FDA Clears AI to Quantify Liver, Visceral Fat on MRI

SwiftSight Body CompositionWhat to know

30-second read

For Radiology · Gastroenterology · Endocrinology

Who

Any adult referred for an abdominal or whole-body MRI, including self-pay preventive/longevity-imaging programs — not ordered for a specific liver or metabolic diagnosis.

What it detects

Visceral and subcutaneous fat, abdominal/thigh muscle volume and quality, and liver fat fraction from one sub-5-minute MRI sequence, scored against age- and sex-matched norms.

Evidence

Cleared via 510(k) substantial equivalence, not a published accuracy trial. Built on PDFF, a validated biomarker: ~87% sensitivity, ~86% specificity for steatosis vs. biopsy.

Workflow entry

Layered onto scans already being acquired by imaging networks — SimonMed first — not a new order a physician places or a result arriving through a usual referral.

Key consideration

ACR says evidence is insufficient to recommend whole-body MRI screening in asymptomatic adults; AASLD still favors FIB-4 and case-finding over incidental imaging for MASLD.

Availability

Cleared 5 October 2026. Category III CPT codes exist for multi-organ MR tissue-composition analysis (0697T/0698T) but carry no fixed Medicare fee; most access is self-pay.

Practice impactNo change to practice yet

A well-engineered quantitative layer on scans many patients are already paying for, not a new screening test. An abnormal liver- or visceral-fat number deserves the same work-up any incidental finding would get — not a shortcut around one.

Read the full physician analysis

Full analysis · 10 min readAIRS Medical clearance notice, 5 Oct 2026 · ACR total-body MRI statement · AASLD MASLD guidance

The U.S. Food and Drug Administration cleared AIRS Medical’s SwiftSight Body Composition on 5 October 2026, AI software that turns a single, under-five-minute MRI sequence into a quantified read on a patient’s visceral fat, subcutaneous fat, muscle and liver fat, benchmarked against age- and sex-matched reference populations. AIRS Medical named SimonMed Imaging, the network already running its SwiftMR reconstruction software, as the first customer — folding the new report into SimonMed Longevity, its nationwide self-pay whole-body MRI screening service.

What one MRI scan now reports

SwiftSight Body Composition does not add a scan; it adds a reading to one that is often already being done. AI segmentation takes the images from a single sub-five-minute MRI sequence and quantifies visceral and subcutaneous adipose tissue, abdominal and thigh muscle volume and quality, bilateral muscle balance, and liver fat fraction. Those figures are set against age- and sex-matched reference populations and assembled into a Body Composition Report with percentile rankings, visual summaries and a standardized liver-fat severity scale — the kind of structured output radiologists can drop into a report and patients can read for themselves.

The number that will draw the most questions: liver fat

Of everything in the report, the liver-fat fraction is the one most likely to reach a physician’s desk. Metabolic dysfunction-associated steatotic liver disease (MASLD, formerly NAFLD) affects more than a quarter of the global population, most of it undiagnosed until advanced stages. MRI-based proton density fat fraction (PDFF) has become the reference imaging biomarker for hepatic steatosis, replacing biopsy in research and showing strong diagnostic performance in independent validation: a Youden-based threshold of 4.4% liver fat identifies MASLD with roughly 87% sensitivity and 86% specificity against histology, with far better inter-reader reproducibility than biopsy itself.

The biomarker SwiftSight is built on

25%+of the global population is estimated to have MASLD, most of it undiagnosed

4.4%PDFF threshold separating normal liver fat from MASLD in independent validation studies

87% / 86%sensitivity / specificity of that threshold against liver biopsy

Figures describe MRI-PDFF, the published reference biomarker for liver-fat quantification by MRI — not a SwiftSight-specific accuracy study, which AIRS Medical has not published. FDA cleared the device on substantial equivalence, the standard the 510(k) pathway requires.

That distinction matters for how much weight a physician should put on any single number. SwiftSight’s clearance establishes that its measurements are substantially equivalent to existing quantitative MR body-composition software, not that it has been validated against biopsy in its own trial. The underlying technique it draws on is well established; the product’s own accuracy, reproducibility and agreement with reference methods have not been published independently as of this clearance.

Who actually orders this scan

The more unusual fact for a physician to register is how most patients will encounter this report at all. SimonMed launched SimonMed Longevity in January 2026, a nationwide division offering whole-body MRI on 3T scanners, self-pay, starting at $899 and covering “100+ body structures” in one visit — expanding from 30 to more than 70 sites by the end of the first quarter. SwiftSight Body Composition is the first product AIRS Medical has cleared to run inside that pipeline. A patient does not get this liver-fat number because a physician ordered a test to investigate it; they get it because they paid for a preventive scan, and the AI read it along with everything else.

That is a different referral pattern than the one most specialists are used to triaging. A radiologist reading the underlying MRI has to decide how — or whether — to flag a body-composition finding that was not the indication for the study. A primary care physician, endocrinologist or gastroenterologist may see the number for the first time in a patient’s hand, arriving with no requisition, no pre-test risk assessment and no clinical question attached.

Where SwiftSight sits among MRI body-composition tools

SwiftSight is not the first MRI body-composition software FDA has cleared. AMRA Medical, a Linköping, Sweden company, won clearance for AMRA Profiler in 2021 for the same category of measurement — visceral and subcutaneous fat volumes, liver fat fraction and lean-tissue volumes from a whole-body MRI scan — and has been in clinical use in Sweden, the UK and Germany for years. The two differ mainly in scan time and where they are deployed, not in what they claim to measure.

Two cleared paths to the same report

	SwiftSight Body Composition	AMRA Profiler
U.S. status	Cleared 5 Oct 2026; deploying first through SimonMed Longevity	Cleared Nov 2021 (K211983); in clinical use
Scan	Under 5 minutes, added to an MRI already being acquired	About 6 minutes, dedicated whole-body protocol
What it measures	Visceral/subcutaneous fat, abdominal & thigh muscle volume and quality, liver fat fraction	Visceral/subcutaneous fat volume, liver fat fraction, thigh muscle volume
Typical access route	Self-pay preventive/longevity imaging networks in the U.S.	Clinical and research use in several European health systems
Published accuracy data	Not published by AIRS Medical as of this clearance	Validated against reference methods in independent research literature

AMRA Profiler particulars from AMRA Medical's FDA clearance coverage and company materials; prescribe clinical weight from each product's own validation record, not this table.

What radiology and hepatology societies still say

The clearance does not come with a professional-society endorsement of the pathway it mostly travels through. The American College of Radiology said in an April 2023 statement that it does not believe there is sufficient evidence to recommend total-body MRI screening for patients with no symptoms, risk factors or family history suggesting disease, citing the lack of documented cost-effectiveness or survival benefit and the risk that non-specific findings drive unnecessary follow-up testing and expense. AASLD’s own practice guidance on MASLD favors a different first step entirely: FIB-4 risk stratification in primary care for patients with cardiometabolic risk factors, prediabetes, obesity or abnormal liver enzymes, reserving imaging and specialist referral for those an indeterminate or high FIB-4 score flags for further workup — not incidental liver-fat quantification on a scan ordered for something else.

The billing question

Two new Category III CPT codes, 0697T and 0698T, took effect 1 January 2026 for quantitative MR analysis of tissue composition across multiple organs — 0697T when no diagnostic MRI of the same anatomy is performed in the same session, 0698T as an add-on when one is. Category III status means these are tracking codes, not guaranteed payment: there is no national Medicare fee schedule for them, coverage is carrier-by-carrier where it exists at all, and commercial payers can and do treat the analysis as investigational. For now, that uncertainty mostly stays out of the physician’s office: SimonMed Longevity is a self-pay program, so the first wave of SwiftSight reports will arrive already paid for, with no claim for a payer to deny and no code a referring physician needs to know — until the report crosses into a workup that does get billed.

• Expect a finding, not a referral. A patient may bring an elevated liver-fat or visceral-fat number from a scan you never ordered, with no accompanying risk assessment.
• Triage it through the standard pathway. Cardiometabolic risk factors, liver enzymes and FIB-4 decide whether it warrants non-invasive fibrosis testing or hepatology referral — the report itself is not a diagnosis.
• Radiologists reading the underlying study face a new judgment call on how prominently to flag a body-composition finding that was incidental to the scan's original indication.
• Do not assume coverage. Category III CPT codes exist but carry no guaranteed Medicare or commercial payment; most current access is self-pay.
• This is not a guideline-endorsed screening pathway. Neither ACR for whole-body MRI nor AASLD for MASLD has recommended it as one.

How SwiftSight got here, and what is next

1. 2018

   AIRS Medical founded

   Seoul, South Korea; builds its first products around AI-enhanced MRI reconstruction (SwiftMR).

2. Nov 2025

   SwiftSight launches

   AIRS Medical's first quantitative-reporting product, starting with brain imaging.

3. Jan 2026

   SimonMed Longevity launches

   Nationwide self-pay whole-body MRI preventive-screening division, expanding toward 70-plus sites.

4. Jan 1, 2026

   New CPT codes take effect

   Category III codes 0697T/0698T for multi-organ quantitative MR tissue-composition analysis.

5. Jun 2026

   TA Associates invests

   Strategic growth investment in AIRS Medical, after adoption across 1,700-plus institutions.

6. Oct 5, 2026

   FDA clears SwiftSight Body Composition

   SimonMed Imaging named as the first customer to deploy it.

7. Ahead

   Deployment and coverage decisions

   No national Medicare coverage determination or society screening recommendation yet exists for this use.

Frequently asked questions

Does a SwiftSight report replace an MRI-PDFF study or liver biopsy ordered to work up suspected MASLD?

No. SwiftSight Body Composition was cleared on substantial equivalence to existing quantitative MR body-composition software, not validated head-to-head against liver biopsy or a dedicated MRI-PDFF protocol. Treat its liver-fat number as a flag worth confirming through the standard pathway (FIB-4, then imaging or biopsy as indicated), not as a diagnostic result on its own.

Will insurance pay for the body-composition analysis?

Usually not yet. Category III CPT codes exist for multi-organ quantitative MR tissue-composition analysis (0697T, 0698T), but Category III status means no national Medicare fee schedule and inconsistent commercial coverage. The first deployment, SimonMed Longevity, is a self-pay program; patients may arrive with a report their insurer never billed.

A patient brings in a whole-body MRI report with an elevated liver-fat reading. What now?

Work it up as you would any incidental finding of hepatic steatosis: cardiometabolic risk factors, liver enzymes, and a FIB-4 score to decide whether it needs non-invasive fibrosis testing or hepatology referral, per AASLD guidance. The report is a data point from a scan the patient chose, not a validated screening result from a pathway your specialty society has endorsed.