Events Add an event Speakers Talks Collections
 
Duration 02:58
16+
Video

B19 Preparing a Clinical Support Model for Silent Mode in General Internal Medicine

Bret Nestor
PHD Researcher at Vector Institute
+ 9 speakers
  • Video
  • Table of contents
  • Video
Machine Learning for Healthcare
August 7, 2020, Online, Los Angeles, CA, USA
Machine Learning for Healthcare
Request Q&A
Machine Learning for Healthcare
From the conference
Machine Learning for Healthcare
Request Q&A
Video
B19 Preparing a Clinical Support Model for Silent Mode in General Internal Medicine
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Add to favorites
85
I like 0
I dislike 0
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
  • Description
  • Transcript
  • Discussion

About speakers

Bret Nestor
PHD Researcher at Vector Institute
Liam McCoy
Junior Fellow at Massey College
Amol Verma
Internist and Clinician Scientist at University of Toronto
Chloe Pou-Prom
Sessional Lecturer at University of Toronto
Joshua Murray
Director, Advanced Analytics at MSc
Sebnem Kuzulugil
Data Science and Advanced Analytics at Unity Health Toronto
David Dai
Data Scientist at St. Michael's Hospital
Muhammad Mamdani
Director at St. Michael's Hospital
Anna Goldenberg
Scientist at The Hospital for Sick Children
Marzyeh Ghassemi
Assistant Professor at University of Toronto

I completed my undergraduate degree in mechanical engineering from the University of British Columbia, where I focused on control systems and digital microfluidics. I wanted to apply my skills to solve pressing human problems so I decided to pursue my masters degree in bio-engineering from the University of California, Berkeley. With one foot in the biological world and one foot in the mechanical world, I worked with engineers, biologists and clinicians at the Wyss Institute for Biologically Inspired Engineering at Harvard Medical School for two years. My research evolved from fabrication of microfluidic devices to computational methods to evaluate drug screening systems for sepsis. I am now at the University of Toronto pursuing my PhD in Machine Learning for Healthcare under the supervision of Marzyeh Ghassemi and Anna Goldenberg.

View the profile

I am a medical student with a deep interest in innovation and the effective and ethical integration of emerging AI and data science technologies into medical practice. In addition to my M.D. studies, I am completing the U of T Computing for Medicine program, and developing my skills in machine learning and data science with a focus upon effective use of clinical data. I am also pursuing a Master of Science degree in Systems Leadership and Innovation, examining health issues from a larger systems lens, and considering the micro- and macro-scale changes necessary for effective change. Projects I am involved in include the development of cancer precision medicine guidelines at the Princess Margaret Research Institute, work on early prediction of sepsis from clinical data, and applied ethics work regarding the relationship of AI to clinical decisionmaking and medicine's evidentiary burdens. I am constantly interested in meeting new people and exploring new opportunities - reach out if you are interested in collaboration, or you are simply also interested in these topics and want to connect!

View the profile

I am an internist and epidemiologist, with specialization in observational methods, ‘big data’ projects, and global health. I often collaborate with researchers across many disciplines, such as geographers, economists, and computer-scientists on research areas that require diverse expertise. My work is primarily on the General Medicine Inpatient Initiative (GEMINI) a data driven quality improvement and research network for all general medicine hospital patients in Ontario. I am Co-Principal Investigator (Co-PI) with Amol Verma (University of Toronto) on GEMINI, and we work closely with physicians and health care leaders across the province. GEMINI has developed infrastructure to extract and standardize electronic clinical data from hospital IT systems (laboratory data, radiology, admission-discharge-transfer, etc.) at 7 large hospital sites and is expanding to include many more hospitals in the next 3 years. We are harnessing this ‘big data’ for diverse research applications and to form the ‘backbone’ of a quality improvement network.

View the profile

I work at the Li Ka Shing Centre for Healthcare Analytics Research & Training (LKS-CHART) of St. Michael's Hospital, where we leverage healthcare data in order to make better decisions, help save lives and transform patient care. I previously worked as a research assistant at Toronto Rehabilitation Institute and at the Vector Institute for Artificial Intelligence, where I looked at natural language processing (NLP) and machine learning (ML) for health applications. When I'm not working, I love reading and bouldering.

View the profile

I worked on a social network data analysis of long term care facilities in Alberta. The work included handing out weekly feedback reports, secretly following the nurses, and writing notes like “threw report in garbage (*crumpled first)”. I also had a lot of fun building a web application to analyze time series data here at CHART.

View the profile

Dr. Mamdani is Vice President of Data Science and Advanced Analytics at Unity Health Toronto. Dr. Mamdani’s team bridges advanced analytics including machine learning with clinical and management decision making to improve patient outcomes and hospital efficiency. Dr. Mamdani is also Professor in the Department of Medicine of the Temerty Faculty of Medicine, the Leslie Dan Faculty of Pharmacy, and the Institute of Health Policy, Management and Evaluation of the Dalla Lana Faculty of Public Health. He is also adjunct Senior Scientist at the Institute for Clinical Evaluative Sciences (ICES) and a Faculty Affiliate of the Vector Institute, which is a leading institution for artificial intelligence research in Canada. Dr. Mamdani holds a Doctor of Pharmacy degree from the University of Michigan, a fellowship in pharmacoeconomics from the Detroit Medical Centre, a Master of Arts degree in econometric theory from Wayne State University, and a Master of Public Health from Harvard University with a focus on statistics and epidemiology. He has previously been named among Canada’s Top 40 under 40. Dr. Mamdani’s research interests include pharmacoepidemiology, pharmacoeconomics, drug policy, and the application of advanced analytics approaches to clinical problems and health policy decision-making. He has published over 500 studies in peer-reviewed healthcare journals.

View the profile

Dr. Anna Goldenberg is a Senior Scientist in Genetics and Genome Biology program at the SickKids Research Institute, and is also the first Varma Family Chair in Biomedical Informatics and Artificial Intelligence. She is an Associate Professor in the Department of Computer Science at the University of Toronto, faculty member and an Associate Research Director, Health at Vector Institute and a fellow at the Canadian Institute for Advanced Research (CIFAR), Child and Brain Development group. Goldenberg trained in machine learning at Carnegie Mellon University, with a post-doctoral focus in computational biology and medicine. The current focus of her lab is on developing machine learning methods that capture heterogeneity and identify disease mechanisms in complex human diseases as well as developing risk prediction and early warning clinical systems. Goldenberg is a recipient of the Early Researcher Award from the Ministry of Research and Innovation. She is strongly committed to creating responsible AI to benefit patients across a variety of conditions.

View the profile
Share

Welcome to our talk on preparing the clinical support model for silent mode. In General, Internal Medicine, my name is bread and I'll be presenting on behalf of collaborators. At the University of Toronto Unity Health Toronto Vector Institute in the hospital for sick. Children received an early warning system for the General, Internal Medicine unit that alerts clinicians. If a patient will have an adverse event within the next 24 hours, due to Adverse Events are defined as an Unexpected death transferred to the ICU or issuance of new palliative, orders, clinicians request that they're no

more than two false alarms per true alarm. This roughly corresponds to a positive predictive value of 0.4. We don't operate independently train models that are combination of generalized out of models lasso g r, u & G, R U D, train on various modalities and subsets of features during validation. We modulate the individual model thresholds that went three or more models, do it simultaneously, we obtain a positive predictive value of 0.4 on the validation set in the end. The Ensemble is tested on a prospective test that in pink. When we simulate the deployment of this model received at the

positive, predictive value remains near 0.4 on the prospective test set. We also show the corresponding sensitivity for the chosen thresholds. However, the reporting a slightly more nuanced while we formulate our task to raise wires within 24 hours of an adverse event. We see that more than half of our true, positive Lyme's happened before 24 hours. And somewhat ameliorate this by reporting on a person, a basis, rather than a per window basis, this means that if a patient receives an alarm at any point before they experience an adverse event,

it is considered a true positive. This increases the appearance of the sensitivity and positive predictive value of the model. However, neither aggregate metric shows. How many alarms are truly actionable? We also observe the performance changes depending on how long the patient encounter is, the patient is discharged, or experienced an adverse event in the first four days. We have a higher ppv at the expense of sensitivity for patients. With longest days, we just talked over half of them who at some point in their stable experience an adverse outcome. Finally, we look at the clinician

alarm strategy under the regular alarm strategy. We receive 29 true positive than her tests set, this corresponds, to 1.28 alarmed for True positive. We also experienced called false positive corresponding 21.25 alarms for false positive when we added an additional requirement at for an extra vote in the first 30 hours, as well as the 24-hour silence, you. After the first alarm, we can enrich the alarms for True positive. While simultaneously, decreasing the number of wires for false positive patient. Thank you for listening to our talking. Please read the full paper for

expanded commentary.

Cackle comments for the website

Buy this talk

Access to the talk “B19 Preparing a Clinical Support Model for Silent Mode in General Internal Medicine”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free

Ticket

Get access to all videos “Machine Learning for Healthcare”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Ticket

Interested in topic “Artificial Intelligence and Machine Learning”?

You might be interested in videos from this event

February 4 - 5, 2021
Online
26
147
ai, application, bot, chatbot, conversation, data, design, healthcare, ml

Similar talks

John Morrison
Assistant Professor of Pediatrics at The Johns Hopkins University School of Medicine
+ 6 speakers
Ali Jalali
Senior Data Scientist at Biofourmis
+ 6 speakers
Hannah Lonsdale
Clinical Research Associate at The Johns Hopkins University
+ 6 speakers
Paola Dees
Physician at Johns Hopkins All Children's Hospital
+ 6 speakers
Brittany Casey
Pediatric Hospital Medicine Fellow at Johns Hopkins All Children’s Hospital
+ 6 speakers
Mohamed Rehman
Eric Kobren Professor of Applied Health Informatics at The Johns Hopkins University School of Medicine
+ 6 speakers
Luis Ahumada
ДолжностьDirector Center for Pediatric Data Science and Analytic Methodology at Johns Hopkins All Children's Hospital
+ 6 speakers
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
George Adam
President and CEO at VentriPoint
+ 3 speakers
Chun-Hao Chang
Senior BA Consultant at Dapasoft
+ 3 speakers
Benjamin Haibe-Kains
Senior Scientist at Princess Margaret Cancer Centre, University Health Network
+ 3 speakers
Anna Goldenberg
Scientist at The Hospital for Sick Children
+ 3 speakers
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Nicholson Price
Professor of Law at University of Michigan Law School
+ 2 speakers
Leora Horwitz
Director, Center for Healthcare Innovation and Delivery Science at NYU Langone Health
+ 2 speakers
Michael Sjoding
Assistant Professor at University of Michigan
+ 2 speakers
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free

Buy this video

Video
Access to the talk “B19 Preparing a Clinical Support Model for Silent Mode in General Internal Medicine”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free

Conference Cast

With ConferenceCast.tv, you get access to our library of the world's best conference talks.

Conference Cast
946 conferences
37592 speakers
14370 hours of content
Bret Nestor
Liam McCoy
Amol Verma
Chloe Pou-Prom
Joshua Murray
Sebnem Kuzulugil
David Dai
Muhammad Mamdani
Anna Goldenberg
Marzyeh Ghassemi