Ministers say new scanners and AI delivered a million extra NHS scans this year. The release's own figures credit the million to new machines; AI's full effect, even by 2029, is a sixth of that.
Estimated reading time: 5 minutes
In short. The Department of Health and Social Care's 3 November 2026 release says "one million more scans were delivered thanks to new scanners and cutting-edge AI technology" [1]. The release's own breakdown attributes that million specifically to community diagnostic centres, which carried out "nearly 2.2 million CT and MRI scans in the 12 months up to March 2026," a figure it says is "an extra million compared to the 12 months up to March 2024" [1]. AI acceleration software, installed in "more than half (54%)" of compatible MRI scanners, is credited separately, and its own stated national ceiling, once "every NHS trust" completes an upgrade due by 2029, is "up to 162,000 more checks per year" [1], a sixth of the headline figure. The release is titled around catching cancer sooner, but the only waiting-time figure it quantifies, a 250,000 fall in the elective care list, covers every condition, not cancer specifically. Two documented mistakes: four stars.
The new Manchester and Trafford Community Diagnostic Centre, a £19.2 million facility, opened its doors on 29 October 2026 [1]. Four days later, the Department of Health and Social Care used it as the backdrop for a national claim: that new scanners and artificial intelligence, working together, are now delivering a million extra CT and MRI scans every year, helping catch cancer sooner. The release bundles two different interventions into one headline number. Only one of them, on the release's own figures, has actually produced it.
The claims, tested
| The release's own words | What the release's own breakdown shows | Verdict |
|---|---|---|
| Minister of State for Health Karin Smyth: "The 125 new scanners and rollout of AI technology are helping deliver around a million extra scans every year" [1] | "Most of the new machines added since March 2024 (107 of 125) have gone into community diagnostic centres, which carried out nearly 2.2 million CT and MRI scans in the 12 months up to March 2026," an extra million against the year to March 2024 [1]. AI acceleration software, separately, is installed in 54% of compatible MRI scanners so far and would add "up to 162,000 more checks per year" nationally only once every trust finishes upgrading, a target set for 2029 [1] | The million is the new machines' own number; AI's own disclosed national effect, even finished, is a sixth of that and is not yet finished |
| Headline: "One million more scans to help catch cancer sooner." Body: "Cancer waiting times are better than last year" [1] | The only waiting-time figure given anywhere in the release is "the overall waiting list for elective care is down by over 250,000 since July 2024" [1], a whole-system measure covering every elective condition, not cancer specifically. No cancer-specific waiting time figure appears in the release at all | A release titled around catching cancer sooner backs its one measurable improvement with a number that is not about cancer |
A number that belongs to the scanners, not the software
The release's "Notes to Editors" source the headline figures to "NHS England's 2025/2026 National Imaging Data Collection," due to be published the same morning [1]. What it discloses is specific: 125 more CT and MRI scanners in service since March 2024, 107 of them installed in community diagnostic centres, and those centres' own annual scan count rising from roughly 1.2 million to "nearly 2.2 million" over the same period [1]. That is where the "million extra scans" figure comes from: more machines, doing more scans, in more buildings.
AI is described in the same release, but as a separate, smaller and still-incomplete programme. "AI-powered MRI technology is being introduced across the NHS, helping services carry out on average 14% more scans while maintaining image quality," and the acceleration software behind that figure "has already been installed in more than half (54%) of compatible NHS MRI scanners" [1], meaning nearly half of even the scanners that can run it do not yet have it. The release's own ceiling for what the technology could eventually add, once every trust has had the "opportunity to upgrade compatible MRI scanners by 2029," is "up to 162,000 more checks per year" [1]: a future, upper-bound, nationwide figure that is a sixth the size of the million the release puts in its headline.
Minister Karin Smyth's quoted line, "the 125 new scanners and rollout of AI technology are helping deliver around a million extra scans every year" [1], joins the two together as if they were two legs of the same achievement. The release's own numbers put almost all of that achievement on one leg. AI technology, by the document's own account, is still being rolled out to the other half of the compatible fleet, and its own disclosed effect even when finished does not come close to a million.
A cancer claim with no cancer number
The release's headline and framing are specific: "One million more scans to help catch cancer sooner," with CT and MRI scanners described as "a vital part of diagnosing cancers including lung, brain, breast and prostate cancer" [1]. NHS England's national clinical director for cancer, Professor Peter Johnson, is quoted saying "the earlier cancer is diagnosed, the more treatment options there are and the better the chance of a successful outcome for patients" [1].
When the release turns to evidence that waiting is actually falling, the only figure it gives is this: "cancer waiting times are better than last year, and the overall waiting list for elective care is down by over 250,000 since July 2024" [1]. The second half of that sentence carries the only number. It is the NHS elective care waiting list in its entirety, every condition a patient might be waiting for treatment on, not a cancer diagnosis or cancer treatment figure. "Cancer waiting times are better than last year" is asserted without a percentage, a day count or a reference to any of the NHS's own published cancer waiting time standards. A reader is given one hard number in a release about catching cancer sooner, and that number is not about cancer.
Credit where due
The release is specific and checkable where it counts most: it names Manchester and Trafford's new centre, states its cost, £19.2 million, and its opening date, 29 October 2026, rather than describing investment in the abstract [1]. It is honest that AI acceleration software is installed in "more than half," not all, of compatible scanners, and frames the 2029 full rollout as a future "opportunity" for trusts rather than claiming it is already delivered [1]. The 125-scanner and community diagnostic centre figures are specific, dated and internally consistent with each other, which is more than some government press releases manage.
The star score counts two documented mistakes. First, the release credits "the 125 new scanners and rollout of AI technology" jointly for "around a million extra scans every year," but its own breakdown attributes that million to community diagnostic centres fitted with new machines, while AI acceleration software, installed so far in 54% of compatible MRI scanners, has a disclosed national ceiling of 162,000 extra checks a year once every trust finishes upgrading by 2029, a sixth of the headline figure and not yet achieved. Second, a release titled and framed around catching cancer sooner gives only one quantified waiting-time improvement, a 250,000 fall in the entire elective care waiting list since July 2024, which covers every condition, not cancer specifically; no cancer waiting time figure appears anywhere in the release. This piece finds no wrongdoing by DHSC, NHS England or any named individual; Karin Smyth and Professor Peter Johnson are quoted only in their ministerial and clinical capacities, from the department's own published release. The release was fetched directly on 3 November 2026 and returned HTTP 200.
Sources
- One million more scans to help catch cancer sooner, Department of Health and Social Care and NHS England, published 3 November 2026