硅藻检测。
Forensic Sci Int. 2019 Sep;302:109922. doi: 10.1016/j.forsciint.2019.109922. Epub 2019 Aug 8.
8
Application of the microwave digestion-vacuum filtration-automated scanning electron microscopy method for diatom detection in the diagnosis of drowning.微波消解-真空过滤-自动扫描电子显微镜法在溺死诊断中硅藻检测的应用
J Forensic Leg Med. 2015 Jul;33:125-8. doi: 10.1016/j.jflm.2015.03.009. Epub 2015 Apr 2.
9
Analysis of false-positive results of diatom test in the diagnosis of drowning-would not be an impediment.分析硅藻试验诊断溺死时的假阳性结果——不会成为障碍。
Int J Legal Med. 2019 Nov;133(6):1819-1824. doi: 10.1007/s00414-019-02021-4. Epub 2019 Feb 15.
10
Application of Diatoms Quantitative Analysis in the Diagnosis of Drowning.硅藻定量分析在溺死诊断中的应用。
Fa Yi Xue Za Zhi. 2022 Feb 25;38(1):110-113. doi: 10.12116/j.issn.1004-5619.2021.410707.

引用本文的文献

1
The application of artificial intelligence in forensic pathology: a systematic literature review.人工智能在法医病理学中的应用：一项系统的文献综述。
Front Med (Lausanne). 2025 Jul 24;12:1583743. doi: 10.3389/fmed.2025.1583743. eCollection 2025.
2
"UDE DIATOMS in the Wild 2024": a new image dataset of freshwater diatoms for training deep learning models."野外 UDE 硅藻 2024"：一个用于训练深度学习模型的淡水硅藻新图像数据集。
Gigascience. 2024 Jan 2;13. doi: 10.1093/gigascience/giae087.
3
The application of vibrational spectroscopy in forensic analysis of biological evidence.

本文引用的文献

1
An efficient method for building a database of diatom populations for drowning site inference using a deep learning algorithm.利用深度学习算法建立用于推断溺水地点的硅藻种群数据库的有效方法。
Int J Legal Med. 2021 May;135(3):817-827. doi: 10.1007/s00414-020-02497-5. Epub 2021 Jan 3.
2
Research advances in forensic diatom testing.法医硅藻检验的研究进展
Forensic Sci Res. 2020 Mar 23;5(2):98-105. doi: 10.1080/20961790.2020.1718901. eCollection 2020.
3
Deep learning-based diatom taxonomy on virtual slides.基于深度学习的虚拟切片上的硅藻分类学。
振动光谱在生物证据法医分析中的应用。
Forensic Sci Med Pathol. 2025 Mar;21(1):406-416. doi: 10.1007/s12024-024-00866-9. Epub 2024 Aug 24.
4
U-Net and ResNet50-Based Automatic Pipeline for Bacterial Colony Counting.基于U-Net和ResNet50的细菌菌落计数自动流程
Microorganisms. 2024 Jan 18;12(1):201. doi: 10.3390/microorganisms12010201.
5
Validation and optimization of the diatom L/D ratio as a diagnostic marker for drowning.验证和优化硅藻的长宽比作为溺死的诊断标志物。
Int J Legal Med. 2023 May;137(3):939-948. doi: 10.1007/s00414-023-02970-x. Epub 2023 Mar 4.
6
Apple-Net: A Model Based on Improved YOLOv5 to Detect the Apple Leaf Diseases.苹果网络：一种基于改进的YOLOv5检测苹果叶部病害的模型。
Plants (Basel). 2022 Dec 30;12(1):169. doi: 10.3390/plants12010169.
Sci Rep. 2020 Sep 2;10(1):14416. doi: 10.1038/s41598-020-71165-w.
4
Automated diatom searching in the digital scanning electron microscopy images of drowning cases using the deep neural networks.利用深度神经网络对溺亡案件的数字扫描电子显微镜图像进行自动化硅藻搜索。
Int J Legal Med. 2021 Mar;135(2):497-508. doi: 10.1007/s00414-020-02392-z. Epub 2020 Aug 13.
5
Fully Convolutional Neural Network for Detection and Counting of Diatoms on Coatings after Short-Term Field Exposure.基于短期野外暴露后涂层上硅藻的检测和计数的全卷积神经网络。
Environ Sci Technol. 2020 Aug 18;54(16):10022-10030. doi: 10.1021/acs.est.0c01982. Epub 2020 Jul 28.
6
[Deep learning network-based recognition and localization of diatom images against complex background].基于深度学习网络的复杂背景下硅藻图像识别与定位
Nan Fang Yi Ke Da Xue Xue Bao. 2020 Feb 29;40(2):183-189. doi: 10.12122/j.issn.1673-4254.2020.02.08.
7
Digital whole-slide image analysis for automated diatom test in forensic cases of drowning using a convolutional neural network algorithm.基于卷积神经网络算法的数字全玻片图像分析在溺死法医案例中的自动化硅藻检测。
Forensic Sci Int. 2019 Sep;302:109922. doi: 10.1016/j.forsciint.2019.109922. Epub 2019 Aug 8.
8
Focal Loss for Dense Object Detection.用于密集目标检测的焦散损失
IEEE Trans Pattern Anal Mach Intell. 2020 Feb;42(2):318-327. doi: 10.1109/TPAMI.2018.2858826. Epub 2018 Jul 23.
9
The Diagnostic Value of Quantitative Assessment of Diatom Test for Drowning: An Analysis of 128 Water-related Death Cases using Microwave Digestion-Vacuum Filtration-Automated Scanning Electron Microscopy.硅藻检验定量评估对溺水的诊断价值：应用微波消解-真空过滤-自动扫描电子显微镜对128例与水相关死亡病例的分析
J Forensic Sci. 2017 Nov;62(6):1638-1642. doi: 10.1111/1556-4029.13455. Epub 2017 Feb 15.
10
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks.更快的 R-CNN：基于区域建议网络的实时目标检测。
IEEE Trans Pattern Anal Mach Intell. 2017 Jun;39(6):1137-1149. doi: 10.1109/TPAMI.2016.2577031. Epub 2016 Jun 6.