Add us as a preferred source
Siân Phillips, consultant radiologist chief medical officer of Adaptix, argues that better imaging could improve the first diagnostic decision and protect capacity.
Every radiologist I know has watched the same story unfold over the past decade: demand for diagnostic imaging rises year on year while the capacity to deliver it struggles to keep pace.
A 2025 analysis of NHS England’s Diagnostic Imaging Dataset illustrates the scale of change. CT activity rose by 60% between 2016/17 and 2023/24 while MRI activity increased by 31%. This is not simply a post-pandemic effect, but a sustained increase in the amount and complexity of imaging modern healthcare requires.
Much of that growth is positive. Earlier and wider use of imaging in cancer, neurological and other pathways can improve diagnosis and treatment. GP direct access has removed some of the traditional bottlenecks between primary and secondary care, and Community Diagnostic Centres are bringing additional diagnostic capacity closer to patients.
But adding scanners is only part of the answer.
Imaging capacity depends on having the workforce to protocol, perform, interpret and act on examinations. The Royal College of Radiologists’ latest workforce census reports a 32% shortfall in UK clinical radiology consultants, with more than 2,300 additional radiologists needed to meet current demand. The College projects that gap could rise further by 2030. The result is not simply pressure within radiology departments but delays throughout patient pathways, including substantial reporting backlogs after imaging has already been performed.
Problems with capacity
Minor injury and emergency care provide a useful example of where pathways might be improved. Conventional X-ray remains the appropriate first-line examination for most suspected extremity fractures: it is fast, relatively inexpensive, low dose and widely available. But a two-dimensional image inevitably superimposes anatomy, so subtle fractures in complex areas such as the wrist, ankle and foot can be difficult to identify.
When an X-ray is normal or equivocal, but clinical suspicion remains, patients may be immobilised and brought back for follow-up, undergo repeat radiographs, or be referred for CT or MRI. Each option can be entirely appropriate, but each also consumes additional appointments, clinical time and, in some cases, scarce cross-sectional imaging capacity.
NHS Resolution’s review of missed-fracture claims illustrates some of the vulnerabilities in this pathway: In many cases, an appropriate X-ray had been obtained, but image interpretation and follow-up were points of failure. The lesson is not that radiography is unsafe; rather that safe pathways depend on the quality of information available when decisions are made, alongside robust reporting and safety-netting.
This is where digital tomosynthesis deserves consideration. It uses multiple low-dose X-ray projections to reconstruct sectional images, reducing the problem of overlying anatomy. In musculoskeletal imaging, this can provide more anatomical information than a conventional radiograph without requiring a CT examination.
It is equally important to be clear about what tomosynthesis is not. It does not replace CT or MRI. CT will remain necessary for complex trauma and some surgical-planning questions, while MRI is required where marrow, ligament, tendon, cartilage or other soft-tissue pathology is the principal concern. The potential value of tomosynthesis is more targeted: Resolving selected cases at the point of first presentation where conventional radiography does not provide enough confidence.
Published musculoskeletal literature provides an emerging evidence base for this role, including applications in occult fracture detection and scaphoid assessment. However, the evidence is not yet uniform across indications and much of it comes from relatively small or retrospective studies. That makes prospective pathway evaluation essential.
That distinction matters because the potential system benefit comes not from replacing cross-sectional imaging entirely but from avoiding some unnecessary escalation to it.
If a proportion of patients who currently progress from an equivocal radiograph to CT or MRI can instead have the clinical question resolved at first presentation using tomosynthesis, some cross-sectional imaging referrals could be avoided. That could shorten individual patient pathways while releasing CT and MRI capacity for patients who genuinely require those modalities.
We have modelled that potential impact using early-adopter data and assumptions around appropriate extremity imaging pathways. On that basis, national adoption could potentially release more than 1 million CT and MRI appointments annually in England.
This is a modelled estimate, not an observed NHS outcome, and it will need prospective validation in routine clinical pathways.
That validation should test what proportion of cases are genuinely suitable for tomosynthesis; how often it changes diagnostic confidence or management; how many CT and MRI referrals are actually avoided rather than delayed; and the effects on repeat attendance, reporting workload and overall cost.
Consistent with policy
This direction is consistent with wider NHS policy. The 10 Year Health Plan emphasises moving care from hospital to community and from analogue to digital, while GIRFT has focused on ensuring imaging is delivered in the right place and at the right time. Community Diagnostic Centres are one expression of that ambition. Another should be considering whether clinicians seeing patients first have access to the most appropriate imaging technology for the decisions they need to make.
There may be future applications for compact point-of-care 3D imaging beyond current orthopaedic and trauma pathways, including other community or resource-limited settings. Those possibilities require separate evidence, evaluation and regulatory approval and should not be confused with established capability today.
None of this removes the need to address radiology workforce shortages. But if imaging demand continues to grow faster than the workforce, capacity needs to be approached from both directions: Expanding services where they are required while safely reducing avoidable demand on CT and MRI.
Better imaging, available earlier and closer to the patient, could help us do both, improving the first diagnostic decision while protecting higher-end imaging capacity for the patients who need it most.
