Seoul National University Hospital has developed a groundbreaking medical large language model (LLM) using a corpus of 38 million clinical texts. The project, which began in March of last year, involved de-identifying and anonymizing various records to create a base for model learning. By early 2025, the team integrated department-specific knowledge bases into the LLM using retrieval-augmented generation. These bases included local medical laws, treatment guidelines, and medical terminology in Korean. When tested on the Korean Medical Licensing Examination from the past three years, the LLM achieved an impressive score of 86.2%, surpassing the average human score of 79.7%. Additionally, the model demonstrated high translation performance, capable of processing 50,000-word texts simultaneously. The hospital plans to verify the LLM’s performance and safety for a year before applying it to clinical work and research, with intentions to expand its use across various medical fields.
Andreas Melhede bio/acc
@AMelhede
·
16h AI in healthcare is inevitable
AI just outperformed doctors in diagnosing common health issues during virtual visits:
Beat them 2-to-1 when recommendations differed
But here’s the catch – docs were better at adapting to changing patient stories & visual diagnoses.
When do you
allon bloch
@allonbloch
·
Apr 4 We are announcing today the results of a major study of K Health’s AI Physician Mode study with Cedars Sinai. Published in the Annals of Internal Medicine.
The bottom line is K’s AI has performed better than doctors in a whole host of acute complaints in real life primary care
npj Digital Medicine
@npjDigitalMed
·
Apr 6 This large-language model predicting fitness for surgery exhibited an absence of hallucinations and produced more consistent output than humans.
https://nature.com/articles/s41746-025-01519-z…
The Jerusalem Post
@Jerusalem_Post
·
13h Study | Artificial intelligence may be better than physicians at making clinical decisions in virtual urgent care settings – at least when it comes to common medical conditions.
Vazquez, Ø
@VGO3696
·
14h Code Beats Coat: AI’s Got the Cure, Humans Still Have the Charm
AI-Driven Diagnoses and Treatment Recommendations Outperform Physicians, Study Reveals
K Health
@khealth
·
Apr 7 K Health’s AI surpasses physician performance in real-world acute care, landmark study reveals. #HealthcareAI #DigitalHealth #AI
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
