Recently, an original clinical research study supported by Eieling Technology was officially published in Liver International (2024 IF: 5.2, JCR Q1, Ranked 24/148 in Gastroenterology & Hepatology), a leading SCI-indexed journal in hepatology. The prospective study, titled “Identification of Hepatic Fibrosis and Steatosis via A Point-of-Care Transient Elastography System With Integrated AI,” comprehensively validated the outstanding performance of Eieling Technology’s Liverscan® platform with unprecedented methodological rigor and an exceptionally large sample size. This marks the highest-level international academic recognition for China’s independently developed innovative products in the field of non-invasive liver disease diagnosis.
Prestigious International Endorsement – Validated by a World-Class Research Team
This study was jointly led by Prof. Ming-Hua Zheng (Director of Hepatology, The First Affiliated Hospital of Wenzhou Medical University), Prof. Wai-Kay Seto (Li Ka Shing Faculty of Medicine, The University of Hong Kong), and Prof. Yong-Ping Zheng (Core Inventor of Liverscan® technology, Department of Biomedical Engineering, The Hong Kong Polytechnic University). The research brought together top-tier hepatology and engineering experts from mainland China, Hong Kong, and Taiwan.
Rigorous Study Design – Highest Level of Clinical Evidence
The study design reflects the “gold standard” for medical device clinical validation:
- Prospective Study Design– All patients were enrolled prospectively, ensuring significantly higher reliability than retrospective analyses.
- Liver Biopsy as Gold Standard– 138 patients underwent both liver biopsy (the definitive diagnostic gold standard) and Liverscan® examination. Pathologists independently assessed biopsy slides while blinded to ultrasound results, ensuring objective comparison.
- Large-Scale Head-to-Head Comparison with Industry Benchmark– 1,455 patients received both Liverscan® and FibroScan® on the same day, completing the world’s largest head-to-head validation of a portable transient elastography device against FibroScan®.
Key Findings – Data-Driven Performance Matching the Gold Standard
| Finding | Performance |
| Near-Perfect Cirrhosis Detection | Liverscan® achieved an AUROC of 0.97 for cirrhosis (F4) detection, approaching a “perfect classifier” level for identifying advanced liver fibrosis. |
| Excellent Steatosis Screening | Liverscan®’s proprietary next-generation fat quantification technology (MAP) achieved an AUROC of 0.92 for steatosis (≥S1) screening, providing reliable technical support for community-based large-scale fatty liver early screening. |
| High Diagnostic Agreement | Liverscan® demonstrated strong correlation with FibroScan®: r = 0.86 for liver stiffness measurement (fibrosis) and r = 0.80 for fat quantification (MAP vs CAP), confirming highly consistent diagnostic results. |
Revolutionary Efficiency Breakthrough – Just 4 Measurements with B-Mode Guidance
One of the study’s most impactful findings: With Liverscan®’s real-time B-mode ultrasound guidance, only 4 valid measurements are required to achieve statistical non-inferiority to the industry-standard 10-measurement protocol. This more than doubles screening efficiency, significantly reducing examination time and increasing throughput in community screening settings.
The underlying reason: B-mode guidance ensures every measurement is precisely targeted on liver tissue, eliminating the need for multiple “blind” attempts to compensate for positioning errors.
MAP – A Generational Leap in Fat Quantification Technology
| Technical Dimension | CAP (Conventional) | MAP (Liverscan®) |
| Sampling Volume | Single RF line (1D) | 32 RF lines with 2D ROI (32× larger sampling volume) |
| Measurement Depth | Fixed ROI | Adaptive ROI (auto-adjusts to patient anatomy) |
| Analysis Method | Frequency-domain only | Time-frequency multi-domain analysis |
| Liver Segmentation | None | AI-powered segmentation (BiSeNet V2) |
Traditional technology is like “looking at a room through a straw,” while MAP is like “opening a window” – offering a wider, clearer view for more stable and accurate assessment of hepatic fat content.
Four AI Deep Learning Models – Defining the Next-Generation Product
The paper details how Liverscan® leverages four purpose-built AI deep learning models to achieve a technological leap:
| AI Model | Function |
| YOLOv8 | Intelligent mode switching – automatically identifies obesity and switches detection modes without manual probe replacement. |
| ConvNeXt V2 | Elastogram quality control – automatically evaluates and rejects substandard measurements. |
| Transformer-Enhanced Hough Transform | Precise shear wave velocity measurement – accurately calculates liver stiffness, improving fibrosis diagnostic accuracy. |
| BiSeNet V2 | Liver segmentation for fat quantification – precisely segments liver parenchyma to ensure MAP measurements target only hepatic tissue, with 32× the sampling volume of conventional CAP for more stable, accurate results. |
These four models are integrated end-to-end, allowing AI to “supervise every step” – from image acquisition to final output – minimizing operator-induced variability.
What This Means in Practice
A nurse at a community health center, with minimal training, can operate Liverscan® to perform standardized liver disease screening – without relying on scarce ultrasound specialists. For China’s vast primary healthcare network, this represents a true “technological democratization.”
Broad Application Scenarios – Empowering Diverse Healthcare Settings
With its top-tier validated performance, portable wireless design, and highly efficient workflow, Liverscan® is well-suited for:
- Community & Primary Care Screening– Portable design + efficient protocol + ease of use = significantly reduced costs for large-scale population screening.
- Health Checkup Centers– Adds high-value fatty liver quantification as a premium screening item in routine health packages.
- Hepatology & Gastroenterology Clinics– With 4 measurements, auto-positioning, and intelligent QC, non-specialists can perform professional-grade liver assessments.
- Clinical Research & Drug Trials– Serves as a non-invasive endpoint for monitoring hepatic fat changes in GLP-1 agonist weight-loss therapies or other clinical trials.
Conclusion
Validated by prospective research published in Liver International: Liverscan® is the world’s first AI-integrated, B-mode-guided portable transient elastography system – proven equivalent to FibroScan® for both fibrosis and steatosis detection across nearly 1,500 patients, yet more portable (wireless handheld), more intelligent (4 AI models), and more efficient (4 measurements vs 10). It ushers in a new era of non-invasive liver assessment for community screening, health checkups, and clinical care.
Looking Ahead – From Academic Excellence to Global Accessibility
The publication in Liver International is a starting point, not an endpoint. It demonstrates that China-developed AI medical devices are fully capable of matching – and even surpassing – international gold standards.
From the laboratories of The Hong Kong Polytechnic University to the prestigious pages of Liver International; from community-wide screening programs in Hong Kong to the industrialization base in Nanjing Jiangning – Liverscan® is charting an innovative path “from research to clinic, from paper to product, from China to the world.”
Eieling Technology remains committed to its mission of “AI-Empowered Precision Medicine,” collaborating with global clinical partners to drive continuous innovation and widespread adoption of non-invasive liver disease diagnostic technologies – enabling more patients to benefit from early screening and precision diagnosis.
Liver Health.Anytime.Anywhere.
Original Article Link:
https://onlinelibrary.wiley.com/doi/10.1111/liv.70634