BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Institute of Biomedical Engineering (BME) - ECPv6.17.4.1//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Institute of Biomedical Engineering (BME)
X-ORIGINAL-URL:https://bme.utoronto.ca
X-WR-CALDESC:Events for Institute of Biomedical Engineering (BME)
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/Toronto
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20210314T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20211107T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20220313T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20221106T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20230312T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20231105T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/Toronto:20221102T120000
DTEND;TZID=America/Toronto:20221102T123000
DTSTAMP:20221101T203731Z
CREATED:20220828T170755Z
LAST-MODIFIED:20221101T203731Z
UID:10000072-1667390400-1667392200@bme.utoronto.ca
SUMMARY:Graduate Seminar Series: Clinical Stream - Pratik Kumar Mishra
DESCRIPTION:Graduate Seminar Series: Clinical Stream\nGraduate Seminar Series for the Institute of Biomedical Engineering (BME). This day is for clinical stream presenters.\nIf you would like to invite your Principal Investigator\, please add their email via the ‘Add Guest’ button and they will also be notified of your presentation.\nPresentation Title: Automatic Detection of Behaviours of Risk in People with Dementia using Unsupervised Deep Learning\nAbstract: Behavioural symptoms of dementia present a significant risk within long-term care homes. Due to limited staffing resources\, supporting residents and monitoring their safety becomes more challenging. Many long-term care facilities have video surveillance systems installed in common areas to help staff observe residents; however\, the videos are not always monitored by staff. The videos contain vital information that can be used to build deep learning solutions that automatically identify clinically important behaviours of risk and alert the staff. Such an environmental monitoring system can facilitate timely intervention by staff to prevent accidents or harm to the residents and better support their care. We developed a customized spatio-temporal convolutional autoencoder that learns the characteristics of the normal video scenes and identifies the scenes with events of behaviours of risk as anomalies. To avoid privacy-related concerns associated with video data\, as they contain information related to facial identity and the overall appearance of the participants\, the participant’s identity was anonymized before being fed as input to the autoencoder by extracting only the postures or masking the appearance of the participants. We present preliminary results using video surveillance data with a single dementia participant and a single camera collected from a dementia care unit. Our approach gave an AUC of ROC of 0.808 for the best-performing privacy-protecting approach\, surpassing the performance of raw video input. This research paves the way for leveraging existing surveillance infrastructure in care homes to detect behaviours of risk in people with dementia while preserving their privacy\nSupervisor Name: Shehroz Khan\nYear of Study: 2\nProgram of Study: PhD\nZoom link: https://us02web.zoom.us/j/89610372821?pwd=azd4SCtYVWtreVovaGNPV1c2NGY2Zz09\nMeeting ID: 896 1037 2821\nPassword: 483329\nPowered by Calendly.com
URL:https://bme.utoronto.ca/event/graduate-seminar-series-clinical-stream-pratik-kumar-mishra/
CATEGORIES:Graduate Seminar Series
END:VEVENT
END:VCALENDAR