Evolution & outlook
Field strength went up for forty years. Then it came back down.
Every return to lower field was not a technical retreat. It happened when the supporting technology caught up and made cost, access and practicality the deciding factors.
Part one · The past
A forty-year argument about field strength.
- 1970s
Resistive magnets
First-generation MRI uses resistive electromagnets, typically 0.05–0.35T. The principle is proven; the hardware is limited.
- 1980s
The superconducting turn
Superconducting magnets replace resistive ones. For four decades the industry moves up: 1.5T, then 3T, then higher.
- 1990s
Low field retreats to the basement
High field owns the flagship. Low-field MRI survives as permanent-magnet open systems, typically 0.2–0.5T — cheaper to buy and easier to house, but treated as entry-level machines rather than a direction of travel.
- 2012
0.35T cleared for radiotherapy guidance
A 0.35T system receives FDA clearance for MRI-guided radiotherapy; the first clinical treatments follow in 2014. Low field is back in clinical use — because the application suited it, not because it was cheaper.
- 2020
First bedside low-field system cleared
In February, the first bedside MRI system is cleared by the FDA at 0.064T with NdFeB permanent magnets, for brain imaging of patients aged two and up. In August the indication is extended to all patient ages. MRI finally reaches the bedside.
- 2021
Reconstruction moves into software
Deep-learning image reconstruction is cleared for the 0.064T portable system, improving the T1, T2 and FLAIR images acquired at the bedside. The image-quality gap starts closing in software rather than in hardware.
- 2024
Low-field enters the reimbursement path
In November, IAC updates its advanced imaging accreditation standards to include ultra-low-field portable MRI.
The pattern: a field-strength route never comes back because the technology regressed. It comes back when the surrounding technology catches up and reshuffles the trade-off between cost, access and practicality.
Part two · The present
How low field closed the image-quality gap.
The hardware got lighter
NdFeB and SmCo permanent magnets with Halbach, C-type and H-type open geometries removed cryogenics and cut weight and footprint by orders of magnitude.
The noise got cancelled
Sensing coils plus deep learning estimate and subtract ambient electromagnetic noise, so the shielding shrinks from a room built on site to a modular enclosure delivered in days.
The algorithms took over
CNN denoising, super-resolution, compressed sensing and parallel imaging trade computation for hardware signal-to-noise, cutting average scan time.
The sequences were rewritten
SPEN, MR fingerprinting and relaxation-aware parameter optimisation address the distortion that comes with the homogeneity limits of a permanent magnet.
Part three · What comes next
Five directions.
These are industry trends, not product claims.
Magnets keep shrinking
Permanent-magnet materials and geometry push magnet weight into the tens-of-kilograms range in prototype portables. Deployment logic shifts from building a room to placing a device.
AI moves from compensation to native design
Instead of imitating high-field output, sequences and diagnostic pathways are designed for low-field physics.
From adjunct to independent platform
In some settings — brain follow-up, post-operative orthopedic review, bedside monitoring — an independent diagnostic pathway becomes plausible.
Decentralized imaging
Imaging spreads from large hospitals to clinics, imaging centers, mobile units and remote facilities.
Imaging as a service
From selling a machine to selling an imaging capability: per-scan pricing, per-cycle service agreements.
Industry-level discussion only. No competing product is named or compared, and no forward-looking statement is made about our own regulatory or commercial outcomes.
Part four · The market
The market, in numbers.
Third-party market research, each figure with its report and forecast window. These are industry forecasts, not our own projections.
USD 1.8B (2025) → USD 3.9B (2034)
Global low-field (<1T) MRI market — CAGR 8.9% (2026–2034)
Dataintelo
USD 1.7B (2026) → USD 4.1B (2036)
Portable MRI systems market — CAGR 8.8% (2026–2036)
Future Market Insights
USD 991M (2026) → USD 3.3B (2036)
Portable brain MRI systems market — CAGR 12.9% (2026–2036)
Future Market Insights
USD 240M (2025) → USD 619M (2032)
Ultra-low-field MRI market — CAGR 14.52% (2025–2032) · North America 46.4% (2025)
PW Consulting
36.8%
Low-field share within the portable MRI market — 2026
Future Market Insights
USD 276.4M (2026) → USD 404.6M (2031)
Global veterinary MRI market — CAGR 7.92% (2026–2031)
Mordor Intelligence
62.0%
Below 0.3T share of veterinary permanent-magnet MRI — 2025. That segment: USD 210M (2025) → USD 340.5M (2032)
PW Consulting
44.62% (2024) · 44.15% (2025)
North America share of veterinary MRI — Largest single market
Mordor Intelligence
30,141
US orthopaedic surgeons on record — 2018 census — surgeons, not clinics
AAOS, Orthopaedic Practice in the US
Market figures are third-party estimates and are reproduced with the report name and forecast window. They describe the industry, not our own revenue, share or outlook.
Want the technical argument in full?
The whitepaper covers the same ground with references and figures.