Digital Pathology Solution Could Speed Up Prostate Cancer Diagnosis
Blog Jun 26, 2018 | By Anna MacDonald, Science Writer for Technology Networks
Histological slide (H & E stain at x300) showing prostate cancer. Credit: NIH (Creator: Otis Brawley)
ContextVision recently participated in the 5th Nordic Symposium on Digital Pathology, where they introduced their upcoming AI product, INIFY™ Prostate.
To learn more about ContextVision, their move into the field of digital pathology, and how INIFY™ could help benefit pathologists and patients, we spoke to Lena Kajland Wilén, Director Business Unit Digital Pathology.
Anna MacDonald (AM): Can you tell us about ContextVision, the company mission and its goals?
Lena Kajland Wilén (LW): ContextVision is a medical technology software company that specializes in image analysis and artificial intelligence (AI). We are the global market leader within image enhancement and a software partner to leading medical imaging manufacturers all over the world. Our technology helps doctors accurately interpret medical images, such as ultrasound, XR and MRI, a crucial foundation for better diagnosis and treatment.
As an industry pioneer for more than 30 years, we are significantly investing in R&D to develop new applications of the latest AI technologies and expanding into the growing digital pathology market.
Our vision is: Better healthcare for more people, and our mission is: Providing intelligent technology to improve medical imaging and empower healthcare. Our goal is to become a major player within decision support software for key cancers.
We are based in Sweden, with local representation in the U.S., Russia, Japan, China and Korea.
AM: What are some of the major challenges pathologists currently face, and how can AI solutions help?
LW: People today are living longer, while cancer treatments are becoming ever more successful. However, these positive developments also mean an overall increase in both screenings and routine follow-ups – and, in turn, a rapidly growing number of pathology samples.
Valuable time and effort could be saved by quickly identifying cancer areas right from the start, with swift, effective analysis tools – truly empowering pathologists in their everyday workflow.
AM: You recently launched your first digital pathology product, INIFY™ Prostate. Can you tell us more about the product, and the benefits it could bring to pathologists and patients?
LW: The product will automatically identify and classify areas with cancerous tissues in digitalized H&E stained whole slide images. The first version of the product will focus on identification, the next one to automatically suggest Gleason classifications.
The first version, INIFY™ Prostate, will automatically separate slides containing cancer from those that do not, before the pathologist has even opened the case. Thus, pathologists’ worklists will be presented in a worst-first order, meaning that the patient cases that contain most cancer – as well as the slides within each case – are immediately shown at the top.
The INIFY™ algorithms are based on the MasterAnnotation™ training tool – a unique, patent-filed method developed by ContextVision.
Drawing on vast amounts of high-quality annotated training data, confirmed by multiplex antibody staining, MasterAnnotation™ lays the foundation for an objective, fast method of accurately identifying and outlining cancer and non-cancer structures in prostate biopsies.
We are heavily focusing on creating a product that will truly support the pathologist´s daily work. I.e. a product that will save time, reduce inter and intra-variability, increase accuracy and confidence and create quantified reports.
Moreover, we aim for a clinically documented product from the start and have secured quality assurance and regulatory compliance throughout the product development.
AM: In addition to prostate cancer, what other cancers are you focusing on?
LW: We have ongoing research within breast, lung and colon.
AM: What barriers are there to greater adoption of digital pathology solutions?
LW: There are actually quite a lot of barriers, even though at the same time quite some drivers.
The barriers include the rather high costs and major change of workflow that usually is needed to transform the lab to digital. Moreover, since one does not really remove any steps, just add the scanning, the benefits lie more into what you can do when slides are digitalized than what you can save by removing workflow steps. The lack of standardization is also an issue, the scanner vendors all use slightly different file formats, however there is now a big push towards standardizing to DICOM.
AM: How big a role do you see digital pathology playing in the future of healthcare?
LW: I believe that the majority of pathology labs will be digitalized within the next 5-10 years. Driving forces include the possibility to send slides for review by sub-specialists, and/or to ease workload peaks by sending slides for remote reviews and of course, that you can use decision support tools on the digitalized slides which really can help the pathologist saving time, increasing accuracy & confidence and creating quantified reports.
Lena Kajland Wilén was speaking to Anna MacDonald, Science Writer for Technology Networks.