Peking Union Medical College Hospital Licenses AI-Powered Ultrasound Imaging Technology for Trigeminal Nerve Identification in Precision Interventional Therapy

Recently, Peking Union Medical College Hospital released a public notice announcing its intention to commercialize a scientific and technological achievement titled “Key Technologies for AI-Based Ultrasound Image Recognition in Precision Interventional Treatment of Trigeminal Neuralgia.” It is reported that this achievement will be licensed on an exclusive basis as know-how.

This technology was developed by Professor Cui Xulei’s team from the Department of Anesthesiology at Peking Union Medical College Hospital. The team has long been dedicated to research in pain medicine and ultrasound-guided nerve blocks. Dr. Cui Xulei is an Associate Chief Physician in the Department of Anesthesiology at Peking Union Medical College Hospital, specializing in the diagnosis and treatment of acute and chronic pain, as well as the application of ultrasound-guided interventional techniques. In 2013, she pursued advanced training in ultrasound-guided nerve block techniques at the Prince of Wales Hospital, The Chinese University of Hong Kong. In 2015, she won the sole championship in the “National Excellent Nerve Block Video Competition” organized by the Chinese Society of Anesthesiology, Chinese Medical Association. She has also played a leading role in promoting multimodal analgesia protocols and patient-controlled analgesia techniques.

The transferee of this technology is Beijing Visual Perception Intelligence Technology Co., Ltd. The company specializes in the fields of computer vision and intelligent analytics. The core technology underlying this transaction leverages artificial intelligence (AI) to achieve precise identification of the trigeminal nerve in ultrasound images. It aims to support precision interventional therapy for the trigeminal nerve by enhancing the accuracy and efficiency of ultrasound image recognition, thereby providing critical technical support for physicians performing interventional procedures.

The trigeminal nerve is the fifth of the twelve pairs of cranial nerves in the human body and the largest mixed nerve in the craniofacial region, primarily responsible for neural conduction across its three branches: the ophthalmic, maxillary, and mandibular divisions. Due to its dual roles in sensory conduction and motor innervation, the trigeminal nerve is also hailed as the “sensory and motor hub” of the craniofacial region.

Recurrent, paroxysmal, and severe pain occurring within the distribution area of the trigeminal nerve is known as trigeminal neuralgia, a common cranial nerve disorder. Due to its intense pain severity and frequent attacks, it is referred to as “the world’s most painful condition.” This disease predominantly affects middle-aged and elderly individuals, with a slightly higher incidence in women than in men, and right-sided involvement is more common than left-sided.

Trigeminal neuralgia can be classified into primary and secondary types based on etiology. The exact cause of primary trigeminal neuralgia remains incompletely understood; however, most studies suggest that it is associated with vascular compression at the root of the trigeminal nerve. Chronic pulsatile stimulation by blood vessels leads to demyelination of the nerve, resulting in abnormal neuronal discharges and subsequent pain symptoms.

Secondary trigeminal neuralgia is caused by identifiable etiologies, such as intracranial tumors (e.g., acoustic neuroma, meningioma), inflammation (e.g., trigeminal neuritis), trauma, and cerebrovascular diseases. These pathologies directly compress or damage the trigeminal nerve, leading to pain episodes. Compared with primary trigeminal neuralgia, secondary trigeminal neuralgia may also be accompanied by other neurological deficits, such as facial hypoesthesia and masticatory weakness.

Current clinical treatment methods mainly include pharmacological therapy, surgical intervention, and interventional therapy. Among these, interventional therapy has become one of the important treatment modalities due to its minimally invasive and safe characteristics. However, during interventional procedures, precise nerve localization is not only key to surgical success but also a major clinical challenge.

On the one hand, experience-dependent operational methods entail a steep learning curve, requiring young physicians to accumulate extensive practice over time to achieve proficiency, thereby limiting the widespread adoption of the technology.

On the other hand, due to the complex anatomical structure and numerous branches of the trigeminal nerve, traditional localization methods largely rely on clinicians’ experience and manual manipulation. Currently used imaging-assisted techniques, such as X-ray or CT, still suffer from insufficient localization accuracy and a risk of injuring surrounding normal neural tissues. It is important to note that the trigeminal nerve has extensive branching and critical functions; intraoperative injury can lead to complications such as facial hypoesthesia and numbness.

To address these pain points, the team leveraged AI algorithms to develop a “Precision Interventional Therapy” recognition technology, aiming to overcome the limitations of existing clinical protocols, such as reliance on manual identification by physicians, imprecise localization, and a steep learning curve.

In the field of neurointerventional medical devices, technological innovation and iteration are accelerating, with intelligence and precision emerging as clear R&D trends.

From the perspective of market layout, domestic enterprises are actively achieving technological breakthroughs through independent R&D and cross-border acquisitions. Among them, Neva MedTech is committed to building a “full product line” matrix covering hemorrhagic, ischemic, and access-related diseases. Its products, including the Captor Thrombectomy System, flow-diverting stents, and the “Maihe” occlusion hemostasis device, have obtained CE/FDA certifications, accelerating its global expansion.

Sinomed has acquired core technologies for flow-diverting stents through its acquisition of the U.S.-based company eLum. The launched AUCURA™ flow-diverting stent can reach deep, small vessels via a 0.017" microcatheter, featuring excellent deliverability and deployment rates, effectively addressing the clinical challenges of poor radiopacity and limited deliverability associated with traditional stents.

On the international front, Siemens Healthineers has partnered with Stryker to develop neurovascular interventional robots, aiming to integrate imaging, consumables, and robotic systems to enhance surgical precision and efficiency; meanwhile, industry giants such as Medtronic and Boston Scientific are integrating AI and real-time imaging technologies into the field of neuromodulation and stimulation, applying them to stroke rehabilitation and chronic pain management.