WED-212

Circulating extracellular vesicles and ferritin enable machine learning based detection of hepatic steatosis in metabolic dysfunction-associated steatotic liver disease

Eleni-Myrto Trifylli, et al.  Greece

This study used iLivTouch FT100 as the transient elastography platform to assess hepatic steatosis in MASLD patients, providing the non-invasive imaging backbone for the biomarker analysis. iLivTouch served as the key phenotyping tool that enabled the machine-learning model to be trained and externally validated against steatosis status. The model combining circulating extracellular vesicles and ferritin achieved strong external performance, supporting the value of iLivTouch-based elastography plus blood biomarkers for MASLD risk stratification.

Key message

·        iLivTouch provided the non-invasive steatosis measurement platform for the study.

·        It anchored the imaging phenotype used for model training and external validation.

·        The study highlights the value of combining iLivTouch elastography with circulating biomarkers for MASLD risk stratification.