By Liam Canavan, Healthcare Lead at Loadbalancer.org
Liam partners with medical imaging providers and hospitals worldwide, building strategic alliances and using cutting-edge technology to tackle tough data storage problems and ultimately enhance patient care.
The ARPA-H INDEX programme and DPA have outlined a broad digital strategy for pathology, yet simultaneously acknowledge that infrastructure shortfalls are impeding progress. But should they be held partly accountable for these persistent challenges?
It is misguided to criticise pathology’s progress by comparing it to other diagnostic disciplines such as radiology and cardiology. Attempting to align pathology with their trajectory will not only impede its digital roadmap but also compromise efforts to address its substantial infrastructure shortcomings.
Fundamental differences
A customised infrastructure approach is essential for this unique discipline.
1. Physical not digital
Unlike cardiology and radiology, pathology’s historical dependence on glass slides means its journey to digitalistion is considerably longer and more complex.
2. Large images
The sheer size of pathology images vastly exceeds those in other diagnostic disciplines.
A single high-resolution pathology image can be 80 times the size of a radiology image, starkly revealing the immense inadequacy of current IT infrastructure for its true needs.
3. Workflow management
Liam Canavan
Pathology’s metamorphosis to digital is uniquely complex because its Laboratory Information Systems (LIS) manage everything from specimen processing to result validation.
This is a stark contrast to radiology’s more segmented approach, where Radiology Information Systems (RIS) handle workflows and reports, and Picture Archiving and Communication Systems (PACS) manage images independently.
Technical obstacles
While ARPA-H’s INDEX program and the DPA acknowledge that inadequate digital pathology infrastructure significantly impedes progress, they offer few concrete solutions to tackle these challenges.
Data storage
By a vast margin, digital storage is dominated by pathology. A large hospital, for instance, might require 1 petabyte (PB) of storage per year just for its pathology needs, necessitating storage solutions that are compliant, scalable, and high-performing.
Data fragmentation
Inconsistent file formats and fragmented data lead to significant hurdles, hindering data exchange, slowing workflows, and making cross-departmental communication more complex.
Data transfer
Due to their immense size, WSI files require substantial network bandwidth and high-speed connectivity. However, systems often struggle with these vast data volumes, resulting in slow transfers and frequent service interruptions.
The way forward
The allure of digital pathology is strong, yet we must critically assess our infrastructure’s ability to support it at scale. It’s pointless to follow radiology and cardiology’s lead if our current healthcare network infrastructure keeps hindering digital pathology’s advancement.
A collective issue
To move forward, interdepartmental collaboration is essential for sharing infrastructure and budgets to ease the storage burden. Furthermore, integrating with existing DICOM systems is vital for standardised image formats and interoperability, securing pathology data within a unified system for long-term universal benefit.
A new approach
Adopting digital workflows means altering deeply ingrained lab practices, making team buy-in crucial. We also need to overcome key obstacles such as the lack of real-time magnification control. This will require a complete redesign of pathology diagnostic processes to ensure accuracy and mitigate these disadvantages.
A workaround
Digital pathology, heavily reliant on Whole Slide Images (WSI), necessitates a robust network. High-throughput scanners alone demand 1 gigabit/s to database servers and 10 gigabit/s to storage. To manage these demands, data compression is vital. While lossy compression reduces file size, the potential void of diagnostic information makes lossless compression the preferred, albeit incomplete, solution.
A different race
The extensive digital infrastructure required for comprehensive Whole Slide Imaging (WSI) poses a unique hurdle for pathology. Ironically, the current intense focus on AI is actually hindering progress.
While a gradual, deliberate infrastructure build-out might not offer the immediate excitement of AI breakthroughs, enabling pathology to develop its digital imaging foundation from the ground up will ultimately create a more stable and beneficial ecosystem for everyone, and support future AI innovations.
