By the end of the twentieth century, revolutions took place, which gave rise to digital pathology. A new technology then was developed which allowed clinicians and researchers to view an entire tissue section to be scanned on an objective scale. Originally, it was called virtual microscopy, but now it is known as Whole Slide Imaging.
Though much more effective than the previous technologies, Whole Slide Images also comes with its set of challenges. These include difficulty in reading, visualization, storage, and analysis. This is the reason for the development of several technologies – to facilitate the handling of these images.
In this article, we will analyze some of the most popular and widely used technologies in the field of digital pathology. These technologies range from specialized libraries for the reading of images, to end-to-end Platforms for Digital Pathology that will allow reading, visualization, and analysis.
Whole Slide Imaging
The technology of whole slide imaging has taken shape due to the multiple efforts made to digitize the whole tissue slides into high-resolution images. Digging into a little history of WSI, Ferreira et al. in 1997 the first to describe and carry out efforts towards this evolution. It was then defined as a “virtual microscope”. Later on with the contribution of Wetzel and Gilbertson, who were then working for Interscope Technologies, the cost and speed of these WSI systems became affordable. Since then, there has been a rising interest in the commercial use of it.
Whole Slide Imaging systems consist of 4 main elements –
- Light source
- A microscope with multiple lenses
- A digital camera
- A system for repositioning the camera view along the sample
The main advantage of these WSI systems is that they can easily capture very high-resolution images, which are usually in the range of gigapixels. These high-resolution images then can be scanned with multiple magnitudes, the most commonly used are that of x20 or x40, and some of these higher magnitudes are used in specific cases. For example blood smears.
Currently, these advanced WSI scanners can use one or several imaging modes, these include bright-field, fluorescent, and multi-spectral imaging. The following image shows examples of WSI modalities.
With every mode, researchers can get a view of different anatomical structures or even the physiological events in the tissue. Each mode uses different light sources. White light is used in the bright field within the visible spectrum, in another mode of fluorescent imaging, the tissue is exposed to light of a certain wavelength, which results in the emission of lower-energy light.
In the spectroscopic mode of imaging, the images are obtained by getting the spectral information of each pixel of an image. This spectral information is also called the data cube because the information is not only on the x and y axes, but also each contains the z plane, and has intensity information at different wavelengths, this helps in creating a stack of spectral images.
What are the Different Kinds of Image Formats?
Currently, there is no particular data format for Whole Slide Imaging which is widely accepted by scanner manufacturers. This lack of data then led to the creation of multiple types of proprietary file formats like NDPI, SVS, and SCN. However, there are two formats that have become the standard for WSI’s:
Digital Imaging & Communications in Medicine
Tagged Image File Format
Software Systems Created Especially For Digital Pathology
There are two groups of Platforms for Digital Pathology specifically designed to overcome any kind of challenges. These are:
- Open-Source Platforms
- Closed-Source Platforms
There are two types of closed-source Platforms for Digital Pathology: those developed by the scanner makers themselves and those developed outside of them. Manufacturers of scanners frequently create their own tools for WSI visualization, manipulation, and analysis.
Here are 10 Software Tools and Platforms for Digital Pathology:
1. AperioImageScope (Developed by Leica Biosystems)
Features:
- Entire-Slide Scanning and Analysis
- One ScanScope Replaces a Number of Automated Microscopes
- Scanning First Saves Time
- Extremely Fast Viewing
- Entire Slides Analyzed in Minutes
- No file size limitations
- Algorithms for IHC and Stain Intensity
- High Resolution
2. HCImage Hamamatsu software (Hamamatsu Photonics)
Features:
- Capture Presets save basic settings such as capture mode, channels, filters, and exposure times, as well as output trigger settings and advanced camera properties. Presets can be saved and loaded when needed.
- Support for the ORCA-Lightning
- Support for the ORCA-Fusion
- High Speed Streaming to DISK or RAM (circular buffer option)
- Line Profile, plots pixel intensity values for monochrome or color images while live, on a single image or an image sequence. The data can then be exported to a spreadsheet.
- Includes capture for single images, or image sequences over time
- Objects are identified by drawing manually or through intensity thresholding
- Measurement of multiple objects with analysis of Area, Length, and Intensity in a single image
- Intensity Plot against time for a single object through an image sequence.
3. OptraASSAYS (OptraScan)
Features:
- Morphology-based image analysis algorithms
- Enabled through IMAGEPath Image Management
- Libraries provide IHC, HE Tumor vs. non-tumor cell identification
- ANN (artificial neural network) based machine learning of algorithms
- Processed image information in XML overlay
- C libraries/API’s/Web Services User Interfaces to tweak algorithms without altering code
- Available as an On-Demand subscription module or stand-alone software
On the other hand, the closed-source Platforms for Digital Pathology are developed by third parties. These include:
4. Image-Pro (Media Cybernetics)
Features:
- Image Management & Display: Open, display, and manage nearly any image and linked data.
- Processing & Adjustments: Use a range of filters and editors to make images analysis-ready.
- Usability: Software designed to improve your productivity, not slow you down.
- Data Analytics & Reporting: Automatically compile data from a series of images, review & report.
- Auditing & Authentication: Ensure the authenticity of images by tracking usage & logging changes.
- Scripting: Automate routine tasks with macros or the built-in scripting language.
5. APP Center (Visio Pharm)
Features:
For all of your image analysis needs, the APP Centre is the one-stop shop. With the help of its tissue image analysis software, each of the more than 140 modular analysis units in the APP Centre is intended to address a particular issue. All tissue, stain, and modality types are covered by our extensive selection of APPs, making it simple for you to identify the ideal answer for your particular issue.
By offering pre-built apps that can be used as a jumping-off point for creating and improving workflows, the APP Centre saves time. Expertly designed apps can also provide you with new ideas for tackling problems and insights.
6. Aiforia, Pathobox, and Halo image analysis platform (Indicalab)
Features:
Pharmaceutical, healthcare, and research organizations worldwide are using HALO® for high-throughput, quantitative tissue analysis in oncology, neuroscience, metabolism, toxicology, and more due to its unparalleled ease of use and scalability, strong analytic capabilities, and the fastest processing speeds available for digital pathology.
- Accessible Modules and Add-ons
- Extensive Analytics
- Scalable Platform
The majority of these are too costly for researchers, students, or small labs, and as they are all closed designs, it is not feasible for external users to add features.
For open-source platforms, the main advantage is that it gives users access to the source code and allows them to add new functions to suit their needs.
Various participants are involved in open-source projects; these participants can be active contributors or passive end users who provide the contributor’s feedback. We refer to these individuals as community members. Bigger communities enable better bug-fix support and more regular product updates.
In Digital Pathology, the open-source Platforms for Digital Pathology that are the most popular and have the most developer support are:
7. Digital Slide Archive (DSA)
DSA is a web-based platform that integrates tools for image annotation and analysis and enables the storing, maintenance, and visualization of WSIs. It is based on CDSA (Cancer Digital Slide Archive) and was developed at the University of Atlanta.34 Memcached, a distributed memory object caching system that enables quick access to tiles, is used in conjunction with MongoDB as the database system and Girder, an open-source, web-based base system created by Kitware, as part of DSA’s operational framework.
8. Quantitative Pathology & Bioimage Analysis (QuPath)
Written in Java, QuPath is a desktop cross-platform utility developed by Queen’s University Belfast. It follows supplement 145 (with a pyramid organization) and supports reading a variety of formats, including Bio-Formats, OpenSlide, and DICOM. QuPath is utilized extensively in DP applications, but it’s also applied in cell biology, oncology, and other bio-imaging fields.
9. Orbit Image Analysis
Actelion Pharmaceuticals Ltd., which is currently known as Idorsia Pharmaceuticals Ltd., developed the Java program Orbit Image Analysis. It supports bright-field and fluorescent multi-channel pictures and reads WSIs using Bio-Formats. Pixel classification, object segmentation, and object classification are the main areas of study for Orbit.
Various machine learning approaches, including Support Vector Machine (SVM) and deep learning, are employed by the algorithms put into place for the analysis. Because of its Map-reduce methodology, which enables it to handle images at the tile level, additional algorithms can be readily added even if they were not intended for processing WSIs.
10. Cytomine
A web platform called Cytomine was first created at the University of Liège. Cytomine is a multi-technology, highly sophisticated piece of software, but it’s easy to deploy because Docker containers, which are lightweight, modular virtual computers, simplify the deployment process. Four major groups comprise Cytomine’s architecture: the core server, which is the center of the system and houses the technologies related to the web application’s backend, as its name suggests.
General Bio-Imaging Software Used in Digital Pathology
Bio-imaging systems are extensively utilized in the field of DP, in addition to instruments designed especially for the processing of DP images. These Platforms for Digital Pathology frequently lack sufficient capabilities for WSI reading, visualization, and analysis.
ImageJ is a free and open-source image processing program. Since its initial release in September 1997, ImageJ has grown to become one of the biological sciences’ most popular applications.45 in ImageJ’s structure makes it possible to install additional modules (such as macros and plugins) that increase its functionality.
With ImageJ’s limited support for Bio-Formats, it is possible to integrate Bio-Formats and read WSIs.
Anne E. Carpenter and Thouis R. Jones invented CellProfiller, a cell image analysis program, in the Sabatini and Golland laboratories. Since its 2005 debut, a number of enhancements have been made. It was redone in Python for the second version that was released in 2011 after being developed in MATLAB initially.
Deep learning and three-dimensional analysis were supported by its third version. The fourth version enhanced various image analysis techniques and made the switch from Python 2 to Python 3 to boost performance. Because of its adaptability, constant development, and simplicity of use, CellProfiller has grown to be one of the most popular software packages for microscopy image analysis.
ICY is an open-source program developed in 2011 by Institut Pasteur’s BioImage Analysis Lab. The greatest range of biological applications might be covered by it, cutting-edge image analysis methods could be easily accessed, and reproducibility research could be promoted. ICY is a Java desktop program that also includes a website with a central repository for sharing and adding plugins, scripts, and protocols.
Ilastik is an instrument designed specifically for the tracking, counting, segmentation, and classification of objects in biological images. Its machine-learning tools for image analysis are numerous. Its functioning is comparable to CellProfiler’s since both utilize pipelines for workflow and have batch-processing systems that allow numerous photos to be analyzed after one pipeline.
Ilastik takes photos up to 5D as input for certain tasks, such as object classification, tracking, and pixel classification. It can be used to send a batch of the image’s tiles for additional analysis, despite the fact that it is unable to read WSIs.
Also Read: Accelerating Forward: How Digital Pathology Speeds Up Slide Review
Source: Journal of Pathology Informatics
