DBTB INT Sanghamitra Deb r
summary
Describes an NLP pipeline that extracts side effects and target age groups from public clinical text, not by running new trials but by mining existing records and labels; shows that attributing outcomes to weight, ethnicity and gender produces clinically meaningful splits (example: 200‑lb vs 100‑lb patients get different optimal drugs). Emphasizes that data cleaning and label quality matter more than fancy models and that assembling dispersed datasets is the main engineering challenge.
AI-written summary · about
related talks
26mNatural Language Processing for Healthcare: Current Trends...
33mLessons Learned Applying Large Language Models in Healthcare
31mBig data for tiny patients: Applying ML to pediatrics
11mEnsuring Reliability in Healthcare Data Engineering
21mDataEngBytes 2023 - SYD-T2-04 - Suneeta Mall
35mBias in NLP 101
33mData Science in Healthcare: Beyond the Hype
06mAdvancements in Natural Language Processing (NLP) and Text Analytics