BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20181221T160905Z
LOCATION:C2/3/4 Ballroom
DTSTART;TZID=America/Chicago:20181114T083000
DTEND;TZID=America/Chicago:20181114T170000
UID:submissions.supercomputing.org_SC18_sess323_post201@linklings.com
SUMMARY:Fast and Accurate Training of an AI Radiologist
DESCRIPTION:Poster\nTech Program Reg Pass, Exhibits Reg Pass\n\nFast and A
 ccurate Training of an AI Radiologist\n\nWilson, Gundecha, Varadharajan, F
 ilby, Yang...\n\nThe health care industry is expected to be an early adopt
 er of AI and deep learning to improve patient outcomes, reduce costs, and 
 speed up diagnosis. We have developed models for using AI to diagnose pneu
 monia, emphysema, and other thoracic pathologies from chest x-rays. Using 
 the Stanford University CheXNet model as inspiration, we explore ways of d
 eveloping accurate models for this problem with fast parallel training on 
 Zenith, the Intel Xeon-based supercomputer at Dell EMC's HPC and AI Innova
 tion Lab. We explore various network topologies to gain insight into what 
 types of neural networks scale well in parallel and improve training time 
 from days to hours. We then explore transferring this learned knowledge to
  other radiology subdomains, such as mammography, and whether this leads t
 o better models than developing subdomain models independently.
URL:https://sc18.supercomputing.org/presentation/?id=post201&sess=sess323
END:VEVENT
END:VCALENDAR