Though artificial intelligence has long made a name for itself in imaging—with hundreds of programs able to automate separate steps during an exam—at this year’s annual meeting of the Radiological Society of North America (RSNA), the three giants of the industry have put forward platforms that they say weave AI end-to-end across the entire workflow.
During the conference in Chicago, GE HealthCare raised the curtain on its latest Signa MRI machines, which are powered by an AI ecosystem employed during pre-patient planning, scanning and reading.
Dubbed Signa One, the machines tap into a live camera feed of an exam room that assists with real-time patient positioning while also contactlessly synchronizing the scanner to the motion of the patient’s breathing for a clearer image.
“We’re dedicated to making every interaction smooth, fast and patient centric by using predictive technology that boosts efficiency,” Bryan Mock, general manager of GE HealthCare’s premium MR segment, said in a statement. “Signa One technology is designed to simplify and improve the efficiency of experienced users, while aiming to shorten the learning curve for new users.”
The platform will debut on two scanners: the Signa Bolt, a wide-bore, 3 Tesla system for diagnostics and clinical research; and the Signa Sprint, a 1.5 Tesla machine featuring the company’s Freelium permanently sealed magnet tech. GE HealthCare said the ventless Sprint can be installed virtually anywhere and uses 1% of the helium traditional scanners have required to cool their magnet.
Signa Bolt and Sprint, as well as Signa One, have all been submitted to the FDA for review, according to the company. Last week, GE HealthCare also announced that it sent an application to the agency for its proton-counting CT system, the Photonova Spectra.
Meanwhile, in Europe, GE HealthCare this week touted two CE mark approvals in nuclear medicine: one for its Omni total-body PET/CT scanner, designed for incorporating tracers and theranostic agents in clinical care and research, and another for its new 4D SPECT/CT system, the StarGuide GX.
The company said StarGuide GX is designed as an all-in-one system capable of imaging virtually all energy levels for general-purpose SPECT scans, as well as being compatible with future alpha-particle-based therapies currently under development.
Back in Chicago, Philips launched its Verida system, which it describes as the world’s first detector-based spectral CT fully powered by AI. Also under review at the FDA, Verida measures how tissues absorb different X-ray energy levels to offer clearer contrasts between the adjoining parts of the body that may appear identical in conventional CT scans, according to the company.
With its latest software, Philips said Verida can reduce energy consumption by up to 45%, in turn lowering radiation doses without compromising image quality—while also completing exams twice as fast, with the ability to conduct up to 270 exams per day. The system previously obtained a CE mark, and the company said it already counts more than 800 global installations of spectral CT systems.
“While photon-counting CT adds complexity, is yet to move from the research arena into clinical practice, Philips spectral CT has been a clinical workhorse for more than a decade and delivers comparable or better clinical outcomes, standing up to the most demanding throughput and at significantly lower total cost of ownership,” offered Philips’ CT business leader, Dan Xu. The company said the latest Verida is slated to become available in the U.S. and select markets in 2026.
Philips also unveiled the industry’s first 3 Tesla MRI scanner with a sealed helium system. Currently under development, the BlueSeal Horizon looks to bring the resource-saving approach to the more powerful magnet seen in widespread use.
The company said it has already installed more than 2,000 of its previous 1.5 Tesla BlueSeal systems worldwide, collectively saving more than 6 million liters of liquid helium since the program's development in 2018.
“With BlueSeal Horizon we’re freeing MRI from dependence on a valuable resource the world can’t replace and bringing advanced diagnostic capabilities to people previously out of reach,” said Ioannis Panagiotelis, Philips’ MR business leader. “But more than that, BlueSeal Horizon is an entirely new 3.0T innovation platform that combines breakthroughs in hardware with AI-powered software, eliminating the trade-off between imaging speed and precision and improving outcomes for both practitioners and patients.”
Additionally at the RSNA meeting, Philips introduced a series of AI programs aimed at improving cardiac MRI scans—including a one-click automation that sets up 14 views of the beating heart in less than 30 seconds.
The programs also aim to provide diagnostic-quality images without forcing the patient to hold their breath, as well as the ability to acquire complete cardiac data within the time of a single heartbeat in patients with irregular rhythms.
Siemens Healthineers also outlined its AI approaches to cardiovascular procedures and other image-guided therapies. Its new imaging chain, Optiq AI, aims to reduce radiation doses while providing clearer visuals for precise, real-time operations.
That includes live algorithms to reduce picture noise during the image processing step, across fluoroscopy and angiography. Optiq AI also automatically finds the optimal exposure parameters by adjusting tube voltage, current, focal spot size, pulse width and detector dose, while accounting for the distance between the device and the patient and any movements or realignments during a procedure.
Optiq AI is set to be available across three new Siemens Healthineers imaging platforms: Artis genio, Artis icono.explore, and Artis icono.vision/Artis pheno.vision. The company said they are designed to tackle the increasingly complex and minimally invasive procedures that are making their way into smaller anatomies.
“The growing need for earlier-stage treatments raises the bar for image quality, and this is where artificial intelligence comes into play. With Optiq AI, we are unlocking AI’s potential for a new generation of interventional systems and are bringing it to all clinical fields,” said Carsten Bertram, Siemens Healthineers’ head of advanced therapies.
Optiq AI and the floor-based version of Artis genio and icono.explore are awaiting an FDA 510(k) clearance; the ceiling-based versions of those systems, as well as Artis icono.vision and pheno.vision, are currently in earlier stages of development.
In addition, Siemens Healthineers presented an AI-powered radiology suite that spans scheduling to image generation to reporting—while also incorporating AI into its advisory services to simulate hospital departments and recommend improvements to providers.
Run on the company’s in-house supercomputer, ActExcell Operational Twin takes customer-specific data and aims to predict future scenarios and pathways to reducing patient wait times, among other targeted needs. That includes the consideration of factors such as staffing, facility design, layout and patient scheduling processes.
Finally, Siemens Healthineers announced new capabilities for its Mammomat B.brilliant system, with new image reconstruction techniques for contrast-enhanced mammograms that aim to provide a more uniform background—allowing areas suspected of harboring breast cancer to better stand out. The ClearCEM software also provides a contrast localizer image for tomosynthesis-guided biopsies.