Scaling diagnostic AI across 13 NHS trusts

How BT and deepc helped the Eastern Diagnostic Imaging Network create a shared foundation for diagnostic AI, starting with lung CT scans and building for future use cases.

A doctor examining a patient with a stethoscope

The challenge

Growing imaging demand and workforce pressure meant EDIN needed more than one AI tool. It needed a shared foundation for AI across 13 NHS trusts.

The solution

BT provided project and delivery support to help deepc implement a vendor-neutral AI platform, giving trusts access to 50+ UKCA/CE-approved radiology applications through one comprehensive integration.

The result

EDIN now has a regional foundation for diagnostic AI. Its first use case supports lung nodule detection and tracking, with potential to release capacity equivalent to two full-time radiologists.

The challenge

Eastern Diagnostic Imaging Network (EDIN) brings thirteen NHS trusts together to coordinate imaging services, workforce planning, training, standards and shared equipment.

It is doing this at a time when demand for diagnostic imaging continues to grow. The Royal College of Radiologists reports that complex imaging activity is growing by 9% a year, while the clinical radiology workforce shortfall stands at 32%.

For patients, pressure on diagnostic services can mean waiting longer for a scan, results and the next stage of care. At the end of May 2026, more than 1.9 million people were waiting for a key diagnostic test in England.

The NHS is also looking to digital technology and AI to help increase diagnostic capacity. The government's 10 Year Health Plan for England sets an ambition to make the NHS the most AI-enabled health system in the world. Plans to expand lung cancer screening will also increase demand for CT scans.

AI is already being used to support some diagnostic services.Early national data shows radiologists using AI-supported chest X-ray tools can cut analysis times for the most complex cases from eight days to four. More than four million patients have received a faster lung cancer diagnosis or all-clear using the technology.

Government investment is creating real opportunities for AI diagnostics, but that comes with new challenges.

Each application needs to work with existing clinical systems and meet requirements for data protection, governance and clinical safety. Introducing each application separately can mean repeating integration and governance work.

EDIN had secured funding through the NHS AI Diagnostic Fund (AIDF) to introduce AI to support the assessment of lung nodules on CT scans. Rather than creating an isolated solution, EDIN wanted a foundation it could build on.

BT and deepc proposed a shared clinical AI platform. It would meet EDIN's immediate need while providing a foundation for other diagnostic AI applications.

Our radiologists work incredibly hard and were stretched thin, with more cases than we had capacity for. We wanted to give them a second pair of eyes, a foundation that could support our workforce across every part of radiology, and something the whole region could build on for years to come. BT and deepc were the right partners to get there.”

Bryce Travers, EDINDigital Lead

The solution

EDIN adopted deepc's clinical AI platform, which can support AI applications from different suppliers through shared infrastructure.

deepc provides the clinical AI platform with data privacy, compliance and governance built in. BT brings project management and delivery support as well as its long track record delivering NHS transformation. Together, BT and deepc helped EDIN move from a targeted AI deployment towards a scalable approach to diagnostic AI across the region.

BT also supported local teams with training and guidance on adopting the technology.

The platform connects with the clinical imaging systems already used by participating trusts. These include picture archiving and communication systems (PACS), radiology information systems (RIS) and electronic patient records (EPR).

This means AI results can be available within clinicians' existing workflows rather than through separate systems.

The solution also pseudonymises patient data within the trust environment before it is processed by the AI platform.

Starting with lung CT scans. The first use case focuses on pulmonary nodules, which can be an early sign of lung cancer.

Measuring lung nodules manually takes time, and measurements can vary between radiologists. EDIN introduced Coreline Soft's AVIEW technology to support radiologists with nodule detection, measurement and comparison across scans.

The technology automatically provides measurements that would otherwise need to be completed manually. It can also compare nodules with previous scans, helping radiologists track changes over time.

This gives EDIN a way to support its immediate lung CT requirements while using the same infrastructure as a foundation for other AI applications.

The results so far

EDIN now has a shared clinical AI platform designed to support diagnostic services across its 13 trusts.

For the first lung CT use case, AI acts as an additional tool for radiologists by helping to detect, measure and track pulmonary nodules across scans. Clinical decisions remain with healthcare professionals.

Automated 3D measurement can be completed in around one-tenth of the time required for manual 2D measurement.

Early estimates suggest the solution could release capacity equivalent to two full-time radiologists.

Because AI results feed into existing clinical imaging systems, radiologists can access the information as part of their normal workflow rather than moving between separate systems.

The benefit extends beyond this first use case. EDIN now has infrastructure that can support additional diagnostic AI applications without creating a separate integration from scratch for every tool.

Through deepc’s clinical AI infrastructure and BT’s delivery support, the EDIN now have access to more than 50 diagnostic AI applications through the same infrastructure. This allows the Network to establish regional capability and more flexibility to add new tools as its clinical need change, without rebuilding the underlying integration each time.

We saw the opportunity not just to introduce AI into a single pathway, but to create the foundations for the future of diagnostics across the region. It’s a brilliant example of what a platform approach can deliver. The Network gets value today, and the flexibility to adopt new innovations as healthcare needs continue to evolve.

Professor Sultan Mahmud
Ready for future AI use cases

EDIN can use the same infrastructure as it considers other AI applications.

  • Potential use cases include:
  • fracture detection
  • chest X-ray prioritisation
  • head CT triage
  • prostate MRI
  • bone health analysis

Each new use case would remain subject to the appropriate clinical, regulatory, information governance and safety requirements.

The approach can also support services delivered through Community Diagnostic Centres, helping NHS organisations use shared technology across different diagnostic settings.

Instead of building the technical foundation again each time, EDIN has infrastructure it can develop as its clinical needs change.

Looking ahead

What started as funding for one AI use case has given EDIN a shared capability that can support diagnostic AI across the region.

The project started with a specific need: introduce AI to support lung CT analysis. BT and deepc used that opportunity to create infrastructure that can support further applications across 13 trusts.

deepc provided the clinical AI infrastructure, with clinical safety and NHS governance designed in from day one, while BT provided project management and delivery support, helping to coordinate implementation across the Network.

This gives EDIN a consistent foundation for introducing further AI applications as clinical needs, technology and evidence develop.

For other NHS organisations considering diagnostic AI, Phase 2 of the AIDF funding is now open. EDIN offers a practical example of how to meet an immediate clinical need while building infrastructure that can support future use cases.

Planning diagnostic AI across your trust or imaging network?

Email our healthcare team about building the infrastructure to support it.