~/talk

A novel way to manage and represent unstructured clinical data

links:none on file
slides:none on file

An automated pipeline extracts text from PDFs with Document AI, runs the Google Healthcare NLP API to pull entities, relationships, temporality and a per-entity confidence score, maps linked concepts from 17 vocabularies into a FHIR bundle (with Device/DeviceDefinition, Provenance and extensions for confidence and offsets), persists results to BigQuery and Firestore, and surfaces dashboards; handles paper-to-FHIR at scale (example dashboard finds three COPD phrase variants) and leaves low-confidence items for downstream review.

AI-written summary · about

  1. 0:00Introduction and agenda
  2. 0:23Overview of Google Healthcare NLP API
  3. 1:03Unstructured data challenges in healthcare
  4. 3:10Knowledge extraction from clinical text
  5. 4:13Context assessment and certainty analysis
  6. 5:17Standard vocabulary mapping with NLP
  7. 6:00Live demo of NLP API in action
  8. 8:05Mapping NLP output to FHIR resources
  9. 10:57Using FHIR extensions for NLP insights
  10. 13:04Complete FHIR bundle example
  11. 14:53Building scalable data pipelines
  12. 16:17Automating NLP with Apache Beam
  13. 17:21Preprocessing PDFs with Document AI
  14. 18:24End-to-end automation example
  15. 21:17Storing NLP results in BigQuery and Firestore
  16. 21:58Creating dashboards from NLP insights
  17. 23:45Summary and ecosystem integration
  18. 24:50Q&A: Market standards and vocabularies
  19. 27:20Q&A: Provenance and NLP source indicators
  20. 28:02Q&A: Context and patient narrative importance
  21. 29:27Q&A: Common NLP use cases in healthcare
  22. 31:19Q&A: Structured vs unstructured data integration
  23. 33:04Q&A: NLP for non-clinical documents
  24. 33:51Q&A: Using confidence scores in practice
  25. 35:18Q&A: Language and vocabulary support
  26. 36:44Q&A and closing remarks
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.