West China Hospital, Sichuan University Licenses Urine-Based Epilepsy Prediction Technology for RMB 500,000

Recently, West China Hospital of Sichuan University released a public notice on the transformation of scientific and technological achievements, proposing to grant an exclusive license for an authorized invention patent.Hangzhou Dian Medical Laboratory Co., Ltd., the licensing fee isRMB 500,000. The inventor of this patented technology is West China Hospital, Sichuan UniversityProf. Chen Lei and Her Team。

The invention patents proposed for transfer in this instance involveA Panel of Urinary Protein Biomarkers for Predicting Epileptic SeizuresThis panel of biomarkers can be used to develop diagnostic kits that effectively predict the risk of seizures within the next 24 hours in patients with epilepsy by measuring the levels of specific proteins in their urine. This technology provides patients with invaluable advance warning, facilitating timely intervention, thereby mitigating the risk of sudden seizures and alleviating the psychological burden associated with long-term prediction uncertainty.

Epilepsy is one of the most common chronic neurological disorders, with the number of patients worldwideOver 70 million. Its onset is fundamentally rooted in the sudden emergence of abnormal, excessively synchronized, and self-sustaining neuronal discharges within the brain's neural networks. This aberrant electrical activity typically lasts from a few seconds to several minutes and may trigger a variety of symptoms, including sudden sensory disturbances, loss of consciousness, or limb convulsions, with symptoms potentially affecting any part of the body.

Due to the high degree of suddenness and unpredictability of epileptic seizures, epilepsy directly or indirectly causes worldwide each yearApproximately 120,000 peopleDeath. Seizures occur without warning, making patients highly susceptible to accidents such as drowning, traffic collisions, severe falls, and even sudden unexpected death in epilepsy (SUDEP). Beyond the direct threat to life, the unpredictability of seizures imposes a significant risk of accidents, social embarrassment, and substantial emergency medical costs.

This uncertainty imposes extensive and profound limitations on patients’ family life, social interactions, education, and career development, rendering the social consequences of the disease often more severe than the physical impact of the episodes themselves.

Furthermore, the social stigma associated with epilepsy and its unpredictable nature often lead to intense feelings of shame, irritability, and chronic anxiety in patients. Notably, this anxiety itself can become a potential trigger for further seizures, creating a vicious cycle of “seizure–anxiety–more seizures,” which further deteriorates the patient’s overall health status.

Therefore, achieving advance prediction and early warning of epileptic seizures is key to improving patients’ quality of life and enabling them to take proactive measures. Currently, the primary prediction techniques used in clinical practice and research rely onElectroencephalogram (EEG)EEG is a technique that records the brain's electrical activity by placing electrodes on the scalp.

Current methods aim to predict whether a seizure will occur within a very short timeframe (e.g., one hour) by continuously monitoring patients’ EEG signals over extended periods and analyzing pattern changes. However, this EEG-based prediction approach has significant limitations: it requires the prolonged use of specialized monitoring equipment, and the warning window it provides is extremely brief. This is often insufficient for patients to comfortably take preventive medication or adjust their schedules and activity plans in advance, thereby limiting its practical clinical value.

Given the limitations of existing prediction technologies in terms of warning lead time and convenience, there is an urgent clinical need for an early warning method that provides a more sufficient response window and is easy to implement. The core advantage of this patented technology lies in its provision ofA Novel Seizure Warning System, this method has achieved significant breakthroughs in principle, convenience, and warning window.

Traditional electroencephalogram (EEG) monitoring relies on capturing abnormal electrical signals from the cerebral cortex, whereas this technology shifts focus toSystems Biologyfrom the perspective of seeking answers. It does so by detecting a specific set of proteins in urine ——Anti-chymotrypsin, transthyretin, immunoglobulin gamma-1 heavy chain constant region, and eukaryotic translation initiation factor 2 alpha subunit——predict risk based on changes in the levels of these proteins. These proteins were not selected arbitrarily; they are closely associated with neuroinflammation and immune responses.

Research has found that one day before a seizure, the patient’s brain may already be in a “subclinical” state of abnormal hyperexcitability. Although this state is insufficient to trigger visibly apparent seizures, it is sufficient to activate immune cells within the brain (such as neutrophils), thereby inducing mild neuroinflammation. This inflammatory response leads to the release of specific proteins into the systemic circulation, which are subsequently filtered by the kidneys and excreted into the urine. Therefore, changes in these urinary proteins serve as early, non-invasive “biological signals” reflecting an impending abnormal storm in the brain.

This type of based onBody Fluid BiomarkersThis approach offers unprecedented convenience. It completely eliminates the need for prolonged, continuous reliance on complex electroencephalography (EEG) equipment; instead, it requires only a single urine sample from the patient, which can be analyzed using established mass spectrometry or immunochemical methods. This enables standardized, large-scale testing in routine clinical laboratories, significantly lowering implementation barriers and reducing the burden of patient compliance.

More importantly, this technology providesApproximately 24 hourswarning window, which constitutes its key clinical advancement. Existing EEG-based predictions typically provide only an ultra-short-term warning of a few hours, often leaving patients insufficient time to implement effective interventions. In contrast, a one-day lead time enables patients and physicians to calmly adopt preventive measures, such as temporarily adjusting antiepileptic drug dosages, avoiding known triggers (e.g., sleep deprivation and alcohol consumption), or arranging for safety supervision in advance.

This not only reduces the risk of accidents on a physical level but also significantly alleviates the chronic anxiety patients experience due to the “unpredictability of seizure onset,” thereby improving their quality of life at its root.

Furthermore, the design of this biomarker panel demonstrates a high degree of flexibility and reliability. The four protein biomarkers can be used in combination to enhance predictive accuracy, or any one or two of them can be selected for combined analysis based on practical testing conditions and cost considerations. This facilitates the broader adoption of the technology across diverse application scenarios.

Clinical validation data indicate that even a change in a single biomarker can provide effective risk alerts, while concurrent changes in multiple biomarkers strongly predict an extremely high risk of onset, thereby achieving stratified early warning of risk levels.

The current landscape of epilepsy detection is characterized by the parallel development of multiple technological pathways, encompassing dimensions such as EEG monitoring, wearable devices, biochemical testing, and genetic testing. Different technologies and products have distinct focuses, respectively addressing diverse needs including clinical diagnosis, home-based monitoring, and medication guidance.

Electroencephalogram MonitoringIt remains the core technology for epilepsy diagnosis and seizure capture, with breakthroughs in algorithm optimization and device form factors in recent years. In traditional clinical settings, long-term video electroencephalography (EEG) is the gold standard for differentiating epileptic from non-epileptic seizures. By synchronously recording electrical brain activity and clinical manifestations, it precisely localizes the seizure onset zone, providing critical evidence for pre-surgical evaluation and the formulation of diagnosis and treatment plans. However, due to limitations in device size and operational complexity, its application is primarily confined to hospital settings.

In the field of intelligent algorithms, Brain-Computer Interface Haihe LaboratoryProfessor Liu Shuang's TeamJointly proposed by the First Medical Center of the Chinese PLA General HospitalCross-Domain Hybrid Self-Supervised Attention Network (MCAN)Achieved a major breakthrough. This model integrates self-supervised learning strategies across the time, spatial, and frequency domains, combining multi-scale feature modeling with an attention mechanism guided by electrode sparsity priors. It addresses key challenges in traditional EEG detection, such as the time-consuming nature of manual annotation, poor cross-subject adaptability, and class imbalance. Validated on four real-world epileptic EEG datasets, the model demonstrates superior detection accuracy and robustness compared to existing methods, with particularly notable performance in reducing false positive rates. This provides a novel framework for the automated and precise analysis of EEG signals.

To address the challenges of out-of-hospital monitoring, wearable devices have become a focal point of industry innovation, breaking through the scenario limitations of traditional EEG equipment.RealViconLaunchedYiruitong S2 Watchis a representative product in this field, equipped withEpicare™ Epilepsy Monitoring Management System, specifically designed for monitoring generalized tonic-clonic seizures (GTCS), integrates medical-grade monitoring capabilities into a conventional smartwatch form factor, balancing privacy protection with wearing comfort.

When a seizure is detected, the device can immediately alert pre-set emergency contacts via phone calls and text messages, addressing safety concerns for patients when they are alone. It also features an intelligent medication reminder function and a mobile application that automatically records seizure data and generates analytical reports, providing continuous data support for diagnosis and treatment. Currently, the global market size for epilepsy monitoring devices is steadily growing. Such domestically produced wearable devices are gradually breaking the dominance of international brands, extending monitoring scenarios from hospital beds to daily life.

Furthermore,Keyue MedicalOther companies are also laying out integrated devices for non-invasive neuromodulation and monitoring, with their developedPolysomnography (PSG) EEG Monitoring DeviceIt enables synchronized acquisition of EEG and ECG data during sleep, and, combined with proprietary analysis software, provides a comprehensive solution for seizure monitoring and treatment efficacy assessment; some products have entered the clinical trial phase.

Electroencephalogram (EEG) monitoring combined with intelligent algorithms demonstrates significant advantages in the accuracy of seizure localization, but is limited by portability and warning duration; wearable devices enable scenario expansion and real-time alerts, yet their ability to recognize non-tonic-clonic seizures still requires improvement; biochemical tests, such as those for urine protein and blood drug concentration, offer distinct benefits in terms of warning duration or medication guidance, but cannot achieve real-time monitoring; genetic testing focuses on etiological diagnosis and struggles to cover all types of epilepsy.

The epilepsy detection industry will accelerate its evolution in three major directions:First, deep integration of multiple technologies: By developing integrated monitoring and neuromodulation devices and linking biomarker detection with wearable real-time monitoring technologies, a closed-loop management system encompassing “early warning, intervention, and assessment” is established, thereby overcoming the limitations of single-technology applications. Second, continuous enhancement of intelligence levels: Leveraging iterative optimization of AI algorithms, this approach further improves adaptability across diverse populations and the precision of seizure recognition, effectively reducing clinical false-positive rates and promoting automation and efficiency in detection processes. Third, accelerated substitution with domestically produced alternatives: Domestically manufactured wearable monitoring devices and diagnostic kits are gradually expanding their market penetration, driven by superior cost-effectiveness and controllable data security, thereby accelerating the dissemination of advanced technologies to primary healthcare settings.

* Patent transaction information is provided by CSTT