EP. 26 AI Predicts Long Term Disease Risk | Podcast | MinuteSci
**Artificial Intelligence (AI)** and **Machine Learning (ML)** with medical applications, specifically addressing prediction models and regulatory oversight. One source from the **FDA** outlines the regulatory framework being developed for **AI/ML-based Software as a Medical Device (SaMD)**, detailing the agency’s recognition that traditional device regulation is inadequate for adaptive technologies and highlighting various action plans and guiding principles. A systematic review and meta-analysis from **Oxford Academic** assesses the efficacy of ML models, particularly **deep learning** and **random forest**, in predicting **cardiovascular disease (CVD) risk** using electronic health records (EHRs), concluding that while ML models exhibit superior performance compared to conventional methods, significant heterogeneity and methodological issues limit current clinical adoption. Complementary news sources describe a new generative AI model, **Delphi-2M**, which can simultaneously forecast the risk and timing of over 1,000 diseases, demonstrating the advanced capabilities of AI in **predictive medicine** but emphasizing the need for further testing before clinical use.
@MinuteStory1st
#daily #science #story #podcast #ai
**Artificial Intelligence (AI)** and **Machine Learning (ML)** with medical applications, specifically addressing prediction models and regulatory oversight. One source from the **FDA** outlines the regulatory framework being developed for **AI/ML-based Software as a Medical Device (SaMD)**, detailing the agency’s recognition that traditional device regulation is inadequate for adaptive technologies and highlighting various action plans and guiding principles. A systematic review and meta-analysis from **Oxford Academic** assesses the efficacy of ML models, particularly **deep learning** and **random forest**, in predicting **cardiovascular disease (CVD) risk** using electronic health records (EHRs), concluding that while ML models exhibit superior performance compared to conventional methods, significant heterogeneity and methodological issues limit current clinical adoption. Complementary news sources describe a new generative AI model, **Delphi-2M**, which can simultaneously forecast the risk and timing of over 1,000 diseases, demonstrating the advanced capabilities of AI in **predictive medicine** but emphasizing the need for further testing before clinical use.
@MinuteStory1st
#daily #science #story #podcast #ai
