Skip to content


Automated MRI Measurement Of Liver Fat

HepaFat-AI automatically analyses MRI datasets to assess liver fat in patients, providing doctors with a comprehensive, multi-metric solution for use in the assessment of individuals with confirmed or suspected fatty liver disease. From one quick, non-invasive MRI scan, doctors have access to a suite of metrics relating to a patient’s liver health.

HepaFat-AI produces these patient results for clinician interpretation within seconds, providing accurate, reliable and reproducible results across MRI centres and scanner models.

HepaFat-AI is ideally suited for single and multi-center clinical trials in non-alcoholic steatohepatitis (NASH) and non-alcoholic fatty liver disease (NAFLD).

HepaFat-AI has received 510(k) clearance from the US Food and Drug Administration (FDA), European CE marking, and approval from the Australian Therapeutic Goods Administration (TGA).

HepaFat-AI can be deployed either in the cloud or on premises and can be integrated directly into radiology workflows. For current availability, please contact us directly.

Comprehensive all-in-one Reporting

HepaFat-AI performs fully automated and comprehensive all-in-one liver fat analysis and reporting:

  • VLFF – (Volumetric Liver Fat Fraction)
  • PDFF – (Proton Density Fat Fraction)
  • Steatosis Grade
  • Liver Fat Distribution Map

In the event that extremely high liver fat content is detected, the HepaFat AI analysis process will generate a different report with a warning message that indicates the value is outside the validated calibration range and that the associated uncertainty makes a precise estimate too unreliable for reporting.

In the event that high liver iron concentration is suspected, or that there is possible lung crest present due to incorrect slice position, a warning is given and no result will be shown on the generated HepaFat-AI report.

Steatosis Grading

Proton Density Fat Fraction (PDFF)

Volumetric Liver Fat Fraction (VLFF)

Key HepaFat-AI Features

Why Use HepaFat-AI to Measure Liver Fat

HepaFat-AI is quick, easy and painless, with an MRI scan time of only two minutes. HepaFat-AI does not require any contrast agents and being MRI based there is no risk from ionising radiation. Due to the automated nature of the technology, HepaFat-AI can also return a patient report in real-time back to the clinician.

Histopathological assessment of a patient’s liver fat levels from a liver biopsy is commonly considered the gold standard for clinical assessment of liver fat. However, it is subjective, has a poor reproducibility and significant sampling error and suffers from the risks associated with an invasive procedure.

Ultrasound, while widely available, only has a sensitivity of approximately 70% and this is further reduced in obese patients. It is additionally hampered by factors such as fibrosis, edema and extrahepatic adipose tissue.

CT is also a widely available imaging modality that evaluates hepatic steatosis indirectly based on hepatic x-ray attenuation. Several factors other than fat (e.g. iron, copper, glycogen, fibrosis, edema, ingestion of drugs such as amiodarone and gold) affect CT attenuation values, resulting in unavoidable errors in fat quantification and low sensitivity for mild-moderate steatosis. Moreover, as CT relies on ionizing radiation, it is not suitable for use in children and for longitudinal monitoring of adults with liver fat. MR has also been used for quantitative assessment of fat and is widely accepted as superior compared to ultrasound and CT.

HepaFat-AI is used in conjunction with a specific MRI data acquisition protocol that is a variation of the Dixon Method. Many MR fat quantification methods such as Magnetic Resonance Spectroscopy (MRS) and other Dixon based imaging methods report the fat:water proton ratio. This ratio is not directly comparable to a fat measurement from a liver biopsy specimen as it neither represents the percentage of parenchyma involved with fat nor the volume fraction of fat in the biopsy. Resonance Health has addressed this issue with HepaFat-AI.


Potential Clinical Applications

Gastroenterology and Hepatology Applications

Surgical Applications

Applications by Primary Care Physicians

How does HepaFat-AI work?

HepaFat AI is based on the Dixon principle of measuring the in-phase and out-of-phase MRI signal in liver tissue. Measurements are achieved through the acquisition of a series of spoiled gradient recalled echo images at three specific echo times.

The process from data upload to report generation takes under 60 seconds for completion. HepaFat AI has been extensively validated to ensure the results are accurate and reproducible.

HepaFat-AI via Resonance Health

HepaFat-AI is a stand-alone software application that automatically analyses MRI (DICOM) images that are uploaded via the Resonance Health web portal. These images, acquirable on most makes and models of 1.5 and 3 Tesla MRI machines, are obtained through a unique and standardised scanning sequence to ensure results are accurate, reliable, and reproducible over time and between hospitals and the various makes and models of MRI scanners.

The entire HepaFat-AI analysis process is completely automated.  The HepaFat-AI technology will initially inspect all images to ensure that they have been acquired with the correct protocol.  If errors are detected in the method of data acquisition, HepaFat-AI will not analyse the data but will indicate the corrective steps that need to be taken to acquire compliant images. Uploading and analysis of data typically takes about 30 seconds.

Due to the quality-controlled, standardised nature of the test, a patient may have HepaFat-AI analysis at any HepaFat-verified MRI centre, anywhere in the world.

HepaFat-AI Resources

Click the link(s) below to access our resources. Need something else? Please contact our team at [email protected] for further information.

Select A Login Portal