YUAN High-Tech Powers AI Transformation in Medical Imaging: From Real-Time 3D Surgery to Scalable Cancer Screening
Pioneering AI solutions with Showa Medical University and National Taiwan University Hospital
YUAN High-Tech, a leader in edge AI video and medical imaging solutions, is driving the future of healthcare through groundbreaking real-world deployments. Partnering with top-tier institutions such as Showa Medical University ( Japan ) and National Taiwan University Hospital ( Taiwan ), YUAN is delivering high-impact AI systems — not as prototypes, but as fully operational clinical tools.
A New Dimension in Surgery: 2D-to-3D Visualization Without Hardware Change In collaboration with Showa Medical University
Conventional endoscopic surgery offers only 2D visualization — a limitation that reduces depth perception and increases risk in microsurgical procedures. Traditional 3D systems exist, but their bulky hardware, dual lenses, and glasses requirements limit real-world usability.
Together with Dr. Yosuke Sato, Professor of Neurosurgery at Showa Medical University, YUAN has developed a plug-and-play solution that reconstructs real-time 3D imagery from standard 2D endoscope video. Powered by the NVIDIA IGX Orin™ 700 platform and accelerated by RTX 6000 Ada GPU, the system offers ultra-low latency (~40 ms) depth perception — all without changing existing surgical tools.
To further reduce integration complexity, YUAN engineered the solution around its Holoscan Sensor Bridge — a custom-designed interface layer that seamlessly connects surgical video inputs to IGX’s accelerated AI compute. This eliminates the need for virtual camera emulation, reduces memory overhead, and ensures lossless real-time signal transfer, crucial in safety-critical environments like neurosurgery.
“This AI-based 3D visualization lets us operate with greater confidence and precision,” said Dr. Sato. “It feels natural — as if we’re looking directly into the surgical field.”
Already deployed in live surgeries at Showa University Hospital (Tokyo) and Mackay Memorial Hospital (Taipei), the solution is now also being used in immersive VR-based surgical training, giving medical students the ability to experience realistic depth in complex procedures.
“This transforms how we teach surgery,”said Dr. Du, Chief of Pediatric Surgery at Mackay.“It’s no longer just watching — it's seeing and feeling in 3D.”
AI for Early Detection: Real-Time Liver Tumor Diagnosis in Clinics In collaboration with National Taiwan University Hospital ( NTUH )
While surgery demands precision, public health requires scale — and liver cancer remains one of Taiwan’s deadliest diseases due to late diagnoses. In response, YUAN and NTUH co-developed an AI-powered ultrasound diagnostic system that detects liver tumors in real time using Jetson Orin NX, optimized with YUAN’s edge video inference stack.
Trained on over 6,000 annotated ultrasound images, the model identifies malignant and benign tumors with high speed ( 9.05 ms per frame ) and impressive mAP@0.5 accuracy (54.41%). Embedded DICOM and PACS integration ensures clinical readiness, while low-latency inference allows instant feedback during live scans, empowering primary care physicians to screen patients without radiologist support. This liver AI solution is already in active use at NTUH and Good Liver Clinic, showing that edge AI is not just for hospitals — it’s scalable for nationwide screening efforts.
“Our goal is to move AI out of the lab and into every clinic,” said Dr. Nien, physician at NTUH and CEO of Good Liver Clinic.“We’re detecting cancer earlier, with less cost and more speed.”
By integrating Jetson Orin NX, the system directly captures and processes live ultrasound video, enabling frame-level inference at 9.05 ms per frame with clinically validated accuracy (mAP@0.5 = 54.41%). This means physicians can receive AI-generated suggestions during the scan, rather than waiting days for off-site analysis — effectively transforming a general clinic into an AI-assisted diagnostic center.
The system also features built-in DICOM conversion and PACS integration, ensuring seamless documentation into existing hospital workflows. Real-time results help reduce patient anxiety and allow faster decision-making, while also relieving pressure on overburdened radiology departments.
“Patients used to wait days for results. Now we can show them AI-assisted insights immediately, which improves both trust and efficiency,” said Dr. Nien.
Already deployed in clinical settings at NTUH and Good Liver Clinic, the system has proven effective for both routine screenings and follow-up monitoring. With TFDA certification underway, YUAN and NTUH are planning a broader rollout, targeting community clinics, mobile health units, and rural screening programs.
Why YUAN’s IGX and Holoscan Stack Matters
What sets YUAN apart is not just AI model performance, but the complete edge-to-cloud system integration. YUAN’s in-house platforms — such as its custom-configured IGX Orin™ 700 system and Holoscan Sensor Bridge module — are designed specifically for medical-grade AI workloads.
• IGX Orin™ 700 by YUAN: Enhanced thermal design, long-term supply commitment, and medical safety support for OR environments.
• Holoscan Sensor Bridge: Provides direct, low-latency, high-bandwidth pathways between surgical sensors ( endoscope, ultrasound ) and GPU compute, with full support for GMSL2, MIPI, SDI, and HDMI interfaces.
• SDK Optimization: All YUAN solutions are tightly integrated with NVIDIA Holoscan SDK, CUDA Graphs, and DICOM libraries, enabling hospital-ready deployments.
“Our edge AI systems aren’t just powerful — they’re usable, dependable, and tailored to the medical world,” said HP, CEO of YUAN. “We’re here to deliver performance where it’s needed most — in surgery rooms and community clinics.”
From Innovation to Impact
YUAN’s work with Showa and NTUH showcases two ends of the healthcare spectrum: high-end surgical precision, and broad public health accessibility. In both, AI is no longer a vision — it's a working solution, running today.
With over 30 years in professional video and a strong partnership with NVIDIA, YUAN continues to be the bridge between vision science and real-world clinical value.
Whether you're a hospital CIO, a surgical department head, or a medical AI developer — YUAN is ready to partner with you to build solutions that see smarter, operate safer, and diagnose sooner.