BEGIN:VCALENDAR
VERSION:2.0
PRODID:talks.ox.ac.uk
BEGIN:VEVENT
SUMMARY:Scalable and low-cost federated learning in the NHS using micro-co
 mputing - Andrew Soltan (University of Oxford)
DTSTART;VALUE=DATE-TIME:20250306T180000Z
DTEND;VALUE=DATE-TIME:20250306T190000Z
UID:https://talks.ox.ac.uk/talks/id/a0c7161f-9191-4a56-bd4f-3f71e8143c17/
DESCRIPTION:Training fairer medical AI needs diverse data\, but hospitals 
 are restricted in what they can share for privacy reasons. Here\, I will d
 iscuss our new\, easy-to-deploy way for hospitals to take part in AI devel
 opment without sharing data\, and our learnings from a pilot deployment ac
 ross 4 NHS Trusts. Federated learning (FL) was first developed by research
 ers at Google as a way to train AI models without moving data. Researchers
  at NVIDIA\, Rhino Federated Computing and University of Pennsylvania have
  since deployed FL in to hospitals to develop clinical models\, but deploy
 ment relied on specialist technical expertise at every hospital taking par
 t. Using cheap micro-computers\, we built a platform for any hospital to e
 asily take part in training and testing AI models without needing to share
  patient data. We developed software for FL and loaded it on to Raspberry 
 Pi 4B devices\, delivering ‘ready to go’ federated clients to hospital
 s. Using our approach\, four NHS hospital groups developed and evaluated a
  COVID-19 screening test while retaining full custody of their data throug
 hout\, together building a more performant model. By making it easier to t
 rain models without moving data\, we hope our new full-stack federated lea
 rning approach may lead to better and fairer models\, while respecting pat
 ient privacy and data sovereignty.\n\nTeams link: https://teams.microsoft.
 com/l/meetup-join/19%3ameeting_ODFkMmFjODUtNWYxYi00OGFiLWFiZjQtOTBkNzZkZjU
 wYzQx%40thread.v2/0?context=%7b%22Tid%22%3a%22cc95de1b-97f5-4f93-b4ba-fe68
 b852cf91%22%2c%22Oid%22%3a%22e230fa69-f5f7-4a56-935e-7c0582e568dd%22%7d\n\
 nBio:\nClinical academic at Oxford\, Profile: https://www.oncology.ox.ac.u
 k/team/andrew-soltan  \nPaper: https://www.thelancet.com/journals/landig/a
 rticle/PIIS2589-7500(23)00226-1/fulltext \nPodcast: https://open.spotify.c
 om/episode/4KfI1GUjS3nzlYKoW8AgVs\nSpeakers:\nAndrew Soltan (University of
  Oxford)
LOCATION:Wolfson College (Levett Room)\, Linton Road OX2 6UD
TZID:Europe/London
URL:https://talks.ox.ac.uk/talks/id/a0c7161f-9191-4a56-bd4f-3f71e8143c17/
BEGIN:VALARM
ACTION:display
DESCRIPTION:Talk:Scalable and low-cost federated learning in the NHS using
  micro-computing - Andrew Soltan (University of Oxford)
TRIGGER:-PT1H
END:VALARM
END:VEVENT
END:VCALENDAR