BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Can machine learning help identify new health relevant signals fro
 m large accelerometer datasets? - Aiden Doherty (Nuffield Department of Po
 pulation Health\, University of Oxford)
DTSTART:20200907T140000Z
DTEND:20200907T150000Z
UID:TALK149641@talks.cam.ac.uk
CONTACT:Lorena Qendro
DESCRIPTION:*Abstract:* Cardiovascular disease (CVD) prevention strategies
  include the typical use of risk stratification and prediction tools to ta
 rget preventive interventions for people at higher risk of CVD and recomme
 ndations on modifiable CVD risk factors such as levels of physical activit
 y. However\, most people with CVD are identified too late and over 50% of 
 major cardiac events are in patients who were not classified as high-risk.
  In addition\, a reliance on crude self-reported questionnaires could mean
  that CVD behavioural risk factors such as physical activity and sleep dur
 ation are more important than previously thought.\n\nWearable sensors such
  as wrist-worn activity trackers (accelerometers) have the potential to co
 ntinuously\, noninvasively\, and painlessly measure CVD risk factors in pa
 tients’ everyday lives. For example\, our group has worked closely with 
 UK Biobank to measure physical activity status in 103\,712 participants wh
 o agreed to wear a wrist-worn device for seven days. These measurements ar
 e now actively used by health researchers worldwide to demonstrate associa
 tions between physical activity and CVD. Machine learning methods can help
  maximise the utility of data from wearable sensors. However\, there is a 
 broad concern around the lack of reproducibility of machine learning model
 s in health data science. It is critical to carefully consider how to prom
 ote robust machine learning findings and reject irreproducible ones\, to e
 nsure credibility and trustworthiness.\n\nIn this talk I will share my gro
 up’s work on reproducible machine learning of wearable sensor data for t
 he early detection of cardiovascular disease. This will include methods to
  identify physical activity behaviours in a free-living validation dataset
  of ~150 adults. I will then illustrate the genetic architecture of these 
 measurements. I will also show that these measurements have a clear utilit
 y to predict future CVD outcomes. Finally\, I will discuss the opportuniti
 es for wearable sensors to advance the prevention of CVD.\n\n*Bio:* I am a
 n Associate Professor at the University of Oxford and lead Health Data Res
 earch UK’s national implementation project on reproducible machine learn
 ing. My research group at Oxford develops reproducible methods to analyse 
 wearable sensor data in very large health studies to better understand the
  causes and consequences of disease. For example\, we have developed metho
 ds to objectively measure physical activity in UK Biobank which are now ac
 tively used by researchers worldwide to demonstrate new associations with 
 cardiovascular disease\, depression\, mood disorders\, and others. We have
  also developed machine learning methods to identify sleep and functional 
 physical activity behaviours such as walking. In addition\, we have discov
 ered the first genetic variants associated with machine-learned sensor phe
 notypes. This work shows the first genetic evidence that physical activity
  might causally lower blood pressure. In 2015 I was one of only three EU M
 arie Curie Award winners (from ~9000 fellowship holders)\, selected for my
  contributions to health sensor data analysis. I have also contributed to 
 the creation of guidelines on the use of mobile devices in clinical trials
 \, in collaboration with the US Food and Drug Administration (FDA) support
 ed Clinical Trials Transformation Initiative on “Mobile Clinical Trials
 ”.
LOCATION:Virtual (see abstract for Zoom link)
END:VEVENT
END:VCALENDAR