High-throughput microscopy

Digitization and Automation in Microscopy
Ludwig Wildner
Ludwig Wildner

Head of Software at PreciPoint

Nicolas Weiss
Nicolas Weiss

CEO at PreciPoint

Overview

What is High-Throughput Microscopy?

High-throughput microscopy aims to process as many slides per time unit as possible. The degree of effectiveness is measured by defining the analyzed slides per time unit as key figure. Analyzed slides, in this case, mean slides that have fulfilled their primary purpose. They were, for example, assessed or evaluated by a researcher and then documented. After that, they can be archived for later use. In principle, two different methods of high-throughput microscopy exist, as high throughputs can be achieved with both analog as well as digital methods. This article will discuss both respective methods and their differences.

Which Parameters Play a Role in Analog and Digital High-Throughput Microscopy?

With an analog light microscope and an appropriately configured microscope stage, a substantial number of samples can be examined within a relatively short time span. With this method, however, the microscope operator and the sample must always be in the same place at the same time.

Important parameters of analog high-throughput microscopy

  • Analyzed slides / time
  • Costs (predominantly labor costs) / analyzed slides 

Digital high-throughput microscopy allows for location- and time-independent operations. Researchers and other users no longer have to have the physical slide directly in front of them, since they can be viewed any time as virtual slides on their computer. To be able to do this, the slides have to be first digitized by a slide scanner.

Important parameters of digital high-throughput microscopy

  • Analyzed slides / time
  • Costs* / analyzed slides
  • Digitized slides / time
  • Walk-away time (time span during which the laboratory device can operate autonomously; determined by storage capacity, scanning speed, and degree of automation.)

Efficient Diagnosis: Analyzing Lab Samples Quickly

1. Manual and analog

It is possible to process a large volume of samples utilizing analog microscopy. Depending on the throughput, several light microscopes as well as several people for operating these microscopes are needed. As this process, however, requires making various evaluations, it is rather time intensive. Annotations cannot be added directly onto the sample, but rather have to be added to, for example, a separate piece of paper.

2. Manual and digital

For this method, a digital microscope without an automated sample feed is used, with the slides being placed under the microscope manually. For the digitization, an analog microscope with a top-mounted camera or a slide scanner without a slide feed can be used. The samples are then viewed in a viewer software. To achieve a high throughput, several scanning devices have to be operated and loaded simultaneously.

3. Automated and digital

For this method, a slide scanner with an automated slide feed is used, making manual steps obsolete and making it possible to scan slides overnight. The scans can then be examined in a viewer software at any time. Ideally, slide scanners operate without any need for manual interventions, which is often referred to as the above-mentioned ‘walk-away time’. This is determined by the scanner’s slide holding capacity, the scanning speed, and the degree of automation.

How Is Microscopy Digitized?

Digitizing microscopy processes requires a digital microscope or a slide scanner. Slide scanners, however, are not exclusively used for high-throughput purposes only. Among slide scanners, varying models are available, all geared toward differing throughput and quality needs. Their respective slide holding capacities vary from one up to a thousand slides. Slide holding capacity, however, is not the only parameter that influences walk-away time.

How Does a Slide Scanner Work?

The general functionality of scanners is relatively similar, does, however, consist of differing components:

Loading mechanism

The loading systems for scanners can differ significantly. Some are equipped with automatic loading, other have to be loaded manually. To load automatically, a storage system for holding the slides is incorporated into the slide scanner. Generally, the higher the slide holding capacity, the larger the device itself.

Light microscope + camera

Physical slides are inserted under the microscope and scanned using a combination of microscope optics, mechanics (XY stage and Z axis), and a camera to save them as a file.

Image processing software

To be able to automatically generate complete scans of samples, scanners need to be coupled with special software. Digitized samples are also referred to as ‘Whole Slide Images’ (WSI) and the corresponding hardware as ‘Whole Slide Scanner”.

Motorized XY stage and Z axis:

To be able to fully digitize samples, the samples have to either be moved and focused on manually, or the XY as well as the Z axes of the light microscope are motorized, enabling an autonomous moving and focusing of samples.

What Is a Whole Slide Image?

Whole Slide Images (WSI) are fully digitized microscopic samples. To image a sample in its entirety, the Whole Slide Scanner stitches single individual high-resolution images together to form a complete scan. As with tissue sections under a microscope, Whole Slide Images can be navigated through, magnified, and analyzed on a screen. The digital sample allows the viewer to benefit from a large field of view, providing a good overview of and easy navigation within the sample.

What Is the Difference Between a Slide Scanner and a Digital Microscope?

Not all digital microscopes are able to generate Whole Slide Images. The range of differing digital microscopes available is large. Some systems are only transmitting a live image of a camera to a screen. Others allow users to manually digitize samples by moving the stage and scanning the individual sections of the sample. The obtained individual images are then stitched to form a Whole Slide Image. Digital microscopes with an incorporated scanner and an automated XY stage, on the other hand, allow for automatic scanning.

A slide scanner is usually used to digitize microscopic samples as quickly and thoroughly as possible. Most scanners only provide scanned images (WSI); further analyses have to be performed with the finished scanned image.

Some scanners and digital microscopes also offer the option of having the sample undergo automatic quality controls before scanning. An overview image is used to identify particularly relevant areas or to check the quality of the specimen. Additionally, particular digital microscopes also offer the option of examining the sample before scanning, just as when working with an analog microscope.

What Role Does High-Throughput Microscopy Play and What Challenges Does It Pose?

Higher volumes of samples

Today’s demands on laboratories are constantly increasing. Simultaneously, demands on the training and education of laboratory staff and medical technical assistants (MTAs) are also increasing. Research laboratories, for example, often have to analyze a large volume of specimen to achieve statistically relevant results. Pathologies are sent more samples than ever before, partially due to an increase in the number of biopsies performed. In teaching and education, virtual samples have become, partly as a result of the Covid pandemic, more and more attractive and call for efficient digitization.

Staff and specialist shortages

Proportionally to the growing volume of samples, there are currently not enough new specialists to absorb the workload. This leads to a bottleneck, which cannot be sufficiently compensated for by human labor – at least in the short and medium term –, creating growing time pressure within laboratories and making disruptions and deviations even more serious. Aiming to relieve these pressures through digital and automated technologies is thus a natural conclusion. Researchers and other users are often not available directly on site. If at all possible, samples then have to sent by mail. Efficient exchange between colleagues is crucial, considering the complexity of their tasks and the internationalization in research as well as the large variety of possible diagnoses. Managing an archive of physical specimen in hospitals, research institutions, or museums is resource intensive. Finding and preparing samples for discussions and team meetings, thus, requires a lot of time.

Requirements for a modern laboratory work environment

One of the greater challenges poses the desire and necessity to be able to work independently of location and flexibly in terms of time. Another aspect is the handling as well as the ergonomics of operating a microscope. When examining samples with analog methods, i.e., through an eyepiece, the pathologist, for example, has to operate the microscope while dictating their findings. Working in this posture can, on the one hand, lead to orthopedic problems. On the other hand, sorting cases and their associated files is less convenient than with digital methods and the support of a pathology information system.

Retention regulations

Samples that have already been diagnosed are subject to archiving requirements. The same holds true for samples used in studies. As a result of these regulations, samples pile up in the basement and space, funding, as well as staff are needed to build and operate storage systems that allow for a systematic access to these samples. By making use of digital archives, this problem can be mitigated: Samples can be, for example, found through a search mask and filters in the image management system without having to leave your workplace.

How Can High-Throughput Microscopy Aid in Solving these Problems?

Digitization and automation increase efficiency

Microscopy and the analysis of samples gain from a digital working environment through:
  • Analyzing samples is done on-screen without having to have the physical slides present at the workplace, making it independent of location.
  • Documents, photo documentations (for example, of the dissection) and metadata for each sample are immediately accessible through the digital archive.
  • Dictations can be captured directly, using digital speech recognition.
  • By linking the data, the number of errors can be reduced.
  • The effort needed for documentation is reduced.

Automation can improve various steps in the laboratory: from dehydrating the tissue, staining the section and mounting a protective cover slip all the way through to immunohistochemistry. Combined with an automated slide digitization, the following advantages arise:
  • Structured and optimized workflows
  • Optimal utilization of laboratory capacities
  • Avoiding waiting times
  • Digitization of samples can be done overnight
  • Reduction of throughput times
  • Flexible collaboration and modern methods

Flexible collaboration and modern methods

In an increasingly digitalized world, quick and easy communication is expected in all areas. For laboratory and particularly microscopy work, digitized samples allow for working independently of location and time, leading to the following advantages:

  • International collaboration is made easier
  • Conferences with digital images
  • Second opinions on samples are made easier, quicker, and cheaper
  • Home office, location-independent working
  • Working more flexibly in terms of time
  • Experts and organizations benefit from better networking
  • Workload can be spread and distributed more flexibly

How can image analysis with artificial intelligence bring relief?

Digitization generates large volumes of data that can be used to train software regarding specific issues. This means that digitization is the foundation for utilizing artificial intelligence. Within pathology and histology, AI is perhaps the most significant innovation after immunohistochemistry and molecular pathology. With the help of AI software, a number of analyses can be automated and require only a fraction of the time needed when working manually. This leads to an increase in efficiency in pathology, histology, as well as in various quality controls.

What are some examples of AI image analyses?

Cell-based analyses:

  • Counting of specific cells or particles
  • Measuring distances of cells in-between each other
  • Determining cell or particle concentrations

Morphological analyses:

  • Recognition of structures in tissue and other specimens
  • Quantification of morphological structures
  • Automatic classification of specimens

In this context, AI cannot replace humans, however it can support and relieve them. A respective expert will always be responsible for making a diagnosis or assessing the samples.

In research, AI, in combination with large sets of data, enables studies to be conducted on a much larger scale than was previously possible. More complex software programs make it possible for researchers to develop their own AI-based applications to then explore varying questions.

Digital archives enable efficient data access

A digital archive cannot replace physical archives, which are required by law as storage facilities. But it can offer a much more convenient and quicker access to samples. Access times are much shorter and retrieving samples and data sets that are interrelated is made immediate and precise. As opposed to physical sample archives, digital archives are suitable to be used as central archives across locations. In combination with AI image analysis, this allows research to be done more efficiently.

Microscopic databases are a tremendous relief for professional education and training. Teachers and learners alike can profit from standardized and high-quality course content.

What Requirements Does a System for High-Throughput Microscopy Have to Meet?

The requirements for a digitizing solution differ depending on the area of application. The following applications can be distinguished:

  • Research
  • Education and teaching
  • Routine pathology

Slide Scanners and Their Respective Software Should Meet Various Requirements:

High reliability and security of device ​

Research Education  Routine Laboratory

High reliability and security of device

Highest reliability to avoid delays caused by work backlogs

Specimen must not be damaged: often rare specimen

Specimen must not be damaged: Expensive to repeat sample preparations and clean

Simple and intuitive controls, as staff and settings rotate/ change frequently

Full integration with it

Integration with IT as well as with upstream and downstream processes not required by default, optional.

Full integration with IT as well as with upstream and downstream processes required.

Simple and intuitive controls, as staff and settings rotate/ change frequently

Controls adapted to the requirements of routine pathology; trained staff working with the system on a daily basis

Image quality ​

Research Education Routine Laboratory

Various objectives can be used (5x to 100x)

Certain objectives can be used (20x, 40x)

High color accuracy

High resolution: spacial resolution, density resolution

Complete scan of sample optional

Complete scan of sample is an absolute MUST

Automatic tissue detection optional

Automatic tissue detection should be available

Compression of image data

Appropriate number of focal points for consistent sharpness (automatic)

Capacity and throughput ​

Research Education  Routine Laboratory

Low to medium capacity and throughput

High to very high capacity and throughput

Automation: required walk-away time can fluctuate strongly

High walk-away time required

Irregular, often during project peaks

Regular, daily

Samples often uneven; accordingly time-consuming to digitize

Samples very flat; accordingly fast to digitize

Often different dimensions of slides

Standardized dimensions of slides

High walk-away times desirable

Highest walk-away times necessary

Connectivity ​

Research Education Routine Pathology

Integration with Image Management System (IMS)

Integration with LIMS of the research laboratory is optional

Integration of the scanner with pathology information system is an absolute MUST

Option to upload data for conferences, international collaboration, lectures and exams

Easy sharing of digital slides for consults

Viewer software has to be available for all

Combination with pathology-specific viewer software

Barcode scanning optional

Barcode scanning is an absolute MUST

Integration with lab peripherals depending on requirements

Interaction with other laboratory equipment is necessary (e.g. staining machines)

Home office should be possible

AI interfaces should be available

Digital sample archive is linked

Integration not desired by default, optional

Integration with Image Management System (IMS) or Picture Archiving and Communication System (PACS)

Flexibility ​

Research Education Routine laboratory

Differing geometry and dimension of slides possible

Adaption to the dimensions of the repective institute’s slides possible

Connecting the system (see above) with third party products possible

Continuous loading and unloading of the system possible

Prioritization of samples optional

Prioritization of samples should be possible

Project management

Research Education Routine laboratory

Depending on the complexity and urgency of the workflows, digitization should be implemented as part of a project to avoid interruptions/ disruptions

Usage/ access to slide scanner and software by staff and students has to be regulated

Access to data has to be ensured for all users

Access to data has to be ensured for all users

Workflow should be planned holistically and adapted accordingly to the respective institute 

Because of the complex and highly interconnected nature of the procedures in pathology, slide scanners and devices alone are not enough for achieving successful processes

How Can Digitization in A Laboratory Environment Be Done Successfully?

Laboratories with a high throughput of samples can successfully manage it in a number of different ways. Certainly, the most appropriate method in the long term is to make use of the digitization and automation technologies available.

Changes, particularly organizational ones, usually go hand in hand with great difficulties. However, this does not necessarily have to be the case. Before implementing any digitization project, analysis, planning and practicing new procedures should be done, reducing uncertainties and risks this way. This holds true for both making gradual improvements as well as a complete transformation towards digital workflows.

Digital high-throughput microscopy can provide many laboratories with major relief within and improved efficiency of their processes. Over time, more and more technologies in the field of digital microscopy and image analysis will become available. Laboratories that make full use of these changes in due time can expect to gain both competitive advantages as well as more efficient cost structures in the long run.

How to Start with Laboratory Digitization?

This article is intended to provide you with a better understanding of how high throughputs in microscopy can be processed and which requirements to consider.

Whether your goal may be to increase efficiency by means of digitization or to implement automation, significant changes do not have to be overwhelming, and you do not have to do it alone.

PreciPoint offers several options to support you taking your next steps:

Would You Like to Learn More About High-Throughput Microscopy?

By submitting this form, you agree to our privacy policy.