August 13, 2018 – Tissue Image Analysis is a 3-step process: Histology Lab, Digital Pathology (DP) and Pathology AI (Artificial Intelligence).
Histology laboratories create histology slides by processing, embedding, sectioning, and staining tissue specimens taken from biopsies.
Digital Pathology enables a very simple global service model for tissue image analysis. Histology slides are digitized to images using slide scanners. The histology slides and the images are created and remain at the local histology labs. No need to ship tissue or histology slides! Copies of the images can be shared conveniently and securely around the world and transferred to data centers for tissue image analysis. No need to create data centers at local lab sites (minimizing costs and burden on local IT)!
Pathology AI provides tissue data by analyzing images of histology slides using image analysis and machine learning. Since histology slide images are very large and processing can be quite complex, tissue image analysis needs to be performed at High Performance Computing (HPC) data centers. A central single HPC data center on-premises (most cost effective) can support a global tissue image analysis service provider from an operational perspective. However now that we are in the days of virtualization and cloud computing, data centers can easily scale to distributed data centers and data centers in the cloud if needed.
The Pathology AI system itself needs to provide a convenient Digital Pathology interface for transferring images, running tissue image analysis, and viewing and managing the data. Histology labs only need a slide scanner to provide images of the histology slides.
A key challenge for this model is the regulatory environment, which depends on both the country (e.g. US, Europe, China) and the intended use of the tissue data (e.g. research, Clinical Trials, or clinical Diagnostics (Dx), Prognostics (Px) or Companion Diagnostics (CDx)).
Tissue image analysis service providers need to have a thorough understanding of how to run a histology lab, Digital Pathology and Pathology AI in a regulated environment (e.g. CLIA, CAP, FDA single-site medical device in the US). They must have an established Quality Management System (QMS) with the appropriate Standard Operating Procedures (SOP).
Tissue image analysis service providers can scale from a single-site central lab model to a multi-sites model by providing any histology lab the appropriate SOPs and training to achieve a seamless integration between each site’s QMS.
Global service providers need to be able to operate internationally. For service providers already operating successfully in the US, which has high regulatory standards, scaling internationally is often straight forward. The service provider will need to work with regulatory experts in the other countries (many consultant companies are available) to comply with the local regulations and provide a localization of their QMS.
Digital Pathology is a game changer for how we think about a service provider setting up their global network of histology labs. With the adoption of Digital Pathology, virtually any histology lab in the world with a slide scanner can become part of a global tissue image analysis service provider’s network. It comes down to just establishing the global service provider’s SOPs. Sites around the world can quickly be added to the network, driven by business opportunities and strategic partnerships. There is no need to set up and operate a large global network of pre-selected sites. If you want to work with a specific histology lab, simply add it to the network.
What is your experience using Digital Pathology to share histology slide images internationally? Are you interested in global research, Clinical Trials, or Diagnostics (Dx), Prognostics (Px) or Companion Diagnostics (CDx)? Would you like to choose the histology labs to work with?
Flagship Biosciences is the leading full-service provider for tissue image analysis. With our core competencies in operations, science, technology and regulatory, we have created a blue print for a tissue image analysis operation that scales easily to multiple sites and internationally.
If you want to learn more about the technology behind tissue image analysis, check out our previous LinkedIn articles, where we presented a short lecture series about the different aspects about Pathology AI intended for a broad non-technical audience.
Holger Lange, PhD
Chief Technology Officer