Digital Pathology: A Case of Enterprise Imaging and Integrated Diagnosis
By Inderpal Kohli, AVP IT Clinical Applications and Training, Hospital for Special Surgery
The COVID-19 pandemic has accelerated the digital transformation of healthcare both for clinicians and patients, including in the area of clinical imaging. There is, however, one set of clinical imaging that is still analog for the most part and could be better incorporated into the enterprise imaging strategy – and that is anatomic pathology.
Anatomic pathology has primarily been based on preparing glass slides of tissue samples for analysis under a microscope. It is essentially a manual process requiring co-location or near-location to process tissue samples and perform diagnosis efficiently. A review of slides with another clinician requires sitting across the table with a multiheaded microscope; Collaboration across institutions for research and consults involves couriers and time delays. And, teaching and conference presentations involve physical transportation of slides and the setup of microscopes.
The Digital Pathology Landscape
Digital pathology is the process of digitizing glass slides using a whole slide image scanner and a software viewer on a computer monitor to analyze pathology images.
Unlike digital radiology, digital pathology still has a manual component in that tissues on a glass slide need to be scanned for digitization. Another difference is that radiology has a standardized file format called DICOM instead of the multiple formats used in pathology.
That said, digitizing slides and image availability in the enterprise system offers many benefits, including enhanced multidisciplinary collaboration, research, and teaching to name a few. While the practice of using digital images and workflows for pathology diagnosis is well established in Europe and other parts of the world, we are now seeing digital pathology take off worldwide due to the advent of faster whole slide imaging. In the US, the Food and Drug Administration also approved software and scanner combinations for primary diagnosis, which acts as a catalyst for the increased digitization.
Benefits of Digital Pathology at the Enterprise Level
Digital access to all patient images will lead to more informed analysis and a better clinical diagnosis. Institutions that plan to implement an integrated diagnostics setup will benefit from the correlation of pathology and radiology images. Integrated diagnostics enables all imaging to be housed either in a single PACS system or within the enterprise EMR. At HSS, we utilize Sectra PACS for both radiology and pathology images.
A digital capability also decouples the pathologist’s physical location from the histology lab, especially when providing pathology services to remote and rural areas. Furthermore, enabling pathology services beyond physical proximity is helpful as the overall number of pathologists in the United States decreases over time. It also enables efficient collaboration across institutions and industries for research and multisite studies, eliminating the physical transfer of slides.
The software solution also offers image analysis tools to create efficiency and reduce manual tasks such as cell counting or percentage cancer tissue calculation, providing a standardized image analysis across the institution, thus eliminating the discrepancy between diagnoses.
How to Optimize Infrastructure and Integration?
The software solution and infrastructure should support server-side rendering of large images, the lack of which creates a suboptimal experience for the pathologist. Plan for significant storage capacity as the size of pathology images can range between 500 MB- 2 GB, dependent on the magnification.
Consider integrating pathology imaging as part of enterprise imaging infrastructure to realize maximum benefits. The availability of pathology imaging from the same portal/viewer as other imaging eliminates the need for additional training and helps with clinical adoption. The solution should support the most prevalent file formats (svs, ndpi, etc.) while waiting for the DICOM standard to gain traction in pathology imaging.
If a single vendor solution is an option, institutions can benefit from extending the existing PACS infrastructure, knowledge, and administration for digital pathology. A LIS/EMR integrated solution would not require a separate tool outside of the standard workflow, eliminating a significant change for the pathologist.
After a successful pilot, HSS deployed an FDA-approved Digital Pathology solution with its pioneer integrated diagnostics workflow on Feb 11th, 2021.
Integrated Diagnostics at HSS
HSS has taken a leadership role in partnering with our digital pathology and enterprise imaging partners, Sectra and Epic EMR, utilizing the Leica Scanner to cut the trail for the integration path for half of the nation’s healthcare systems. We did comprehensive development work that other organizations on Epic can easily leverage to enable integrated workflow.
Intending to enable an efficient diagnostic process while maintaining a familiar workflow for pathologists, we invested in developing the unique integration between Sectra and Epic to allow image launching from the pathologist worklist, which was implemented in February 2020. After a successful pilot, HSS deployed an FDA-approved Digital Pathology solution with its pioneer integrated diagnostics workflow on Feb 11th, 2021.
How are we doing?
HSS started with a modest target with an incremental increase, but we have already surpassed our initial goal, with 70% of all cases being scanning and diagnosed digitally. Now, clinicians can access digital pathology images using a familiar portal to co-relate with other images while reviewing patient data, which was impossible until a few months ago.
What does the future hold?
The technological evolution of digital pathology is just getting started, and one noticeable advancement includes the use of machine learning for image analysis. Deep learning or AI algorithms can now analyze complex visual features in image data, assisting pathologists by automating the tedious tasks of slide image analysis.
With the advances in image analysis and the ability to provide quantitative information about biological samples, digital pathology will help pathologists make more accurate and consistent diagnoses, leading to improved patient outcomes.
