The motivation stems from an uncomfortable reality in modern medicine: healthcare systems worldwide are responsible for an estimated 5.0 to 8.5 percent of global greenhouse gas emissions, and diagnostic imaging is a significant contributor. Magnetic resonance imaging is indispensable for visualizing soft tissue without ionizing radiation, but fixed MRI systems are voracious consumers of energy, requiring dedicated cooling, specialized electrical infrastructure and extensive shielding. They also generate distinctive waste streams: tons of copper and lead shielding, thousands of liters of liquid helium for their superconducting magnets, and gadolinium-based contrast agents that persist in waterways. Until now, no study had rigorously compared these burdens against the newer generation of portable scanners in the settings where they promise the greatest impact.

The research team focused on Weeneebayko General Hospital in Moose Factory, Ontario, a fly-in community on the James Bay coast that serves roughly 12,000 predominantly Indigenous people through the Weeneebayko Area Health Authority. The hospital has on-site CT and ultrasound but no conventional MRI, so patients needing brain scans are flown to tertiary centers such as Kingston Health Sciences Centre, roughly 800 kilometers to the south. A round trip by air for a single patient generates approximately 879 kilograms of carbon dioxide. The team evaluated a 64-millitesla portable scanner, the only portable MRI currently licensed by Health Canada and commercially available for clinical brain imaging, building on a prior feasibility study at the hospital in which 25 consecutive patients received diagnostic-quality bedside brain scans.

The energy comparison was striking. The portable system draws a maximum of 1,650 watts for the entire unit, including its console, radiofrequency amplifier and three gradient amplifiers, consuming just 1.65 kilowatt-hours per hour of operation. Conventional 1.5 Tesla scanners use about 15.9 kilowatt-hours for a 49-minute head examination, while 3 Tesla units consume roughly 23.9 kilowatt-hours for a comparable study, figures drawn from published audits of radiology department energy use. That makes the portable scanner 11.8 to 17.4 times more energy-efficient per scan. In the study’s model of 100 scans per year, the portable unit would emit just 0.065 metric tons of carbon dioxide annually, equivalent to burning about 27.6 liters of gasoline, compared with the equivalent of 266.5 to 401.3 liters for fixed systems.

The biggest environmental win, however, came not from the machines themselves but from the flights they could prevent. Based on the earlier clinical feasibility data, 56 percent of patients scanned locally did not require transfer to a fixed-MRI center because the portable images were diagnostically sufficient for their condition, which included suspected stroke, hemorrhage, hydrocephalus, mass effect and midline shift. The researchers modeled how those 56 avoided round trips per year would distribute across the three air transport modes serving the community: commercial flights on De Havilland DHC-8 aircraft, charter flights on Cessna 208 Caravans, and air ambulance transfers on Pilatus PC-12 turboprops or Leonardo AW139 helicopters, each with distinct per-passenger emission factors. Avoided travel translated to roughly 14.3 tons of carbon dioxide per year, equivalent to the annual emissions of about 1.9 single-family homes. Notably, the twin-engine helicopter was the most carbon-intensive option per passenger, emitting 2.66 times more than the single-engine turboprop.

Waste analysis revealed further, if smaller, advantages. Conventional superconducting scanners rely on niobium-titanium magnets bathed in liquid helium, and a single scanner may require up to 10,000 liters over its lifetime, with about 2,000 liters held in the unit at any moment. Helium losses occur through boil-off, quenching and limited recycling, and in 2017 MRI operations accounted for roughly 30 percent of United States helium consumption. Helium liquefaction itself is energy-intensive, generating an estimated 0.4 to 2.6 tons of carbon dioxide equivalent over a scanner’s lifespan. Portable scanners instead use permanent neodymium-iron-boron magnets that need no helium at all and are lighter and more easily recyclable. A typical 1.5 Tesla suite also requires around 1,500 kilograms of copper shielding, embodying 5.0 to 7.2 tons of carbon dioxide equivalent, whereas portable systems need minimal shielding thanks to built-in radiofrequency noise cancellation.

The contrast agent story adds another dimension. Gadolinium-based contrast agents are chemically stable, resist degradation, accumulate in aquatic ecosystems and can enter food chains; nearly all administered gadolinium is excreted, with an estimated 21 tons discharged into United States wastewater each year. Each administration also produces 0.28 to 0.56 kilograms of carbon dioxide equivalent from single-use devices and packaging. Because the low magnetic field strength of portable systems limits gadolinium’s effectiveness, most portable scans proceed without contrast, so substituting 100 contrast-enhanced examinations would prevent roughly 28 to 56 kilograms of carbon dioxide equivalent annually. Taken together, the researchers estimated that eliminating helium use, reducing shielding and minimizing contrast waste would prevent an additional 5.4 to 9.8 tons of carbon dioxide per scanner over its lifecycle, equivalent to the yearly emissions of 0.73 to 1.3 homes.

Conservatively summing all three domains, the study calculates a total carbon reduction of 19.7 to 24.3 tons of carbon dioxide per year at the modelled scan volume, or about 2.6 to 3.1 single-family-home equivalents annually. The savings are deliberately modest in absolute terms, reflecting a single small catchment of roughly 12,000 people and about 50 to 100 examinations per year. But the underlying driver, averted long-distance medical transport, applies far more broadly. Approximately 22,400 on-reserve residents live in the 26 Nishnawbe Aski Nation communities in northern Ontario without year-round road access, generating roughly 2,100 emergency air transports annually, and Canada’s territories produce tens of thousands of medical-travel trips each year, with medevac rates of 32 to 53 per 1,000 residents in the Northwest Territories and Nunavut. Even if only a subset of those trips is amenable to substitution by portable imaging, replicated savings across dozens of remote communities could aggregate substantially.

The authors are careful to frame the environmental benefits as a co-benefit of a technology whose primary value is health equity, not as a replacement for conventional MRI. Portable ultra-low-field systems cannot match the spatial resolution, sequence variety or specialized capabilities of fixed scanners: they offer lower signal-to-noise ratio, longer acquisition times, and no routine contrast-enhanced, angiographic, perfusion or functional imaging. In the feasibility cohort, all 25 completed examinations were judged diagnostic quality, though five carried quality caveats such as motion artifact, and 44 percent of patients still required transfer. Crucially, when a portable scan detects an abnormality that warrants further characterization, transfer remains the clinically correct response, even though it adds travel. The technology is therefore best positioned as a local triage and diagnostic tool, reserving flights for patients who genuinely need advanced imaging.

The study also acknowledges its limitations: reliance on literature-derived emission factors rather than on-device metering or ticket-level flight data, uncertainty in annual scan volumes and case mix, and limited generalizability beyond the study region. Future work, the researchers suggest, should include multi-site life-cycle assessments comparing 64-millitesla, 1.5 Tesla and 3 Tesla systems across rural and urban programs, standardized aviation accounting, and investigations into rare-earth magnet recycling. On current evidence, though, the conclusion is compelling: a scanner the size of a small cart, plugged into an ordinary outlet in a fly-in community on James Bay, can deliver diagnostic-quality brain imaging, keep most patients home with their families, and meaningfully shrink the carbon cost of care, all at once.

Subject of Research: Environmental and carbon footprint of portable ultra-low-field MRI compared with conventional MRI in remote Canadian healthcare settings

Article Title: Environmental impact of portable MRI in remote Canadian settings

Article References: Azizi, N., Zhou, Z., Ibrahim, A., & Islam, O. (2026). Environmental impact of portable MRI in remote Canadian settings. Frontiers in Neuroimaging, 5, Article 1706184. https://doi.org/10.3389/fnimg.2026.1706184

Image Credits: AI Generated

DOI: 10.3389/fnimg.2026.1706184

Keywords: portable MRI, ultra-low-field MRI, carbon emissions, greenhouse gases, remote healthcare, Indigenous health, medical transportation, helium consumption, gadolinium contrast agents, health equity, sustainable radiology, life-cycle assessment