Star protocols:用于分析微生物学细胞图像的深度学习框架

Star protocols:用于分析微生物学细胞图像的深度学习框架

Before you begin

Deep learning (DL) has proven to be extremely effective in addressing a range of major biological challenges, including predicting protein structure,4 DNA sequencing,5 and drug discovery.6 The application of DL has expanded into the microbiological field,7 particularly in cellular image analysis. In traditional cellular image analysis, there are several challenges that need to be addressed.

One challenge is that parasites have completely distinctive features in morphology during their complex life cycles,8 and the shape and size of the cells can vary considerably,9 making the classification and detection of different parasites and cells quite difficult. Additionally, obtaining high-quality and in-focus microscopic images can be challenging,7 due to various factors such as the diffraction barrier and defects in optical systems.10

DL-based cellular image analysis can solve these problems to some extent. However, the black-box nature of DL often leads to unexplainable results. Incorporating the knowledge and insights of experts into the modeling process can help to solve it, but most of the DL-based methods have not considered the importance of knowledge from microbiologists in cellular image analysis.11,12 They are highly specialized and lack detailed instructions for most microbiologists. As a result, it can be challenging to develop accurate and easy-to-use DL models for cellular image analysis in microbiology.

To address these challenges, this protocol introduces a knowledge-integrated DL framework for cellular image analysis in microbiology. By building upon the previous studies of our group,1,2,3 this protocol provides a comprehensive guide to implementing a wide spectrum of tasks (i.e., classification, detection, and reconstruction) in cellular image analysis. The following sections describe how the DL model integrates with human expert knowledge and provides step-by-step instructions accessible to both beginners and professionals.

Description of the methods

This protocol introduces three DL models integrated with knowledge from microbiologists, namely deep cycle transfer learning (DCTL),1 geometric-feature spectrum ExtremeNet (GFS-ExtremeNet),2and correcting out-of-focus microscopic images (COMI).3 These models are designed for the classification, detection, and reconstruction tasks of cellular images in microbiology.

DCTL and COMI are both based on cycle generative adversarial networks (CycleGAN),13 as illustrated in Figure 1A. CycleGAN is comprised of two sets of generator-discriminator structures, which are different types of neural networks with distinct functionalities. Generators are used to transform the input images into different styles, while discriminators are used to identify whether the images are synthesized or not. Unlike traditional GANs, the cycle network topology does not require the one-to-one pairing of source images (DomainX) and target images (DomainY), as in the case of DCTL.

In DCTL, X represents the morphologically similar macroscopic objects, while Y denotes the parasites to be recognized. The GeneratorXY transforms the macroscopic images in DomainX into their corresponding parasite images, SyntheticY. Then, GeneratorYX restores the images in SyntheticY back to the original macroscopic images, RestorationX. Another cycle performs the same process in reverse. Finally, the discriminators are used to distinguish between the generated images and the original images, which are used to help the generators improve the quality of the generated images.

Building upon the backbone of CycleGAN, DCTL incorporates human expert knowledge through two supplementary feature extractors, as shown in Figure 1B. Using four groups of extreme points, it calculates the microscopic and macroscopic correlation (MMC)1 to find the morphologically similar macroscopic objects of each parasite as a quantitative knowledge representation (Figure 2A). CycleGAN then learns the morphological information from these two image domains and teaches the supplementary feature extractors to identify different parasites. Each feature extractor is trained on both original images and synthetic images using a Cross-Entropy loss function.14 Once the model training is completed, the supplementary Feature ExtractorY can be applied to classify the four types of parasites.

Key resources table

Software and algorithms
Anaconda Anaconda v2.4.0
Spyder Spyder v5.3.3
Python3 Python v3.7.16
Tensorflow Tensorflow v1.15.0
Tensorboard Tensorboard v1.15.0
Tensorflow-estimator Tensorflow-estimator v1.15.1
Pytorch Pytorch v1.2.0
Torchvision Torchvision v0.4.0
Keras Keras v2.2.4
Keras-contrib Keras-contrib v2.0.8
H5py H5py v2.10.0
Scikit-learn Scikit-learn v1.0.2
Matplotlib Matplotlib v3.5.3
Scikit-image Scikit-image v0.17.2
Opencv-python Opencv-python v4.6.0.66
Pycocotools Pycocotools v2.0.5
Tqdm Tqdm v4.64.1
Pandas Pandas v1.3.5
Numpy Numpy v1.21.6
Protobuf Protobuf v3.19.0
Tensorflow-gpu Tensorflow-gpu v1.15.0
cuDNN cuDNN v7.6.5
Cudatoolkit Cudatoolkit v10.0.130
Codes and Datasets Github
Computing Platform: This protocol was performed on a computer with Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40 GHz Processor, two NVIDIA 2080Ti graphic cards, and 32G memory. Computer with more graphic cards is recommended to accelerate the training and evaluation. Windows 10


打赏 微信打赏,为服务器增加50M流量 微信打赏,为服务器增加50M流量 支付宝打赏,为服务器增加50M流量 支付宝打赏,为服务器增加50M流量
上一篇 08/09/2023
下一篇 08/10/2023


  • 错误使用的实验动物麻醉方法导致拒稿

    近期有一项报道,国内一位科研人员向《Scientific Reports》投稿遭拒,原因是该项研究在动物实验过程中,使用水合氯醛麻醉,而该物质作为麻醉剂,更应该是镇静剂,使用时其镇痛效果较差,并且有导致腹膜炎等的副作用,实验过程对动物不人道…

  • 利用ImageJ Fiji分析WesternBlot条带


    实验方法 02/03/2023
  • RNA immunoprecipitation RNA免疫共沉淀方法

    RNA免疫共沉淀方法:RIP is an antibody-based technique used to map in vivo RNA-protein interactions and RNA modifications such as…

    实验方法 12/17/2022
  • 头颈部CT影像数据集-Head-Neck-CT Dataset

    “该集合包含来自魁北克四个不同机构的 298 名经组织学证实的头颈癌 (H&N) 患者的 FDG-PET/CT 和放射治疗计划 CT 成像数据 所有患者在 4 月之间进行了治疗前 FDG-PET/CT 扫描 2006 年和 2014…

  • ChatGPT 中文使用教程调教指南

    ChatGPT模型是由OpenAI训练的大型语言模型,能够生成类人文本。通过向它提供提示,它可以生成继续对话或扩展给定提示的响应。 直接问它能干什么我是一个训练有素的大型语言模型,可以帮助你回答各种问题,比如关于政治、历史、科学、技术、艺术…