Two main parts are included:
1. AI-based precision screening software platform for personalized treatment drugs for epilepsy patients, combining advanced AI technology with improved health education and access to specialist care opportunities, through advanced imaging, neurocognitive functions, genetics, and other clinical features training AI algorithms to develop clinical decision support models, aiming to establish a central network platform to improve patients with epilepsy, especially in rural areas, to obtain specialist care opportunities.
2. IPSC-based neural models and brain-like organ models, using pluripotent stem cells (IPSC) induced from the patient's cells to obtain human-derived neurons can construct a very ideal epilepsy disease model, and brain-like organs contain multi-layered cells and tissue structures found in the brain to increase disease models. Individualized epilepsy disease models can be used to test other new drug treatment effects, find other unconventional treatments, or finding other unconventional alternative drugs for high-risk epilepsy patients with drug resistance.
The two techniques can be used independently or complement each other. The AI software platform is mainly used for the fast selection of conventional epilepsy drugs; the stem cell model is mainly used for the selection of new drugs and other alternative drugs.
Clinical pain points solved by platform technology:
1. AI-based drug screening platform: For 70% of patients who respond to drugs, the platform will shorten the time to find suitable epilepsy drugs.
2. IPSC platform: For 30% of drug-resistant patients, the IPSC platform will screen new drugs (existing or under development) that are not currently used to treat epilepsy.
Advantages of AI-based software platform for accurate screening of personalized therapeutic drugs for epilepsy patients:
1. Input signal diversification. Man can use the patient's direct EEG signals, imaging signals, etc. to make decisions, or obtain useful information from the patient's medical history, electronic medical records, clinical diagnosis reports, and other indirect information to make decisions.
2. It contains an advanced BIOBERT model, which can analyze indirect or hidden information of unstructured data types such as electronic medical records, clinical diagnosis reports, medical history, etc., to obtain more effective information and improve the effectiveness and reliability of the decision.
3. As a decision-making assistant software for clinicians, it can replace or shorten the treatment cycle of traditional medication "trial and error method", so that patients can receive timely and effective treatment at the lowest cost and reduce the probability of delay in the disease.
4. It contains advanced drug treatment prediction models to predict drug treatment effects, and quickly and accurately screen the most suitable treatment drugs for individual patients.
Technical advantages of IPSC-based neural models and brain-like organ models:
1. IPSC not only carries the patient's genetic information but can also grow or "differentiate" into a variety of cell lines, including a variety of nerve cell subtypes; It can study the combined effects of genetic variation (multiple SNVs identified in a single patient) and unknown gene damage.
2. IPSC-based models can grow indefinitely and do not bring any risk to patients, so they are very useful for high-throughput screening of potential drugs in patient specific context, and can identify novel and targeted antiepileptic drugs. Such a novel drug screening platform based on human cells can overcome our heavy dependence on traditional rodent models, which hinder the development of antiepileptic drugs.
3. a). These models have also been successfully used in high-throughput drug screening for other central nervous system diseases; b). Brain-like organ models increase the complexity of disease models, which is crucial for accurately simulating the dysfunction of various cell types and brain regions that lead to human epilepsy.