Compatibility and Bias Between Ultrasound-guided Attenuation Parameter and Controlled Attenuation Parameter for Ultrasound-based Hepatic Steatosis Assessment
Year: 2026
Type: article
Study design: Observational Study
Abstract: This retrospective observational study aimed to develop a regression model to estimate the controlled attenuation parameter (CAP) from the ultrasound-guided attenuation parameter (UGAP) attenuation coefficient (AC), and to evaluate the agreement, biases, and potential interchangeability between the 2 measurements in a routine clinical setting. A total of 85 patients who underwent both UGAP and CAP measurements on the same day were included. UGAP-AC values were converted from dB/cm/MHz to dB/m to match CAP units. Pearson's correlation and linear regression analyses were performed to derive a predictive equation for CAP based on UGAP-AC. Agreement and biases between the 2 attenuation values were assessed using Bland-Altman analysis. In addition, repeated-measures analysis of variance was conducted to assess the influence of body mass index (BMI) on measurement differences. UGAP-AC and CAP showed a strong positive correlation (r=0.947, P<0.0001). The regression equation was CAP (dB/m)=371.5× UGAP-AC + 12.3 (R²=0.8946). Bland-Altman analysis revealed a fixed bias of 28.6 dB/m and a BMI-dependent proportional bias. However, this proportional bias disappeared observed when analysis was limited to the clinically relevant CAP range (228 to 300 dB/m), suggesting good comparability between UGAP and CAP within the diagnostic range. UGAP-AC can reliably estimate CAP using a simple linear regression model. While some systematic biases exist, their clinical impact is limited. These findings support the potential interchangeability of UGAP and CAP for noninvasive hepatic steatosis assessment in high-throughput clinical practice.
Source: Ultrasound Quarterly
Authors: Daisuke Miura, Rino Hiwatashi, Tomoko Sakata
Institution: Fukuoka Tokushukai Hospital
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7126095258