The past, the present and the future Clinical Research Software for Intensive Care Monitoring Dr Peter Smielewski Dept of Clinical Neurosciences University of Cambridge
Disclosure ICM+ software is licensed by Cambridge Enterprise Ltd, subsidiary of University of Cambridge, UK. Gentlemen pictured above (Dr M. Czosnyka, left, and Dr P. Smielewski, right) have financial interest in part of the licensing fee for the ICM+ software
1980 1990 2000 2010 1982 Warsaw University of Technology University of Cambridge First ICP Analyser Sinclair ZX Spectrum!
1980 1990 2000 2010 1985 Warsaw University of Technology University of Cambridge ICP Analyser for DOS ICP via Analogue output
1980 1990 2000 2010 1991 Warsaw University of Technology University of Cambridge Intensive Care Monitor - ICM
1980 1990 2000 2010 1991 Warsaw University of Technology University of Cambridge ICM trends examples
1980 1990 2000 2010 1995 Warsaw University of Technology University of Cambridge Biological Signals Analyser Software for real time analysis of cerebrovascular reactivity tests
1980 1990 2000 2010 ICM+ begins Warsaw University of Technology University of Cambridge 2000
1980 1990 2000 2010 Warsaw University of Technology University of Cambridge 2002 ICM+ replaces ICM in the NCCU ICM ICM+
1980 1990 2000 2010 Warsaw University of Technology University of Cambridge 2004 Cambridge Enterprise (wholly own subsidiary of Cambridge University) takes over IP rights to ICM+ and starts offering its licenses to other clinical research centres First ICM+ Installation: University Clinique, Brescia, Italy Prof. N. Latronico, Dr F. Rasulo
1980 1990 2000 2010 Warsaw University of Technology University of Cambridge 2005 First Digital Interface: Phillips monitors Department of Neurology, Medical Centre Haaglanden, The Hague, Netherland Joseph T Tans
1980 1990 2000 2010 Warsaw University of Technology University of Cambridge 2009 The plug-in interface The first plugin Non-invasive ICP by B Schmidt and R Plontke, Chemnitz, Germany Real ICP DE 19600983 Real ICP Non-invasive ICP Non-invasive ICP
1980 1990 2000 2010 Warsaw University of Technology University of Cambridge 2010 First ICM+ users group meeting
Data collection interfaces Analogue Digital (Serial Interface) Ascii continuous export Client/server model Proprietary language Coax cable 0.050 0.021 0.859 0.382 0.416 0.000 0.000 0.055 0.016 0.861 0.392 0.416-0.000 0.000 0.060 0.022 0.856 0.398 0.403-0.000 0.000 0.065 0.026 0.859 0.402 0.407-0.001-0.000 0.070 0.024 0.851 0.410 0.427-0.000-0.001 0.075 0.033 0.846 0.408 0.446 0.000-0.001 0.080 0.026 0.858 0.410 0.467 0.000-0.000 0.085 0.031 0.850 0.419 0.475-0.001-0.000 0.090 0.038 0.848 0.387 0.486-0.000-0.000 0.095 0.039 0.834 0.350 0.483-0.000 0.000 0.100 0.047 0.842 0.385 0.486-0.000 0.000 0.105 0.044 0.837 0.441 0.482-0.001-0.000 0.110 0.057 0.833 0.481 0.485-0.001-0.001 0.115 0.054 0.826 0.499 0.529-0.001 0.000 Request Reply
DWL TCD Analogue interface HP Merlin Marquette Drager Spacelab Surgical Display Controller
Signal calibration Analogue interface [V] [mmhg] 400 200 0-200 -400 HP monitor module Coax cable A/D interface configuration A/D Converter (ADC) Label ADC Channel Calibration (voltage correspondence)
Digital Interface network or serial (COM Port) COM Port on a desktop PC Monitor side connectors USB Serial adapter
Continuous ASCII data export Continuous ASCII stream from the monitor COM Port Configuration NIRS 0.050 0.021 0.859 0.382 0.416 0.000 0.000 0.055 0.016 0.861 0.392 0.416-0.000 0.000 0.060 0.022 0.856 0.398 0.403-0.000 0.000 0.065 0.026 0.859 0.402 0.407-0.001-0.000 Parsing configuration
ICM+ Registration Info Client/server interface Proprietary data exchange protocol Installed monitors Monitor connection configuration Available modality selection ICP ART SAT HR CO2 TEMP Valid values range specification
Data collection configuration Analogue ASCII Client/Server
Analysis pipeline Data acquisition Trends display Monitors ICM+ Collected signals Real-time analysis pipeline ICM+ Virtual Signals Primary Analysis Final Analysis Σ Statistical summary function Σ
Configuration example - RAP Input signal ICP Ouput trend RAP Virtual Signals Primary Analysis Final Analysis Mean() Amplitude() Correl() Virtual signals Primary analysis Final analysis ICP = icp ICP = Mean(ICP) Amp = FundFrq(ICP,...) RAP = Correl(ICP,Amp,...)
Configuration example Advanced mode
Configuration example - Tau Input signals FV ABP C CVR CBV a CVR S S a a mabp mfv t t 0 ( FV, C a acbv aabp mfv ) dt Virtual Signals Primary Analysis Secondary Analysis Final Analysis ABP ABP FV FV Mean(ABP) Amplitude(ABP) Mean(aCBV) Mean(aABP) Integrate(FV MA(FV,300)) Mean(FV) Ca * CVR CBV Amplitude(CBV) Mean(ABP) Mean(FV)
Data display
Data display - configuration
Intervention tests analysis Transient Hyperaemic Response Test CO2 reactivity test Leg cuff autoregulation test CSF infusion test Saline infusion period Critical pressure limit Overdrainage test Overdrain age test limit
Reanalysis of recorded signals Alternative data trends File with raw signals Analysis configurations
Delay Gain Phase Gain DAP Filter Design Real-time calculation engine Virtual Signals with DSP support Frequency Step Response Impulse response editor
Real-time calculation engine summary functions with advanced options support
Signals calculator Trend removal using polynomial fit Icp Polyfit(icp,21)
Text import configuration variable names parsing \w+
Text import configuration data records parsing
Text import configuration preview and signals selection
Plug-in system DLL library Implementing DLL library DSP Implementing interface DLL library DSP Implementing interface DSP interface DLL library Implementing DLL library Stats Implementing interface DLL library Stats Implementing interface Stats interface DLL library Implementing DLL library Chart Implementing interface DLL library Chart Implementing interface Chart interface DLL library Implementing DLL library Tools Implementing interface DLL library Tools Implementing interface Tools interface DSP (Signal Calculator) Stats functions (Real-time analysis) Charts ICM+ Tools (Intervention tests)
nicp plugin for ICM+ Real ICP Non-invasive ICP Real ICP Non-invasive ICP Schmidt B at al: Adaptive noninvasive assessment of cerebral autoregulation and ICP. Stroke 34:84-89; 2003 DE 19600983
Artefacts Management Selected artefact period is excluded from the signals Artefacts can be removed either: 1. manually, using artefacts selection tool, or 2. automatically, automatically, using the analysis engine. This establishes valid range for the afvm parameter mfvm parameter is defined as amplitude of pulse wave This ensures the data is process only when afvm parameter is free from artefacts
icp icp manual artifacts markup editor Global Artefacts Series Artefacts Global artefact => Data gaps Series artefact => NAN values
automatic artifacts detection ABP Artefact characteristics loss of pulsatility 9 seconds ABP 24 hours
Results of automatic removal of ABP artefacts
batch analysis process
batch analysis configuration
batch analysis results
ICM+ Tools Written to ease the pain of summarising large volumes of ICM+ batch analysis data Allows for outputting summary measures of column data Mean, Median, Std Dev, Max, Min, etc etc Tabulates results by filename More sophisticated summaries being developed
Summary Spreadsheet
Event data fields
Event categorical data field
Example of an event data input including a link field
Example of an event view dialog showing a link data field. Clicking on the link opens an appropriate viewer
New user login
There are many promising methods of data analysis and display available for intensive care and more are being invented. Most of the methods do not see wider clinical application, usually the clinical studies presented are more of a proof of concept This is mostly because, at the moment, they require specialised software and/or engineering support, not widely available in clinical centres. Industry support is essential to incorporate new methods into general clinical practice. However this normally requires large scale studies/clinical trials to justify new investments and ensure medical governing body approval. Interim/hybrid solutions of integrating research orientated data analysis solutions into standard clinical monitoring might be a way of facilitating transfer of new technologies into industry supported medical practice.
ICM+: A multicentre collaborations platform Steadily growing community - critical mass for multicentre collaborations?