Introduction
The prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) is growing worldwide with the increase in obesity. Among the many forms of MASLD, metabolic steatohepatitis (MASH) has attracted attention, as it can progress to liver cirrhosis and hepatocellular carcinoma due to hepatocyte apoptosis, inflammation and fibrosis. Traditionally, liver biopsy has been the gold standard for the diagnosis and assessment of hepatic steatosis. However, this method has some problems such as sampling error and inter-pathologist variability. In addition, the invasive nature of the procedure creates a risk of complications. More recently, MR proton density fat fraction (PDFF) has been accepted as a non-invasive reference standard, but limited access and high cost prevent widespread use, especially with respect to regular follow up exams.
The liver echogenicity on ultrasound B-Mode is widely used for the detection of hepatic steatosis. However, this technique does not enable a quantitative assessment since liver texture or brightness may vary depending on the imaging parameters used or the examiner's technique. Therefore, an objective ultrasound quantification method is desired for steatosis grading in current and potential MASH patients. Recently, a novel non-invasive tool that utilizes attenuation of the sound wave was developed. However, it may be susceptible to multi-reflection artifacts from subcutaneous tissues as well as disruptive structures such as vessels or diaphragm since the measurement area is not guided by imaging.
This paper describes Ultrasound-Guided Attenuation Parameter (UGAP), a real-time, image-guided method of measuring the attenuation of the sound wave. The principles of the method as well as the clinical evaluation results are presented.
Ultrasound Attenuation
When the ultrasound wave propagates in an organ, such as the liver, it is gradually weakened due to diffusion, scattering and absorption. Known as sound attenuation, this results in less signal returning to the ultrasound transducer, causing the image to get darker with depth. If an image of a healthy liver has uniform image brightness over depth, it is because of the time-gain compensation (TGC) capability of the ultrasound scanner, which applies a different gain for each depth. In the case of fatty liver, the presence of many lipid droplets in the hepatocytes becomes the dominant factor of the attenuation, sometimes causing insufficient echo signals in the deeper area.
The amplitude of ultrasound wave u propagating in the x direction is expressed u = u₀e⁻ᵃˣ, where u₀ is the amplitude at x = 0 and α is attenuation rate. Since the sound attenuation increases nearly proportional to the frequency (ranging between 1 MHz to 10 MHz), the attenuation rate can be approximated by α = α₀f, where α₀ is the attenuation coefficient (dB/cm/MHz) and f is frequency in [MHz]. Attenuation rate and attenuation coefficient are used primarily to evaluate ultrasound attenuation in human tissues.
Figure 1: Ultrasound attenuation compensation by TGC. Shows how Time-Gain Compensation (TGC) adjusts signal gain over depth for uniform image brightness in a normal liver, contrasted with a fatty liver where deeper areas may appear darker due to increased attenuation.
Ultrasound Attenuation Evaluation Method by UGAP
Principle of Measurement
Measuring the attenuation slope would provide insight into liver attenuation. However, the sound profile in the real signal is not so simple, since it is curved. This complexity is caused by a focused sound beam from transmission as well as reception conditions. To cancel out or compensate for this complexity, several methods have been reported. UGAP performs the compensation based on a Reference-Phantom Method (RPM). The profile of the echo amplitude for a tissue-mimicking phantom is measured in the depth direction and stored in the ultrasound system as a reference. In this case, a frequency of 3.5 MHz is used. This industry-standard phantom includes glass bead particles for attenuating materials and the attenuation coefficient is known. In the UGAP mode, the transmission and reception conditions are fixed to the same values as were used on the reference phantom, and the acquired echo profiles of the target (liver) are compensated by the reference data. As a result, the compensated sound profiles represent only decay caused by attenuation. If the compensated sound profile is flat, the attenuation is the same as the reference phantom.
Figure 2: Compensation of ultrasound signal to enable UGAP measurement. Illustrates the process of measuring echo amplitude profiles and compensating for system-dependent characteristics using a reference phantom method (RPM) to isolate true attenuation.
Measurement Algorithm
Although the system-dependent sound profile is compensated, there are still problems in performing a successful measurement. For example, structures such as large vessels and diaphragm may deform the slope profile. Or, multi-reverberation in the subcutaneous fat may generate artifacts into the liver parenchyma. Or, information needed to determine the slope may be diminished if the attenuation is very large. To avoid these problems, UGAP includes an automated measurement algorithm to find and then analyze the optimum measurement range. The start point of the range is determined by analyzing linearity and discontinuity of the echo profile close to the liver surface, thereby avoiding the multi-reverberation artifacts. In addition, the algorithm automatically detects and avoids depths where the signal-to-noise ratio (SNR) is insufficient. This enables the algorithm to employ the deepest, usable end point. The diaphragm is also automatically excluded. Finally, the angle of the slope is measured across this optimum range to provide a representative attenuation coefficient. Because this measurement takes place on the raw data of a frozen or recalled image, it is not dependent on gain or other post processing settings. The goal of these automations is to make the UGAP measurement less dependent on the ROI position and more robust across various liver sizes and conditions.
Color Mapping for Measurement Guidance
To find the right scan-plane and ROI for measurement, B-Mode and two types of new color-mapped images are available in UGAP mode.
(a) Attenuation map: Displays color-coded local attenuation values for each pixel. When the measurement area (denoted by a trapezoid with a center line) has a uniform color, it is suitable to measure. The color will become inhomogeneous if the area includes a disruptive structure such as a large vessel.
(b) Quality map: Displays a color at pixels where signal quality is sufficiently high to perform a measurement. Even though the B-Mode texture may look homogeneous, a lack of color could be the result of unseen artifacts.
To aid in the acquisition and measurement of UGAP, various display formats are selectable: B-Mode only, B-Mode with color map overlay, and a dual display that shows both images side-by-side. Examples of the color-coded attenuation for different degrees of steatosis are shown in Figure 4.
Figure 3: Example of B-Mode/color dual display format for UGAP. Shows a B-Mode ultrasound image alongside a color-coded attenuation map, aiding in selecting a measurement region with uniform color for accurate assessment.
Figure 4: Examples of color-mapped attenuation images showing different degrees of liver fat. These images, courtesy of Prof. Sporea, Victor Babes University of Medicine and Pharmacy of Timisoara, demonstrate how color intensity correlates with steatosis severity (Normal, Mild, Moderate, Severe fatty liver).
Evaluation of Steatosis Grade in Chronic Liver Disease by UGAP
This section summarizes the results of the multicenter prospective study reported by Nayoro City General Hospital, Ogaki Municipal Hospital, Yokohama City University Hospital, Tokyo Medical University Hospital, Iwate Medical University, Musashino Red Cross Hospital and Gifu Kyoritsu University Hospital.
Materials and Methods
A total of 1,010 patients with chronic liver disease who underwent MRI-PDFF (proton density fat fraction) and UGAP were enrolled. PDFF was measured using a multi-echo Dixon method (IDEAL-IQ sequence) with a single region of interest (ROI) (20 × 20 × 20 mm) placed in liver segment VII or VIII. Patients' sex, age, etiology, body mass index (BMI), skin capsule distance and steatosis grade are shown in Table 1.
| Variable | Value |
|---|---|
| Subject (n) | 1,010 |
| Sex (male/female) | 544/468 |
| Age (y) | 61.7 (52.0 – 72.0) |
| Etiology (NAFLD/ALD/HBV/HCV/AIH/others) | 515/90/133/124/38/110 |
| BMI (kg/m²) | 26.2 (23.1 – 28.9) |
| Skin capsule distance (mm) | 19.0 (15.5 – 22.0) |
| Steatosis grade (%) | |
| S0 (< 5.2%) | 356 (35.2%) |
| S1 (5.3% < PDFF < 11.3%) | 281 (27.8%) |
| S2 (11.4% < PDFF < 17.1%) | 168 (16.6%) |
| S3 (PDFF > 17.2%) | 205 (20.3%) |
UGAP measurements were performed using software equivalent to LOGIQ E10, LOGIQ E10s and LOGIQ Fortis. All measurements were performed within three months before or after MRI-PDFF, in fasting conditions for more than four hours, on patients in a supine position, with the right arm in maximum abduction, by intercostal approach, in the right liver lobe. A colored-coded attenuation map was used during measurement to confirm a homogenous area of the liver, free of large vessels. At least six measurements were performed. Reliable UGAP measurements were defined as the median value of six measurements performed in a homogeneous area of liver parenchyma, with an IQR/M < 0.30. UGAP values are expressed in dB/cm/MHz.
Figure 5: Description of (A) MRI-based proton density fat fraction (PDFF) and (B) Ultrasound-Guided Attenuation Parameter (UGAP) measurements. Shows the placement of a region of interest (ROI) and the resulting attenuation map for UGAP.
Results
All UGAP measurements were successful for 1010 cases. The mean MRI-PDFF was 10.8 ± 8.05% (range: 0.3 – 44.81%), and the mean UGAP value was 0.69 ± 0.12 dB/cm/MHz (range: 0.32 – 0.99 dB/cm/MHz). The relationship between UGAP and PDFF is shown in Figure 6. Significantly strong correlation was confirmed between UGAP and PDFF, Spearman's rank correlation coefficient was 0.785 (p < 0.001).
UGAP values in each steatosis grade are shown in Figure 7. The average values of UGAP in S0, S1, S2 and S3 steatosis were 0.59, 0.69, 0.77, and 0.83 dB/cm/MHz, respectively, as UGAP increased with increasing steatosis grade.
Receiver operating characteristics (ROC) curve of UGAP for diagnosis of S1 or higher, S2 or higher and S3 steatosis are shown in Figure 8: (A): ≥ S1, (B): ≥ S2, (C): S3. Area under the ROC curve (AUROC), 95% confidence interval (95% CI), and cutoff value of UGAP are shown in Table 2. The AUROCs of UGAP for the prediction of S1 or higher, S2 or higher and S3 steatosis were 0.901 (95% CI: 0.891 – 0.928), 0.912 (95% CI: 0.894 – 0.929) and 0.894 (95% CI: 0.873 – 0.916), respectively.
Figure 6: Relationship between UGAP and PDFF. A scatter plot illustrating the strong positive correlation (r=0.785) between Ultrasound-Guided Attenuation Parameter values and MRI-PDFF measurements.
Figure 7: Measured UGAP values in each steatosis grade. Box plots showing the distribution of UGAP values for different grades of steatosis (S0, S1, S2, S3), indicating an increase in UGAP with higher steatosis levels.
Figure 8: ROC of UGAP for diagnosis of steatosis grade. Receiver Operating Characteristic curves demonstrating the diagnostic performance of UGAP for predicting (A) ≥S1, (B) ≥S2, and (C) S3 steatosis, with reported Area Under the ROC Curve (AUROC) values.
| UGAP | ≥ S1 | ≥ S2 | S3 |
|---|---|---|---|
| AUROC (95% CI) | 0.901 (0.891 – 0.928) | 0.912 (0.894 – 0.929) | 0.894 (0.873 -0.916) |
| Attenuation coefficient cutoff value (dB/cm/MHz) | 0.65 | 0.71 | 0.77 |
| Attenuation rate cutoff value (dB/m) | 228 | 249 | 270 |
Discussion
In this prospective study based on more than 1,000 examinations – the largest to date – the results showed that UGAP has high accuracy at diagnosing and grading steatosis with MRI-PDFF as reference standard. Liver biopsy is still considered the gold standard for the evaluation of hepatic steatosis, but it is painful and costly. In addition, MRI-PDFF has the advantage of accounting for spatial variability of hepatic steatosis compared to biopsy which has potential for sampling error and may result in grading inaccuracies. Recently, there have been several reports showing the utility of MRI-based PDFF for diagnosing hepatic steatosis, including reports showing the superiority of MRI-PDFF to VCTE-based CAP for diagnosing hepatic steatosis grade in NAFLD patients who underwent liver biopsy. Therefore, MRI-PDFF is a precise and accurate non-invasive imaging biomarker for the diagnosis of hepatic steatosis.
The result of this study indicated that UGAP has significant linearity and negligible bias with respect to the reference standard of MRI-PDFF measurements over the entire range of observed steatosis severity. Therefore, it can be concluded that UGAP has excellent technical performance characteristics as a quantitative method of steatosis for widespread use in clinical trials and patient care.
Evaluation of the LOGIQ Totus Comparing to LOGIQ Fortis
Alina Popescu, Felix Bende, Ioan Sporea et. al. Department of Gastroenterology and Hepatology, "Victor Babeș” University of Medicine and Pharmacy Timișoara, Romania.
Background
The LOGIQ Totus is equipped with the same UGAP technology as the LOGIQ E10, LOGIQ E10s and LOGIQ Fortis. In addition, the same C1-6-D and C1-6VN-D probes are optimized for UGAP. UGAP workflow features, such as measurement tools, worksheet pages, automatic statistics on the summary pages, and the color mapping for measurement guidance are identical on the LOGIQ E10, LOGIQ E10s, LOGIQ Fortis and LOGIQ Totus.
A clinical study was performed to compare liver measurement values on the LOGIQ Fortis versus LOGIQ Totus and determine whether the LOGIQ E10, LOGIQ E10s, LOGIQ Fortis cutoff chart can be used for LOGIQ Totus.
Materials and Methods
Subjects
141 patients were recruited. The average age and BMI of the subjects were 56 years (age range 42–68 years) and 28.7 (BMI range 24.1–33.3), respectively. The subject demographics and disease etiology are summarized in Table 3.
| Variable | Value |
|---|---|
| Subject (n) | 141 |
| Sex (male/female/unknown) | 73/63/5 |
| Age (y) | 55.8 (42.7 – 68.9) |
| Etiology (NAFLD/ALD/HBV/HCV/ASH/others) | 50/8/27/15/7/17 |
| BMI (kg/m²) | 28.7 (24.1 – 33.3) |
| Steatosis grade by LOGIQ Fortis cut-off chart (%) | |
| S0 | 61 (43.3%) |
| S1 | 27 (19.1%) |
| S2 | 19 (13.5%) |
| S3 | 34 (24.1%) |
Data Collection Methods
UGAP data in the liver were acquired using GE HealthCare LOGIQ Totus and LOGIQ Fortis ultrasound systems. On the LOGIQ Totus, the R4.5.4 software version and C1-6-D probe were used. The LOGIQ Fortis used the R4.0.1 software version and C1-6-D probe. Ten measurements were performed, and a median value was calculated as a final UGAP result for each subject. These measurements were performed consistently across both LOGIQ Totus and LOGIQ Fortis for the same subjects. Although UAGP results were available for both attenuation coefficient in dB/cm/MHz.
Data Analysis Methods
UGAP measurements were performed by placing at least ten measurement regions over the stored images. The measurement regions were chosen by the operators to exclude obvious artifacts in the measurement guidance image. The average attenuation coefficient within each measurement region was automatically recorded by the system in a worksheet. These measurement regions were typically placed on different UGAP image frames so that independent measurements of attenuation coefficient were obtained for each subject. MedCalc® Statistical Software version 22.016 (MedCalc Software Ltd, Ostend, Belgium; https://www.medcalc.org; 2024) was used to analyze data. The correlation coefficient between LOGIQ Totus and LOGIQ Fortis, as well as the data distribution for each system, was examined. Subsequently, normality tests were conducted for both datasets using the Shapiro-Wilk and Kolmogorov-Smirnov tests. This was followed by the application for the Mann-Whitney U test and independent samples t-test.
Results
The attenuation coefficient for all patients measured by LOGIQ Fortis and LOGIQ Totus are shown in Figure 9, grouped by steatosis grade as determined by cutoff values in Table 2.
Figure 9: Median of UGAP measurements for 141 subjects, grouped by steatosis stage for GE HealthCare LOGIQ Fortis and GE HealthCare LOGIQ Totus. Box plots illustrate the UGAP values across different steatosis grades for both ultrasound systems.
Discussion
As expected, attenuation coefficient measured by LOGIQ Totus UGAP increased with the level of steatosis. A strong correlation was observed between LOGIQ Fortis and LOGIQ Totus (R² = 0.90, p < 0.001). The two datasets were found to not followed a normal distribution according to the Shapiro-Wilk (LOGIQ Fortis; W=0.9796, P=0.0330, reject normality, LOGIQ Totus; W=0.9738, P=0.0083, reject normality), and based on the Kolmogorov-Smirnov Test (LOGIQ Fortis; D=0.0557, P>0.10, accept normality, LOGIQ Totus; D=0.0800, P=0274, reject normality) only LOGIQ Fortis conforms to a normal distribution, while LOGIQ Totus does not. Subsequently, the Mann-Whitney U test was employed to compare the LOGIQ Fortis and LOGIQ Totus, yielding a result of P=0.2557. Furthermore, an independent sample t-test was also conducted, showing the difference between LOGIQ Fortis and LOGIQ Totus measurement values was not significant (P=0.2796).
Ultrasound-Guided Attenuation Parameter Conclusions
UGAP is a non-invasive method for measuring a patient-specific, quantitative attenuation parameter that is well correlated to liver biopsy for discriminating hepatic steatosis among patients with CLD. The B-Mode image provides an anatomical guide while the attenuation and quality maps provide an attenuation quality guide. This combination provides extensive user assistance for proper placement of the UGAP measurement ROI. Lastly, automated algorithms optimize the results within the specified measurement ROI. As such, UGAP is an easy and fast tool that, in combination with 2D shear wave elastography, has the potential to aid in the initial diagnosis and follow-up care of CLD patients.
References
- Loomba R, Sanyal AJ. The global NAFLD epidemic. Nat Rev Gastroenterol Hepatol, 10, 686-690, 2013.
- Angulo P. Nonalcoholic fatty liver disease. N Engl J Med, 346, 1221-1231, 2002.
- Ratziu V, Charlotte F, Heurtier A, Gombert S, Giral P, Bruckert E, Grimaldi A, Capron F, Poynard T, Group LS. Sampling variability of liver biopsy in nonalcoholic fatty liver disease. Gastroenterology, 128, 1898-1906. 2005.
- Karlas T, Petroff D, Sasso M, Fan JG, Mi YQ, de Ledinghen V, Kumar M, Lupsor-Platon M, Han KH, Cardoso AC, Ferraioli G, Chan Wk, Wong VW, Myers RP, Chayama K, Friedrich-Rust M, Beaugrand M, Shen F, Hiriart JB, Sarin SK, Badea R, Jung KS, Marcelin P, Filice C, Mahadeva S, Wong GL, Crotty P, Masaki K, Bojunga J, Bedossa P, Keim V, Wiegand J. Individual patient data meta-analysis of controlled attenuation parameter (CAP) technology for assessing steatosis. J Hepatol, 66, 1022-1030, 2017.
- Ultrasound Handbook, Maruze-Yushodo Co. (in Japanese) 1999.
- Fujiwara Y, Kuroda H, Abe T, Ishida K, Oguri T, Noguchi S, Sugai T, Kamiyama N, Takikawa Y. The B-Mode image-guided ultrasound attenuation parameter accurately detects hepatic steatosis in chronic liver disease. Ultrasound in Med Biol, 44(11), 2223-2232, 2018.
- Mamou J, Oelze ML. Quantitative Ultrasound in Soft Tissue. Springer, 2013.
- Yao LX, Zagzebski JA, Madsen EL. Backscatter coefficient measurements using a reference phantom to extract depth dependent instrumentation factors. Ultrason Imaging, 12, 58–70, 1990.
- Suzuki Y, Yasuda S, Toyota H, Imajo K, Nakajima A, Sugimoto K, Kuroda H, Yasui Y, Kurosaki M, Kumada T. Evaluation of Hepatic Steatosis in Chronic Liver Disease by UGAP: Comparison With MRI PDFF. Jpn J Med Ultrasonics 48, S275 (Suppl.), 2021.
- Imajo K, Toyoda H, Yasuda S et al. Utility of Ultrasound-Guided Attenuation Parameter for Grading Steatosis With Reference to MRI-PDFF in a Large Cohort. Clinical Gastroenterology and Hepatology, 20, 2533-2541, 2022.
- Wong RJ, Aguilar M, Cheung R, et al. Nonalcoholic steatohepatitis is the second leading etiology of liver disease among adults awaiting liver transplantation in the United States. Gastroenterology 2015; 148:547–555.
- Imajo K, Kessoku T, Honda Y, el. Magnetic Resonance Imaging More Accurately Classifies Steatosis and Fibrosis in Patients With Nonalcoholic Fatty Liver Disease Than Transient Elastography. Gastroenterology. 2016 Mar;150(3):626-637.e7.
MedCalc statistical software - free trial available
