Brain Injury Neuroinformatics

Background

1997 – Formation of BrainIT

  • Formed around the time of the Williamsburg ICP Meeting
  • Need for standardised data collection and shared databases on both compliance and secondary insults.
  • Group interested in brain monitoring technology, specifically the Spiegelberg compliance monitor.
  • Per Enblad, Pelle Nilsson, Tim Howells, Giuseppe Citerio, Juan Sahuquillo, Iain Chambers, Karl Kienning, Charlie Contant and Ian Piper

2001 – 2002: BrainIT -1 - FP5 QLRI-2000-00454 Infrastuctures Programme - Networking Meetings to Define Core Dataset.

  • One year project expanded the group from the 5 members interested in compliance measurement to 22 centres capable of collecting research data.
  • Defined a core-dataset standard for the collection of high resolution intensive care patient data.
  • Conducted a paper based pilot data collection exercise to determine the feasibility of collecting the core dataset in all centres.
  • Technical sub-committee
    • Designed the interface protocols required in each centre to collect the BrainIT data.
    • Discussed and designed a flexible database format to hold the BrainIT group data.

 2003 – 2006: BrainIT-2 - FP5 QLGT-2002-00160 Infrastuctures Programme

  • Developed new software methods including several new tools for collection of the BrainIT core dataset.
  • Acquired patient data from 22 Neuro intensive care centres from 11 EU countries
  • Successfully completed a prospective data collection of data from 262 patients using these new tools. 

2008-2012: AvertIT - FP7 IST-2007-217049 ICT Programme - AVERT-IT- A Bayesian Neural Network Predicting Hypotension

  • Aim to develop a mechanism, for use within intensive and high-dependency care units, to monitor and predict the likelihood of arterial hypotension (low blood pressure).
  • Determined the weighted association between multiple patient parameters and subsequent arterial hypotension.
  • Association used to define the novel Bayesian neural network (BNN), which was trained against the existing BrainIT dataset
  • Undertook an observational study on 60 patients to test the BANN in a live clinical environment.
  • This work was successfully completed and the calculated sensitivity and specificity from using the BANN system in a live clinical environment were found to be 40.09% and 92.57% respectively.

2015 – Chief Scientists Office (CSO) – Prof. Chris Williams, University of Edinburgh

  • Found that approx. 30% of potential arterial hypotension events are not quantifiable due to either missing data or artifact from blood sampling, patient handling or other clinical interventions.
  • Investigated approaches to automatically detecting and cleaning blood pressure artifact from the raw data.
  • Developed two models for detection of artifact in blood pressure time-series data:
  • Factorial Switching Linear Dynamical Systems (FSLDS) model
  • Discrimitive Switching Linear Dynamical Systems (DSLDS) model.