Quantitative Ultrasound(DeepUSFF) vs MRI-PDFF for Liver Fat Assessment in MASLD

Официальное название

Evaluation of a Quantitative Ultrasound Model(DeepUSFF) for Liver Fat Quantification in Patients With Metabolic Dysfunction-Associated Steatotic Liver Disease: A Multicenter Prospective Study Using MRI-PDFF as the Reference Standard

Обзор

This multicenter prospective study aims to evaluate the correlation between quantitative ultrasound fat fraction (USFF) and MRI-PDFF (Proton Density Fat Fraction) for liver fat quantification in patients with metabolic dysfunction-associated steatotic liver disease (MASLD). The study will compare the diagnostic accuracy of quantitative ultrasound imaging against MRI-PDFF as the reference standard.

Подробное описание

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a common liver disease requiring accurate assessment for treatment planning and monitoring. While liver biopsy remains the gold standard, it is invasive with potential complications. MRI-PDFF has emerged as an accurate non-invasive method, but it is expensive and has limited accessibility. Quantitative ultrasound techniques using RF data have been developed to provide objective liver fat assessment.

Objective: To prospectively evaluate the correlation between quantitative ultrasound-derived fat fraction (DeepUSFF) and MRI-PDFF in patients with suspected MASLD across different ethnicities and varying degrees of hepatic steatosis.

Methods: This prospective multicenter study will recruit 62 patients (31 from each participating center) suspected of having MASLD. All participants will undergo both quantitative ultrasound examination and non-contrast liver MRI within one week. The primary endpoint is the correlation coefficient between ultrasound fat fraction and MRI-PDFF. Secondary endpoints include diagnostic accuracy metrics and inter-observer reproducibility.

Вмешательства

• Устройство Quantitative ultrasound (DeepUSFF)
  \*\*Novel Quantitative Ultrasound Technology Assessment\*\* This study evaluates Samsung Medison's proprietary DeepUSFF (Deep Learning-based Ultrasound Fat Fraction) technology, a next-generation quantitative ultrasound method for liver fat assessment that differs from conventional ultrasound approaches in several key aspects: advanced RF data analysis, proprietary technology, standardized protocol, direct MRI-PDFF correlation, specific MSALD poopulation and multicenter design.

Первичные конечные точки

• Correlation between Ultrasound Fat Fraction and MRI-PDFF [Срок оценки: At enrollment (single time point assessment)]
• Correlation between Ultrasound Fat Fraction and MRI-PDFF [Срок оценки: At enrollment (single time point assessment)]

Критерии участия

Критерии включения

• Patients with clinically suspected MASLD based on abnormal ultrasound or liver function tests requiring liver ultrasound or MRI examination
• BMI ≥25 kg/m² or waist circumference >90 cm (male) or >80 cm (female), suggesting high likelihood of fatty liver disease
• Living liver transplant donors requiring preoperative liver ultrasound or MRI examination
• Age ≥18 years
• Understanding and signing informed consent

Критерии исключения

• Significant alcohol consumption in the past 2 years:

Male: ≥30-60g/day average alcohol intake Female: ≥20-50g/day average alcohol intake

-Chronic liver disease:

Histological diagnosis of chronic liver disease HBsAg positive Anti-HCV positive Other suspected chronic liver diseases

-Liver failure:

Serum albumin <3.2 g/dL INR >1.3 Direct bilirubin >1.3 mg/dL

• History of esophageal varices, ascites, hepatic encephalopathy, or acute biliary obstruction
• History of liver cancer diagnosis or treatment
• History of liver surgery
• Pregnancy
• Inability to obtain adequate liver ultrasound imaging:

Patient cooperation impossible Inadequate image acquisition as determined by investigator

-Inability to obtain adequate liver MRI imaging: Patient cooperation impossible Severe obesity preventing MRI examination MRI contraindications (cardiac pacemaker, etc.) Other factors preventing adequate imaging as determined by investigator

Критерии приведены из реестра в оригинале (на английском). Окончательную оценку соответствия проводит исследовательский центр.

Здоровые добровольцы: Нет

Дизайн исследования

Распределение

Не применимо

Модель

Одна группа

Маскирование

Открытое

Основная цель

Диагностика

Центры проведения

США · 1 центр

• southwoods imaging (Northeastern Ohio Radiology Research and Education Fund ) — Boardman

South Korea · 1 центр

• Seoul National University Hospital — Seoul

Публикации

• Xie WJ, Zhang B. Learning the Formation Mechanism of Domain-Level Chromatin States with Epigenomics Data. Biophys J. 2019 May 21;116(10):2047-2056. doi: 10.1016/j.bpj.2019.04.006. Epub 2019 Apr 11. PMID 31053260
• Fan C, Shen M, Nussinov Z, Liu Z, Sun Y, Liu YY. Reply to: Deep reinforced learning heuristic tested on spin-glass ground states: The larger picture. Nat Commun. 2023 Sep 14;14(1):5659. doi: 10.1038/s41467-023-41108-w. No abstract available. PMID 37709759
• Zhou G, Qin Y, Petticord D, Qiao X, Jiang M. Plant-ant interactions mediate herbivore-induced conspecific negative density dependence in a subtropical forest. Sci Total Environ. 2024 Jun 1;927:172163. doi: 10.1016/j.scitotenv.2024.172163. Epub 2024 Apr 1. PMID 38569958

Идентификаторы

NCT: NCT07192159 · 2503-162-1625

Первоисточники (государственные реестры)

• ClinicalTrials.gov (NIH, США)

Открыть это исследование на ClinicalTrials.gov ↗