Researchers have developed TEDDY, a novel foundation model designed to predict the risk of various diseases in children using historical diagnostic data. Trained on millions of ICD-10 diagnoses from over a million children, TEDDY demonstrated superior performance compared to traditional machine learning models and even larger general-purpose language models in predicting disease onset. The model showed significant predictive capabilities for both common and rare conditions, with its accuracy holding across different demographics and remaining effective for predictions made over two years before a diagnosis was officially recorded. AI
IMPACT Establishes a new benchmark for predictive healthcare models using limited data, potentially improving early disease detection in pediatrics.
RANK_REASON Publication of a research paper detailing a new foundation model and its performance benchmarks. [lever_c_demoted from research: ic=1 ai=1.0]
- asthma
- attention deficit hyperactivity disorder
- CNN
- DenseNet
- Hugging Face
- ICD-10
- long short-term memory
- recurrent neural network
- TEDDY
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
