Completed Computing & AI

NPIC: National Pathology Imaging Co-Operative

In plain English

AI plain-English summary

Pathologists in 12 NHS hospitals across Northern England will begin scanning millions of microscope slides into computers to train artificial intelligence systems that can spot cancer. The problem is that digital pathology—converting glass biopsy slides into high-resolution images—produces data that AI needs to learn from, but no large collections of such images exist. Without them, AI cannot be trained to recognise cancerous patterns reliably. NPIC will install scanners in hospitals to generate roughly 760,000 images per year (1.2 petabytes of data), creating the first large-scale NHS digital pathology library. If the project succeeds, AI systems trained on this data could help pathologists diagnose cancer faster, more accurately, and at lower cost. That would directly affect how quickly patients receive a diagnosis and start treatment. The co-operative structure, led by NHS doctors, is designed to keep clinicians in control of how AI is deployed and to ensure public trust in how patient data is used. This is applied research with a clear practical goal: building the infrastructure and training data needed to make AI-assisted cancer diagnosis a routine part of NHS care.

View original technical description
"Pathologists are the **doctors who diagnose disease**. To do this, we examine samples (""biopsies"") using a microscope, to decide if a sample shows cancer, and what type of cancer it is. New scanners allow us to create digital images of microscope slides, so we can look at them on computers, and share them between hospitals more easily. This is called **digital pathology** - it will be a revolution in how we diagnose cancer. But digital images allow us to go one step further - to use computers to help us look at the images. ""**Artificial intelligence**"" (AI) is a computer technology which allows computers to do pattern recognition. AI can be trained to recognise the patterns of disease - for example searching for small areas of cancer in a large sample. So pathologists will be able to use AI to diagnose cancer faster, better and at lower cost. To work, AI needs to ""learn"" the patterns by looking at large numbers of images and learn from them, like a human would. But there are no large collections of digital images that can be used for this. In this project, the **Northern Pathology Imaging Co-operative (NPIC)**, we are bringing together the NHS, industry and scientists to solve this problem. We will put scanners in 12 Northern NHS hospitals to gather digital pathology images training AI systems. This will generate a lot of data - about 760,000 images per year by the end of the project, which is about 1.2 Petabytes. We will then work with industry and scientists to make new AI systems to analyse our images and make better diagnoses. This is a big opportunity for the UK to lead in this new area. Because it is a **co-operative with a shared goal, led by doctors**, this project will ensure that AI systems are safe, and that doctors and the NHS are in control of how they are used. Just as important, we need to ensure the public understand what we are doing and trust that we are using NHS data properly and securely - this will be an important part of our project. We hope that by the end of our project, we have used the funding to create something that is world leading. We want to develop the best systems for gathering and using data to make AI systems, bringing value for the NHS and patients."

View the original record at the funder ↗

Related Research

Grants with similar aims, by meaning.

MICA: InterdisciPlInary Collaboration for efficienT and effective Use of clinical images in big data health care RESearch: PICTURES
PathLEAD: Pathology image data Lake for Education, Analytics and Discovery.
AI-based diagnosis for improving classification of bone and soft tissue tumours across the UK
Integrating Clinical Infrared and Raman Spectroscopy with digital pathology and AI: CLIRPath-AI
Capacity, Confidence, Care - Using Artificial Intelligence and Machine Learning to support Breast Screening

Original classification

Collaborative R&D

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.