Digital Pathology is the process of digitizing pathology samples into image format while attaching all pertinent information, analysis, and annotation. Samples existing in a digital format can be stored, exchanged, viewed, and discussed electronically which allows researchers and clinicians to collaborate in real-time and on-line.
Table of Contents
Digital pathology is the method of converting sample slides to a digital format for the purpose of storing the data instead of the slide as well as allowing for real-time collaboration and analysis of samples by pathologists anywhere in the world. A fully integrated process can allow for fast, efficient, and powerful dissemination of information between experts. This sort of environment can also replace the need to have pathology experts on location at every institution that processes samples.
The first step in the digital pathology environment is rendering a slide into a digital format. This process is called scanning or digital microscopy. There are two basic forms of digital microscopy. The first is stitching and tiling of many images acquired with a digital camera into a larger mosaic image. This is typically done on a standard microscope with an automated stage and software that enables the accurate positioning of the stage to sample the large area without underlap or overlap as well as their stitching and smoothing together. The other modality of digital microscopy is true scanning of the sample. This involves a line scan detector imaging one row of pixels at a time while the slide itself moves past the detector. Once the image has been scanned it must be able to be viewed.
A large component of any digital pathology platform is the ability to easily and quickly view the sample. A digital pathology viewer must have the ability to zoom in and out of the image and navigate over the large image at high magnification. Other features of a viewer could include the ability to choose virtual magnifications, visualize scale bars, adjust brightness and contrast, and turn layers on and off should the sample have multiple stains in fluorescence.
While virtual pathology will likely involve many slides there needs to be an information management system. Minimally, the images would be managed in a system that has security to control access for viewing, an organizational infrastructure to house the data, and a way of attaching metadata related to the images. The metadata could include institution, pathologist, scan date, clinical study, image magnification, imaging technician, etc. All of these components need to be interrelated with high fidelity.
With the sample now rendered digital, viewed, and catalogued the next step is analyze for specific features. This can be done by a pathologist using the traditional methods of hand scoring and annotating or by software algorithms designed to segment and quantify the image into relevant terms. The end result is the same, a report that has pre-clinical or clinical relevance to the entire team of scientists and doctors.
The elegance of digital pathology is truly gained when all of these elements can be tightly integrated into a work flow and can likely be automated. Once all the steps are integrated within themselves, they should ultimately be integrated to the institution(s) information system so that data is generated as efficiently as possible and made available to as wide an audience as is necessary. This way the physical work can be centralized and the data can be made global.
The final element is crucial to the power of digital pathology. Sharing and collaborating on cases worldwide is where the real power lies. Scientists and clinicians from the institution of sample source or from other remote outposts with specialists can view the images, analysis, and clinical data simultaneously while deliberating on the results and what they mean. All the collaborators will need is an internet connection and their expertise.
Digital pathology is an environment in which a sample slide is scanned, viewed, managed, analyzed, integrated, and shared. The power in this process is that it can be automated and centralized which allows for faster processing, reduced cost, increased system uptime, and more expertise brought to bear on the analysis. With the potential for cloud based applications and storage, the scope of digital pathology can be spread to supoort researchers and clinicians across the globe.