Current Ultrasound-Related Strategies for Assessing Liver Fibrosis in Chronic Liver Disease
Liver fibrosis is a significant consequence of various chronic liver diseases (CLDs), and its accurate assessment is crucial for effective patient management. Historically, liver biopsy has been the gold standard for evaluating fibrosis progression. However, its invasive nature, potential for sampling errors, and variability in interpretation have driven the need for non-invasive alternatives. Conventional ultrasound (US) provides some diagnostic information but is subjective and lacks sensitivity in the early stages of fibrosis. In recent years, ultrasound-based shear wave elastography (SWE) techniques have emerged as powerful tools for non-invasive liver fibrosis assessment. These include transient elastography (TE), point shear wave elastography (pSWE), and two-dimensional shear wave elastography (2D-SWE). Additionally, advancements in computer-aided quantitative techniques, particularly deep learning, have further enhanced the analysis of radiological images.
Shear Wave Elastography Techniques
Shear wave elastography techniques measure liver stiffness by generating shear waves within the liver tissue. The velocity of these waves is directly proportional to tissue stiffness, providing a quantitative measure of fibrosis severity. The most widely studied SWE technique is transient elastography (TE). In TE, a piston-mounted transducer induces a mechanical vibration on the skin, creating an elastic shear wave that propagates through the liver. Pulse-echo ultrasound acquisitions are then used to measure the shear wave velocity. TE has proven valuable for evaluating liver fibrosis stages in CLDs, but it has limitations. It does not provide real-time imaging, and the absence of gray-scale images can lead to errors in selecting the appropriate region of interest (ROI).
Point shear wave elastography (pSWE) is another SWE technique that can be implemented using conventional ultrasound devices with modified probes. In pSWE, shear waves are generated in a small ROI (approximately 1 cm³) of the liver by applying a high-frequency acoustic radiation force impulse (ARFI) pulse. The displacement of liver tissue caused by the shear waves is monitored using B-mode imaging. The clinical application of ARFI with virtual touch tissue quantification (VTQ) technology has been validated through comparisons with TE. Both techniques exhibit high predictive accuracy for detecting cirrhosis, with areas under the receiver operating characteristic curve exceeding 90%. ARFI offers several advantages over TE, including the ability to perform the procedure on a standard ultrasound machine, greater applicability in cases with complications like ascites, and the ability to examine multiple regions managed by the operator.
Elastography point quantification (ElastPQ) is another pSWE method that uses ARFI technology. Integrated into conventional ultrasound machines, ElastPQ allows for easy liver stiffness measurements after selecting the ROI on abdominal ultrasound. In both TE and ARFI methods, a single shear wave is temporarily emitted at a single frequency. However, in two-dimensional shear wave elastography (2D-SWE), the transducer emits a plurality of pulse wave beams at increasing depths with a wide frequency band ranging from 60 to 600 Hz. This technique superimposes a real-time color mapping of elasticity encoded pixel-by-pixel over the standard B-mode image, enabling quantitative imaging of tissue elasticity by placing an ROI within the color mapping. 2D-SWE has demonstrated good to excellent accuracy for diagnosing advanced fibrosis and cirrhosis, particularly in patients with viral hepatitis. It is also effective in evaluating portal hypertension, a critical clinical consequence of CLDs.
Comparison of Shear Wave Elastography Techniques
The European Federation of Societies for Ultrasound in Medicine and Biology guidelines recommend TE, ARFI, and 2D-SWE as first-line tools for assessing liver fibrosis severity in patients with chronic hepatitis C. These techniques are particularly effective in ruling out cirrhosis. For patients with chronic hepatitis B, pSWE using VTQ and 2D-SWE are useful for identifying cirrhosis. A comparison of these techniques reveals their unique advantages and limitations. TE, for instance, lacks gray-scale ultrasound guidance, while ARFI and 2D-SWE provide real-time imaging and can be guided by conventional gray-scale images. 2D-SWE, in particular, generates a real-time color mapping of tissue elasticity, offering a more comprehensive assessment.
Computer-Aided Quantitative Techniques
In recent years, computer-aided quantitative techniques have gained traction in supporting human decision-making in radiological image analysis. Traditional machine learning methods involve various feature extraction techniques, but their completeness is often compromised by image distortion. Deep learning, a subset of machine learning, has emerged as a powerful alternative. It can directly process and automatically learn mid-level and high-level abstract features from raw medical image data. Convolutional neural networks (CNNs), a type of deep learning architecture, consist of multiple modules, each typically including a convolutional layer, a pooling layer, and a fully connected layer. Deep learning methods applied during 2D-SWE analysis have shown excellent performance in fibrosis staging.
However, deep learning models require large, labeled training datasets to achieve optimal performance, which poses a challenge in current ultrasound image analysis. Transfer learning has been proposed as a solution to this issue. Transfer learning involves transferring knowledge from other domains to the medical ultrasound domain, thereby improving performance. Studies have demonstrated that liver fibrosis can be staged using transfer learning models based on the combination of grayscale and 2D-SWE images with excellent accuracy. The potential of transfer learning in assessing liver fibrosis is significant, but further exploration in large-scale studies is needed.
Conclusion
Shear wave elastography has revolutionized the non-invasive assessment of liver fibrosis. Techniques such as TE, pSWE, and 2D-SWE have proven effective in evaluating fibrosis severity, with 2D-SWE offering real-time imaging and comprehensive elasticity mapping. The integration of computer-aided quantitative techniques, particularly deep learning, has further enhanced the analysis of ultrasound images. As artificial intelligence systems continue to evolve, computer-aided quantitative techniques will remain a focal point of future research in liver fibrosis assessment.
doi.org/10.1097/CM9.0000000000001136
Was this helpful?
0 / 0