The process of using automated, high-throughput microscopy to capture images, followed by machine learning-enabled image analysis, is known as high-content screening (HCS)¹. This method has completely transformed² how researchers conduct cell-based assays and workflows for drug target discovery³, cellular functional assays⁴ and toxicology studies⁵. As a formally trained scientist, and in my years of helping researchers overcome common imaging and analysis challenges, I’ve found that the most compelling benefits of HCS are reducing time spent doing manual tasks and increasing assay reproducibility and reliability.

Today’s innovative imaging and analysis solutions leverage advanced technology to measure complex and multiparametric phenotypic effects in real time. It’s now possible to extract and analyze biologically relevant data from images faster, and with more accuracy, than ever before.

High quality, high content images

Getting the most accurate biological signal while minimizing noise interference is essential if researchers are to acquire the highest quality raw images. A high-content system offers inherent capabilities to achieve this. Their quantitative component reveals insights beyond manual methods—especially in 3D or 4D cellular models— at a scale unachievable with traditional, qualitative microscopes.

The quality of data generated by microscopic technology is dependent on the number and types of lasers, numerical aperture and working distance of objectives, autofocusing capabilities, and the superiority of the camera. Market-available systems that integrate top-tier attributes increase efficiency, enhance the acquisition process and empower researchers to gather reliable, robust data sets.

High content imaging in shared lab space

HCS systems are typically operated in central or core lab facilities where researchers share costs, centralize the technical expertise required for operation and conduct multiplexed cell-based assays faster and with more consistent results. In shared facilities, significant computing power is required to analyze such large image-based data sets and consideration must be given for data management systems, data security, and long-term storage. According to a systematic literature review of HCS studies⁶, even in multiplexed assays, computational algorithms and statistical requirements limit how much data can be collected to only one or two parameters. New analysis offerings extend this to multiparametric analyses of multiple parameters. Advanced high-content imaging systems have built-in software to perform image data extraction and analysis for any number of users at every analytic or post-analytic step of the experimental workflow.

The benefits of automation

Some labs, particularly in academic settings, still rely on individuals to manually analyze image data, which is time-consuming, limits the number of parameters that can be measured and introduces variability due to subjectivity. With manual data analysis, results tend to be generalized across all cells in the same plate and/or well and normalization to controls can be challenging. For experiments in which hundreds or even thousands of images are generated, manual analysis may be simply impossible. For these reasons, increasing numbers of scientists are adopting AI-enabled analysis platforms that streamline the entire process. Sophisticated AI-powered computational algorithms extract, analyze, and render results meaningful and statistically relevant, in less time and with less variability than a manual approach. Solutions with machine learning-enabled software are found in the training of the software itself, where the training process focuses on distinguishing cells from non-cellular entities, utilizing drawing tools and a deep learning segmentation technique. Deep learning segmentation uses hundreds of thousands of features to separate the elements within a sample and classify them based on specific characteristics such as morphology or signal intensity.

Commercially available HCS systems facilitate the growth of labs from lower throughput experiments with fewer parameters to higher throughput multiplexed and multiparametric experiments. They are capable of handling more samples and producing larger amounts of data with less human intervention.

End-to-end automation

Scalable automation is key to unlocking productivity and removing bottlenecks in workflows. Many commercially available high content screening systems now offer end-to-end automated solutions that minimize hands-on time from sample prep and culturing to data analysis and interpretation. Removing the potential for human error and variability improves the time to results, as well as the quality of those results. Additionally, scalability and modularity in configurable workflows allows labs to accommodate growth in sample volumes and facilitates increasingly complex analyses and multiple study endpoints seamlessly. This is vital in a core lab setting where technical staff, with varying levels of experience and expertise, operate several instruments for multiple projects and investigators. Automation can be scaled to any part of the experimental workflow, in any size of lab, to maximize economies of scale, reduce inter-operator variability and the potential for error, and improve laboratory productivity.

Catalyzing the next scientific breakthrough

The consistent application of identical settings and analyses across all plates minimizes variations, improves efficiency and enhances the reproducibility and dependability of the assays. Facilitated by intensified imaging, increased sample sizes, replicated tests, and swift analysis, high content screening reveals null effects more quickly. By expediating crucial insights, the technology allows researchers to pivot sooner to new projects with potentially positive outcomes.

As computational algorithms improve and facilitate truly multiparametric studies, many more labs will require HCS systems that deliver rapid end-to-end workflow solutions for multiplexed cell-based assays. By increasing throughput and minimizing subjectivity, they enable scientists to acquire, segment, analyze and quantify complex image-based data that pushes the boundaries of scientific and biomedical knowledge.

References

1. Mattiazzi Usaj M, Styles EB, Verster AJ, Friesen H, Boone C, Andrews BJ. High-Content Screening for Quantitative Cell Biology. Trends Cell Biol. 2016;26(8):598-611. doi:10.1016/j.tcb.2016.03.008
2. Gough AH, Johnston PA. Requirements, features, and performance of high content screening platforms. Methods Mol Biol. 2007;356:41-61. doi:10.1385/1-59745-217-3:41
3. Esner M, Meyenhofer F, Bickle M. Live-Cell High Content Screening in Drug Development. Methods Mol Biol. 2018;1683:149-164. doi:10.1007/978-1-4939-7357-6_10
4. Dourlen P, Chapuis J, Lambert JC. Using High-Throughput Animal or Cell-Based Models to Functionally Characterize GWAS Signals. Curr Genet Med Rep. 2018;6(3):107-115. doi:10.1007/s40142-018-0141-1
5. Li S, Xia M. Review of high-content screening applications in toxicology. Arch Toxicol. 2019;93(12):3387-3396. doi:10.1007/s00204-019-02593-5
6. Singh S, Carpenter AE, Genovesio A. Increasing the Content of High-Content Screening: An Overview. J Biomol Screen. 2014;19(5):640-650. doi:10.1177/1087057114528537

Recent posts