Medlinker Launches China's First AI Doctor, MedGPT, and Files IPO Prospectus
In recent years, the rapid advancement of natural language processing (NLP) technology has garnered widespread global attention. Notably, the GPT-3 language model developed by OpenAI sparked a technological revolution in 2020. However, the applications of this technology extend beyond the realm of NLP; it can also be leveraged in the healthcare industry to enhance the efficiency and quality of doctor-patient communication and chronic disease management.
April 28, 2023, China's Leading Internet HospitalMedlinker Announces Successful Development of China’s First Large-Model-Powered AI Doctor—medGPT, the product has currently entered the internal testing phase and is scheduled for official release in May this year. MedGPT by Medlinker currently"Already possesses initial diagnostic capabilities for nearly 3,000 diseases,"Covers over 80% of adult diseases and more than 90% of pediatric diseases in children aged 0–12 years.
MedGPT, developed by WeDoctor, leverages large language model technology and is empowered by a medical knowledge graph that provides rich, accurate, and structured medical knowledge. The model has undergone extensive deep training on nearly 2 billion real-world doctor-patient communication transcripts, laboratory test results, and medical records. Furthermore, reinforcement learning based on authentic physician feedback is employed to ensure the quality, accuracy, and reliability of the model’s reasoning. Notably,medGPT has, for the first time, overcome the challenge of AI doctors being unable to engage in continuous, free-flowing conversations with real patients, and supports multimodal input and output in medical consultation scenarios.Achieve comprehensive intelligence across the four critical stages of disease management: prevention, diagnosis, treatment, and rehabilitation.
To validate the diagnostic accuracy of medGPT, Medlinker extracted medical records from 532 follow-up patients, anonymized the data, and conducted simulated initial consultation experiments. The results showed that the diagnostic concordance rate between medGPT and the patients’ original offline outpatient diagnoses exceeded 97.5%, fully demonstrating the value and potential of AI physicians in the healthcare sector.
Wang Shirui, Founder and CEO of Medlinker, stated, “The integration of AI doctors with internet hospitals will bring about an unprecedented chemical reaction in the healthcare sector. We believe that through Medlinker’s medGPT, the work and learning efficiency of physicians will be significantly enhanced. Meanwhile, high-quality medical resources will be rapidly replicated and widely disseminated. Leveraging medGPT, Medlinker Internet Hospital will provide patients with more convenient, efficient, and accessible medical services, thereby further advancing healthcare equity and striving to resolve the ‘impossible triangle’ of healthcare.”
In the medical field, the accuracy, reliability, and logical reasoning capabilities of language understanding and generation are critically important, as they directly impact physicians’ clinical judgments and treatment planning, thereby safeguarding patients’ safety and health.
The emergence of MedGPT by YiLian is set to bring about a revolutionary transformation in the healthcare sector. Its launch marks the beginning of large language model technology playing a significant role in the medical field. It is foreseeable that, with the advent of MedGPT, the internet healthcare industry will develop more rapidly, efficiently, and precisely, creating greater value in medical technology and delivering an improved service experience.
In the future, Medlinker will continue to increase its R&D investment in medGPT and expand the application of AI doctors across a broader range of disease areas. Meanwhile, Medlinker will strengthen collaborations with medical institutions at all levels to promote the adoption of AI doctor technology in primary healthcare services. By continuously optimizing AI doctor technology, Medlinker is committed to providing patients with more precise, efficient, and accessible medical services, thereby achieving the goal of universal healthcare access.
We sincerely invite partners to join us. Please send emails to: wanglei@medlinker.com
1. Data Collaboration: Tertiary Grade A Hospitals and Relevant Medical Data Enterprises
2. Department Development Collaboration: Grade 3A Hospitals, Medical Experts, Digital Health Enterprises, and Digital Therapeutics
3. Multimodal Model Collaboration: Enterprises possessing multimodal model technologies include, but are not limited to, those offering diagnostic assistance models (text, imaging, and laboratory testing), treatment recommendation models, prognosis assessment models, drug discovery models, and epidemiological models.