BACKGROUND Hepatic steatosis is a major cause of chronic liver disease. Two-dimensional (2D) ultrasound is the most widely used non-invasive tool for screening and monitoring, but associated diagnoses are highly subjective. AIM To develop a scalable deep learning (DL) algorithm for quantitative scoring of liver steatosis from 2D ultrasound images. METHODS Using multi-view ultrasound data from 3310 patients, 19513 studies, and 228075 images from a retrospective cohort of patients received elastography, we trained a DL algorithm to diagnose steatosis stages (healthy, mild, moderate, or severe) from clinical ultrasound diagnoses. Performance was validated on two multi-scanner unblinded and blinded (initially to DL developer) histology-proven cohorts (147 and 112 patients) with histopathology fatty cell percentage diagnoses and a subset with FibroScan diagnoses. We also quantified reliability across scanners and viewpoints. Results were evaluated using Bland-Altman and receiver operating characteristic (ROC) analysis. RESULTS The DL algorithm demonstrated repeatable measurements with a moderate number of images (three for each viewpoint) and high agreement across three premium ultrasound scanners. High diagnostic performance was observed across all viewpoints: Areas under the curve of the ROC to classify mild, moderate, and severe steatosis grades were 0.85, 0.91, and 0.93, respectively. The DL algorithm outperformed or performed at least comparably to FibroScan control attenuation parameter (CAP) with statistically significant improvements for all levels on the unblinded histology-proven cohort and for “= severe” steatosis on the blinded histology-proven cohort. CONCLUSION The DL algorithm provides a reliable quantitative steatosis assessment across view and scanners on two multi-scanner cohorts. Diagnostic performance was high with comparable or better performance than the CAP.
Inflammation has significant impacts on liver fibrosis measurement by ultrasound elastography. The interpretation requires further optimization in patients with or without anti-viral therapy. We prospectively enrolled a consecutive series of patients with chronic hepatitis B who received liver histology analysis and acoustic radiation force impulse (ARFI). 146 patients who underwent liver biopsy (50.9%) or tumor resection (49.1%) were enrolled. 34 patients (23.3%) had been receiving anti-hepatitis B therapy of various duration. The areas under the receiver-operating characteristic (AUROC) for the diagnosis of Metavir F4 by mean ARFI was 0.820 in the non-treatment group and 0.796 in the treatment group. The ARFI tended to be not lower (100%) than the corresponding Metavir grading in patients with treatment within 12 months, equal (75%) from 13 to 31 months, and lower (71.4%) after 32 months. We conclude that ARFI is a reliable tool for measurement of liver fibrosis in chronic hepatitis B patients with ALT (alanine aminotransferase) <5x the upper limit of normal. For those patients under anti-HBV therapy, the optimal timing for ARFI analysis will be over 1-2.5 years of nucleos(t)ide analogue therapy. The ARFI measurement after 2.5 years tends to be lower than the corresponding histology grading.
Fibrosis-4 (FIB4), transient elastography (TE), and acoustic radiation force impulse (ARFI) are popular modalities to assess liver fibrosis. Their cutoff values for degrees of fibrosis vary between studies. The influence of hepatic steatosis on fibrosis measurements for different etiologies was evaluated. Data from a consecutive series of patients who received fibrosis measurement were included for the training group. An additional series with histology served as the validation group. A standardized protocol was performed for both TE and ARFI, mostly by a single technician. Patients with alcoholism, autoimmune disease, active inflammation, or who were receiving therapy were excluded. The training group included 215 patients and the validation group included 221. The correlation of liver stiffness between TE and ARFI was good (R2 linear = 0.798; p < 0.001). Different correlations between ARFI and TE were noted between high and low control attenuation parameter (CAP) values (cutoff: 290 dB/m), especially in the non-hepatitis B subgroups. Relatively lower FIB4 and TE values were seen in the high CAP versus low CAP in patients with histology-proven non-alcoholic fatty liver disease and chronic hepatitis C. FIB4 cutoff values were >25% lower among F2-F4 stages and the TE cutoff value for F4 was 8.5% lower in the high versus low CAP group. Such findings were not observed in chronic hepatitis B. Different fibrogenesis mechanisms between hepatitis B and non-B are discussed. We conclude that hepatic steatosis significantly impacts FIB4 and TE fibrosis measurements in non-hepatitis B-related liver diseases. Fibrosis grade should be interpreted with caution in severe steatosis.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.