Clinical support devices and clinical care environments were not designed with data collection and archiving in mind. The traditional paradigm in medicine is that physiological data is assessed in real time by clinicians at the bedside to draw conclusions about the condition of the patient and not stored for subsequent analysis. Our research group collects all physiological waveform data on all patients at all times during their stay in the Critical Care Unit with the goal of archiving this information indefinitely.

This retrospective archive is used for research purposes as well as assisting in event review and teaching. We believe that incorporation of insight driven by analytics into the care process in real-time or near real-time will require an ecosystem-level modification of the current care paradigm. The objective of our Clinical Translational Engineering Program is to anticipate how medicine will be practiced in the future and proactively re-engineer our patient care environment with the goal of improving the consistency and quality of data capture as well as creating mechanisms for incorporating analytical insight into the direct patient care in measurable ways.

Domains incorporated under the umbrella of our Clinical Translational Engineering Program include:

  • Systems physiology – understanding physiological state, coupled physiological subsystems and patient trajectory
  • Predictive modeling– incorporation of physiological variables in predictive models of evolving patient physiology
  • Decision support systems
  • Causal inference/causal reasoning– the process of drawing a conclusion about causality based upon conditions of the occurrence of an event.
  • Process engineering
  • Human factors engineering/Ergonomics– understanding how to harness and optimize the human factor in care delivery. This includes understanding of workflow patterns, the built environment and cognitive strategies used by caregivers at the bedside as a means of optimally introducing decision support tools
  • Improvement Science– strategies to scientifically measure the effect of any changes introduced by means of process, outcome and balancing measures.
  • Visualization– Integrative, actionable visualization of more complex patient variables, like state and risk change.
  • Closed loop state control– the goal of designing auto regulatory mechanisms to optimize instantaneous patient state
Machine Learning
August 18, 2021

Machine Learning for Healthcare – Virtual Conference 2021

The development of machine learning models for deployment into the clinical setting requires a substantial…
Deep Learning
June 7, 2021

Novel Deep Learning Architecture Addresses Small Dataset Problem

Labelling tasks for Machine Learning, such as ECG arrhythmia classification or detection of pneumonia from…
Artificial Intelligence
June 16, 2020

SickKids VS Cardiac Arrest: AI That Can See Heart Attacks Coming

Episode 6: Cardiac Arrest When cardiac arrest strikes, it's dangerous, even deadly. What if we…
CompressionPhysiological Data
June 8, 2020

3 Trillion Values in AtriumDB

Today we hit a milestone with our database high frequency physiological signals. AtriumDB now contains…