AI image recognition enables automatic identification and classification of individual bacterial cells
Identifying, sorting, and exporting individual bacterial cells rather than populations of them has been incredibly complex, expensive, and often simply doesn’t work without damaging the cells.
Now, researchers from the Qingdao Institute of Bioenergy and Bioprocess Technology (QIBEBT) of the Chinese Academy of Sciences (CAS) and their collaborators have proposed a new system called “EasySort AUTO” that enables single-cell analysis even of bacteria. It is based on artificial intelligence image recognition.
The study was published in lifetime on December 18.
Technologies that allow the classification and analysis of cells they generally employ fluorescence activated cell sorting (FACS), which can sort a mixture of different biological cells in vessels or tubes, one cell at a time.
This works well for relatively large human tissue or organ cells, but not for bacteria, which are typically about 1,000 times smaller in volume. “Even when FACS can be performed in bacteria, it is generally almost impossible to sort cells in an index-based manner, particularly if cell vitality is preserved,” said Diao Zhidian of the QIBEBT Single Cell Center, first author. of the study.
“It’s complicated and expensive to classify bacteria, and even on the rare occasions that you do, you can’t do much with it anyway,” said Kan Lingyan, co-lead author and an engineer at QIBEBT’s Single Cell Center.
As a result, microbiome research has become mired in relatively crude studies of cell populations. An easy-to-operate, low-cost, index-based, and vitality-preserving single cell sorting system is needed to work at the bacterial scale and thus enable analysis of single cell microbiomes. To do this, it is necessary to solve two fundamental problems. First, a method of detecting and sorting cells, and second, exporting the tiny bacterial cells cell by cell into container tubes.
“For the first problem, we turned to artificial intelligence for its ability to detect even hidden properties of cells only from cell images, and thus for cell identification,” said Zhao Yilong, a co-author from the Center for QIBEBT single cells.
“We deployed a deep convolutional neural network, a type of AI inspired by the visual cortex of animals and most commonly used for visual image identification,” said Li Yuandong, an engineer at the QIBEBT Single Cell Center.
“For the automation of the export of single cells to container tubes, we devised a method that does not require any additional complex equipment, but instead involves a capillary tube harvest module coupled to an optical clamp and a positive-mount microscope stage.” said Prof. Ma Bo, corresponding author from the QIBEBT Single Cell Center, who led the study.
Individual target cells are identified by imaging based on artificial intelligence algorithms, and then move cell populations by optical tweezers, followed by export to Polymerase chain reaction (PCR) through an automatic collection platform. During this process, a single cell is packaged into a micro-droplet and automatically exported in a precisely indexed, “one cell, one tube” manner, preserving the vitality of each cell.
The EasySort AUTO system is more than 93% efficient, which means that 93% of the time, the exported drop contains a single cell, identified and indexed. It can perform the task at a rate of about 120 cells per hour.
The researchers tested the technique on yeast cells (which are fungi rather than bacteria, and about 3-4 times larger, thus somewhere in the middle in size between a human cell and a bacterium) and the Escherichia coli bacterium. In both cases, more than 80% of the individual cells could be subsequently cultured, indicating that their vitality had been preserved during sorting and export. High precision of AI object detection was confirmed in yeast samples whose identity was already known.
According to Professor Xu Jian of QIBEBT’s Single-Cell Center, who co-led the study.
In addition, researchers have coupled the EasySort AUTO system with genome sequencing to link single-cell phenotype identification with single-cell genotyping analysis for both bacterial and human cells.
Zhidian Diao et al, AI-Assisted Automatic Index-Based Single Cell Microbial Sorting System for One-Cell-One-Tube, lifetime (2022). DOI: 10.1002/mlf2.12047
chinese academy of sciences
Citation: AI Image Recognition Enables Automatic Identification and Classification of Individual Bacterial Cells (Dec 20, 2022) Retrieved Dec 20, 2022 from https://phys.org/news/2022-12-ai-image- recognition-automatic-identification.html
This document is subject to copyright. Apart from any fair dealing for private study or research purposes, no part may be reproduced without written permission. The content is provided for informational purposes only.