Multiparametric Abdominal Ultrasound as a Source of Imaging Biomarkers in Predictive and Personalized Medicine: A Narrative Review

Multiparametric ultrasound (MPUS) has emerged as a clinically relevant platform for the non-invasive generation of quantitative imaging biomarkers, transforming conventional abdominal sonography from a morphological screening tool into a functional, data-rich instrument aligned with predictive and personalized medicine. This review aims to synthesize current evidence on MPUS-derived imaging biomarkers across the principal abdominal organ systems and to examine their integration with radiomics and artificial intelligence. A narrative structured review was conducted through searches of PubMed/MEDLINE, Scopus, and Web of Science (January 2007–May 2026), combining terms related to multiparametric ultrasound, elastography, contrast-enhanced ultrasound (CEUS), quantitative ultrasound, radiomics, and imaging biomarkers. Articles were selected purposively on the basis of relevance, methodological quality, and recency; the review does not follow PRISMA methodology and does not include a formal quantitative synthesis. MPUS integrates B-mode imaging, Doppler, elastography, CEUS, and quantitative acoustic parameters (attenuation coefficient, sound speed, and viscosity) within a single examination. In hepatology, multiparametric protocols enable non-invasive staging of steatosis, fibrosis, and steatohepatitis in metabolic dysfunction-associated steatotic liver disease, portal hypertension assessment, and focal lesion characterization. Emerging applications span pancreatic, renal, vascular, and digestive pathology. The integration of radiomics and artificial intelligence amplifies biomarker potential, enabling molecular subtype prediction, treatment-response assessment, and prognostic stratification. Multiparametric abdominal ultrasound holds genuine promise as a central, non-invasive platform in precision medicine. Standardization of protocols, multicenter validation of quantitative thresholds, and integration with clinical, biochemical, and molecular data remain the principal challenges limiting broader clinical translation.