FDA clears AIRS Medical’s SwiftSight body-composition software

AIRS says SwiftSight uses AI-powered MRI analysis to quantify muscle and adipose tissue from MR images.[2] Its measurements include visceral and subcutaneous fat, abdominal and thigh muscle, muscle quality, bilateral muscle balance and liver fat.[2] The software evaluates specific muscle groups, including the abdominal wall, psoas and paraspinal muscles, as well as the quadriceps, hamstrings and adductors.[2] For liver fat, AIRS says it uses proton density fat fraction, or PDFF, with automated region-of-interest measurements and visual mapping.[2]

AIRS designed the scan to take less than five minutes, with automated AI analysis following it.[2] The company says AI-driven segmentation quantifies the body-composition measurements, which are benchmarked against age- and sex-matched reference populations.[2] The resulting Body Composition Report combines quantitative measurements, percentile rankings and visual summaries.[2] The announcement describes a scan-to-report workflow intended for high-volume MRI operations, rather than a separate manual analysis process.[2]

AIRS says organizations can add the service without introducing another imaging modality or a manual analysis workflow.[2] The company also says the software supports a standardized approach across supported scanners and locations.[2] AIRS presents the service as a potential new revenue opportunity from MRI infrastructure already in operation, and says it can extend whole-body and preventive imaging, healthy-aging, weight-management and wellness programs.[2] It also points to MRI’s lack of ionizing radiation as a basis for repeat quantitative assessment over time.[2]

AIRS named SimonMed Imaging as the first customer to deploy the cleared software.[2] SimonMed, based in Scottsdale, Arizona, has its own whole-body MRI program, according to Radiology Business.[3] AIRS, which Radiology Business says is based in Seoul, described the clearance as an expansion of its quantitative MRI portfolio beyond brain imaging.[2][3] The company’s stated roadmap planned diffusion tensor imaging for November 2026 and quantitative joint applications for 2027.[2]