The diagnostic imaging industry is continuing to see significant technological innovation. Across CT, MRI, and ultrasound, OEMs are bringing innovations to market that don’t just improve image quality — they have the potential to change where, how, and by whom imaging is performed. Three key forces driving this transformation are: the pursuit of higher diagnostic precision, the imperative to expand access, and the deepening integration of artificial intelligence.
CT: Photon-Counting Technology Enters the Mainstream
The most significant recent innovation in CT is the arrival of photon-counting detector technology (PCCT). Unlike conventional CT, which converts X-rays into light before measurement, PCCT directly counts individual X-ray photons and records their energy. The result is higher spatial resolution, lower radiation doses, and more spectral information available in every scan.
Siemens Healthineers was first to market in this space. Since launching the NAEOTOM Alpha in 2021 — the world’s first commercial photon-counting CT — over one million patients have been scanned worldwide. In March 2025, the FDA cleared the expanded NAEOTOM Alpha class, adding the dual-source Alpha.Pro and the single-source Alpha.Prime, which brings this technology to a broader range of facility types. Siemens has committed $86 million to expanding photon-counting detector production at its Forchheim, Germany facility.
The competitive landscape is shifting. In November 2025, GE HealthCare submitted a 510(k) to the FDA for its Photonova Spectra system, built on proprietary Deep Silicon detector technology featuring 8-bin energy resolution. GE states the system is designed to capture up to 50 times more data than conventional CT models. Canon Medical Systems and NeuSoft have also displayed photon-counting systems, signaling a transition from single-vendor technology to a multi-vendor competitive market.
MRI: Helium-Free Systems and Portability Expand Access
In MRI, one stream of innovation is focused on removing the barriers that have historically limited where systems can be installed and operated — most notably, the dependency on liquid helium.
Conventional MRI systems require from 1,500 to 1,800 liters of helium for superconducting magnet cooling. A new generation of systems is eliminating that dependency, and all three major OEMs are now competing in this space.
Philips has the longest track record, having pioneered its BlueSeal helium-free magnet technology since 2018. With more than 2,000 BlueSeal 1.5T systems installed worldwide, Philips says the technology has saved over 6 million liters of liquid helium to date. At RSNA 2025, the company unveiled BlueSeal Horizon — the industry’s first helium-free 3.0T MRI platform — which encloses just 7 liters of helium permanently in the cryogenic circuit, eliminating the need for refills or vent pipes.
Siemens Healthineers’ MAGNETOM Flow Platform uses DryCool technology with just 0.7 liters of helium in a sealed-for-life design, requiring no helium refills and no quench pipe, all within a 25-square-meter footprint. In June 2025, the FDA cleared the Magnetom Flow.Ace, Siemens’ first helium-free 1.5T system, which the company says reduces annual energy consumption by more than 30% versus the prior generation. GE HealthCare also entered the space at RSNA 2025 with Signa Sprint with Freelium, a ventless 1.5T system using less than 1% of the helium required by conventional MRI.
In another innovation stream, portable MRI is gaining clinical traction. In June 2025, the FDA cleared Hyperfine’s next-generation Swoop system, an AI-powered portable brain MRI weighing just 630 kg that plugs into a standard wall outlet. Early clinical users have reported image quality comparable to head CT — a notable benchmark for a device operating at 0.064 Tesla.
Meanwhile, Philips received FDA clearance for SmartSpeed Precise, which the company describes as the industry’s first integrated dual-AI MRI solution, delivering up to 3x faster scanning and up to 80% sharper images across its 1.5T and 3T installed base.
Ultrasound: The Point-of-Care Revolution
One key area of Ultrasound innovation is defined by miniaturization and intelligence. Ultrasound form factors have expanded from cart-based systems to include smartphone-sized handhelds — exemplified by GE HealthCare’s Vscan Air, Philips’ Lumify, and Butterfly Network’s single-probe whole-body imager. These devices are making diagnostic imaging available at the bedside, in the ambulance, and in rural clinics where traditional imaging infrastructure doesn’t exist.
AI is a central differentiator. Embedded algorithms now assist with image acquisition, automated measurements, and real-time clinical decision support, lowering the skill barrier for non-specialist clinicians. At RSNA 2025, GE HealthCare showcased early-stage concepts for autonomous ultrasound workflows developed in collaboration with NVIDIA.
The Common Thread: AI Across All Modalities
AI is not a feature bolted onto imaging systems — it is becoming embedded in the core imaging chain. As of late 2025, the FDA has authorized more than 1,350 AI-enabled medical devices, with radiology accounting for approximately 77% of all authorizations — surpassing 1,000 radiology AI devices for the first time. Every major OEM now offers native AI capabilities: Siemens’ myExam Companion, GE’s AIR Recon DL, Philips’ SmartSpeed Precise, and Hyperfine’s Optive AI.
What These Innovations Signal
These are compelling innovations occurring across multiple modalities in the same time period. Photon-counting CT is transitioning from a one-vendor novelty to a competitive battleground. Helium-free and portable MRI are lowering installation and operating barriers, opening new addressable settings. Handheld ultrasound is redefining point-of-care diagnostics. And AI is the connective tissue accelerating all three.
With these innovations the diagnostic imaging landscape continues to evolve and provide better information, in more places, and in a format which is accessible to more practitioners and patients, paving the way to continued advances in healthcare delivery.
