Establishing levels of standardization is an essential issue impacting the field of digital pathology. It is seen as one of the major challenges facing the industry today.
While the concept has been around for a significant amount of time, it has only been in recent years that it has shown rapid advancements due to technology being available to support the idea.
Image Credit: science photo/Shutterstock.com
Following these steps forward in technology, there has been a rapidly growing widespread adoption of the methodology that focuses on digitized whole slide images taken from microscopes, storing and analyzing them, and creating databases that can be shared with scientists around the world.
The automated analysis and sharing of images allowed by the method are helping scientists to more rapidly gain insights into the pathology of several diseases, facilitating the fast and efficient development of new and effective therapeutic approaches.
However, since modern digital pathology is still in its early days, it suffers from a lack of standards and regulations. For digital pathology to meet its full potential, standardization must be improved to ensure that the same high standards are upheld across the board.
Standardization will also ensure the uniformity of data that is being added into shared databases, allowing for it to be easily incorporated into analyses run by different teams worldwide.
Currently, there is not just one standardization challenge facing digital pathology. The industry needs to address standardization in terms of the equipment, the analysis software, the protocols, the imaging acquisition systems, and the color displayed by the digital slides. Below we discuss how each of these areas needs to be improved.
Achieving standardization of the color that digital slides display is perhaps the most important aspect of standardization in digital pathology.
However, improving on color standardization can be challenging because variation can arise for a variety of reasons, from variation in the protocols and practices used in each lab to variations in the capture parameters, such as the illumination and filters used, to the software that processes the images.
The resolution to color display standardization comes from establishing standards in these other fields.
Scanners used to produce whole slide images are made by a variety of companies, resulting in equipment that works differently, the main difference being if the scanner works via tile-based or line-based scanning.
The Digital Imaging and Communications in Medicine (DICOM) standards committee working group has laid out standards for tile-based scanning, helping teams across the globe use their equipment in the same way to produce comparable results.
Standards are also required for line-based scanning methods to achieve uniformity from this type of equipment also.
Other hardware, lighting equipment, in particular, needs to be considered. Currently, there are no standards that lay out guidelines for the illumination levels and filters that should be used across labs to ensure that the color display of digital slides remains uniform.
Establishing rules as to which filters and illumination levels should be used is vital to being able to easily and rapidly compare the vast number of images that are stored in the digital database.
Image acquisition systems
The interpretation and comparison of large numbers of slides collected by labs in different locations also rely on the imaging acquisition systems that are integrated with the vendor-neutral archives (VNAs) complying with guidelines, such as those that have been laid out by DICOM.
Without adherence to such standards, the acquired images would lack interoperability. This failure to use these images in useful ways provides a strong barrier to labs and health services adopting the technology, restricting its growth and preventing more data from being shared between professionals.
Differences in protocols can cause significant variation in the images produced, particularly in terms of the colors displayed, which makes it difficult to compare images produced by different labs.
There is a need for guidelines to be set out to reduce the variability in processes between labs. Ensuring that all labs are working, in the same way, reduces the variability seen in images that can result in them being difficult to analyze and compare, restricting the efficacy of digital pathology.
Moving digital pathology with standardization
There is a need for guidelines to be established to facilitate widespread standardization, calibrating the work being conducting at labs across the world, ensuring that the images being shared can be easily analyzed and compared, and adding value to research and healthcare processes.
Research is needed to determine what these guidelines should be. Once such guidelines are set, researchers and healthcare professionals should prepare their teams to adapt and update their processes so that they are in line with the guidelines.
In some cases, this may require an investment of time and resources to retrain staff. However, standardization of processes will add significant value to digital pathology.
- Niazi, M., Parwani, A. and Gurcan, M., 2019. Digital pathology and artificial intelligence. The Lancet Oncology, 20(5), pp.e253-e261. www.thelancet.com/.../fulltext
- Pantanowitz, L., Sharma, A., Carter, A., Kurc, T., Sussman, A. and Saltz, J., 2018. Twenty years of digital pathology: An overview of the road traveled, what is on the horizon, and the emergence of vendor-neutral archives. Journal of Pathology Informatics, 9(1), p.40. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6289005/
- Yagi, Y., 2011. Color standardization and optimization in Whole Slide Imaging. Diagnostic Pathology, 6(S1). https://link.springer.com/article/10.1186/1746-1596-6-S1-S15