Partnership to build on AI Pathology for detection of Cancer

Quantitative digital pathology and image management company Indica Labs and the Industrial Centre for Artificial Intelligence Research in Digital Diagnostics (iCAIRD) are collaborating on the development of an AI-based digital pathology solution for the detection of cancer within lymph nodes from colorectal surgery cases. The primary aim of the project is to develop a tool which in the future may improve the efficiency of pathology teams within the National Health Service Greater Glasgow and Clyde (NHSGGC) reporting colorectal cancer cases and the detection of metastatic cancer in lymph nodes.

Funded by a combination of Innovate UK and industrial partners and supported by the West of Scotland Innovation Hub, Glasgow-based iCAIRD is one of the largest healthcare AI research portfolios in the UK. A collaboration of 30 partners from across the NHS, industry, academia and technology, the programme is currently delivering 35 ground-breaking AI projects across radiology and pathology, having grown from just 10 projects at the outset in 2019. The mission of iCAIRD is to establish a world-class centre of excellence for implementation of artificial intelligence in digital diagnostics.

Anonymised H&E slides from NHS Greater Glasgow and Clyde’s digital pathology archive will be used to train, validate and test the algorithm, which is being developed collaboratively by iCAIRD and Indica Labs. The resulting algorithm will report negative and positive lymph node status and will be compared to pathologist reports. Furthermore, positively involved lymph nodes will be categorised into metastases, micro-metastases, and individual tumour cells.

Dr. Gareth Bryson, Consultant Pathologist at NHSGGC and Clinical Director for Laboratory Medicine of iCAIRD said: “Our belief is that AI powered decision support tools, such as the one we are working on, may help to support pathologists by improving the process’ efficiency, while simultaneously increasing sensitivity in detecting small metastasis – which will direct patient therapy. Colorectal cancer resections are one of the most common cancer resection specimens and a disproportionate amount of pathologist’s time is utilised in screening lymph nodes.”

Indica Labs, based in Albuquerque, New Mexico, offers a suite of digital pathology image analysis solutions including HALO AI and HALO AP, both of which will be utilised by Indica Labs and iCAIRD partners for the development of AI-based pathology solutions and their evaluation in an NHS digital pathology workflow.

More information online