Handling and Analyzing Large Telemetry Data Sets
Download as PPTX, PDF0 likes108 views
The document outlines a series of webinars led by Dr. Sarah-Jane Guild focusing on telemetry data collection and analysis in small animal research, specifically in relation to blood pressure and ischemic stroke. It discusses technical aspects of telemetry systems, guidelines for data recording, automation through macros, and analysis techniques for efficient data management. Key topics include scheduling data collection, file saving automation, and considerations for data reliability and analysis throughout the experimentation process.
Handling and Analyzing Large Telemetry Data Sets
- 1. 1 Sarah-Jane Guild, Ph.D. Chief Scientific Officer Kaha Sciences Handling and Analyzing Large Telemetry Data Sets
- 2. Small Animal Telemetry Mini Series Webinar: Therapeutic Relevance of Elevated Blood Pressure After Ischemic Stroke in the Hypertensive Rats November 4th, 2020: To register: www.scintica.com/webinars Webinar: Introduction to Measurement of Tissue Oxygen in Rats Via Telemetry November 11th, 2020: Webinar: Tips and Tricks for Successful Rat Telemetry Blood Pressure Recordings December 2nd, 2020:
- 4. Dr Sarah-Jane Guild Chief Scientific Officer, Kaha Sciences Senior Research Fellow, University of Auckland Handling and analyzing large telemetry data sets
- 5. 5 • All audio from the attendees will be muted during the presentation • Please ask questions as we go using the Q&A button at the bottom of the Zoom meeting window • We will have a question and answer session at the end of the webinar Webinar guidelines Webinar Guidelines
- 6. 6 A little about me…. Sarah-Jane Guild • PhD in Physiology & Electrical Engineering • Officially started with Telemetry Research in 2009 • Chief Scientific Officer, Kaha Sciences (0.5) • Senior Research Fellow, University of Auckland (0.5) • Research interests • Cardiovascular control • Intracranial Pressure • Implantable devices • Optogenetics Home
- 7. 7 Kaha Sciences Telemetry Measurements Atrial Pressure Pulmonary Artery Pressure Bladder Pressure Ventricular Pressure Intracranial Pressure Pressure Gradients Sympathetic Nerve Activity Heart Rate Aortic PressureCardiac Function dP/dt Ocular Pressure ECG* EEG* EMG* EOG* Tissue Oxygen * Mouse telemeter – only measures biopotentials
- 8. 8 Kaha Sciences Telemetry Technology The hardware Rat Telemetry System Mouse Telemetry System • TR181 SmartPads or MT110 tBases • TRM5 or TR5 series rat telemeters or MT10B mouse telemeters • TR190 Configurator system – one required per laboratory • Data acquisition (third party hardware)
- 9. 9 Kaha Sciences Telemetry Technology Wireless Power • SmartPad & tBase use electricity to generate magnetic wireless power field • Field passes through non-metallic surfaces • Pickup/Inductor coil in telemeter takes power from magnetic field and converts it back into electricity. • Telemeter electronics are powered • 24/7 available power • No telemeter refurbishment Wireless Power Field Telemeter power/charging coil
- 10. 10 Kaha Sciences Rat Telemetry System Arterial and Left Ventricular Pressure (TRM54PP)
- 11. 11 Experimental design webinar Recording available on the Kaha Sciences Knowledge Centre or on Demand link on the Scintica Instruments website
- 12. 12 Webinar Outline • Automatic scheduling of data recording • Options for automating file saving and management for telemetry studies • Ways to use macros to automate basic analysis tasks for telemetry data
- 14. 14 How much data should you record? Questions to ask when planning data acquisition • What is your hypothesis? • How will you present your final data? • How will you analyze your data? • What graphs will you draw? • Hourly averages? Daily averages? • Circadian variation? • What statistical tests will you use? • What signals are you recording? • Sampling rates, variability over time
- 15. 15 Scheduled data collection Maybe you don’t have to record 24/7…..
- 16. 16 Why schedule data recording? • Spread data recording throughout the day and night • Better representation of the “true mean” • No need to be present in the lab • Data recorded at the same time each day • Smaller volumes of data to store and analyze
- 17. 17 Why schedule data recording?
- 18. 18 How can we schedule data collection? Example: Scheduler feature of LabChart
- 19. 19 Things to consider • Are my signals stable, will I get good data from every collection period? • Signal artefacts? Animal activity? • e.g. arterial pressure recorded during development of hypertension over weeks may be well suited to data scheduling • e.g. left ventricular pressure may not be if you need data from every single data period and you don’t have perfect catheter placement. Stehlin et al, J Appl Physiol 2013
- 20. 20 Things to consider (continued) • Will I miss anything between the recording periods? • e.g. Seizure or arrhythmia detection may not be appropriate Seizure in mouse EEG CB1704 day 22 post M 1I.adicht BP(mmHg) 70 80 90 100 110 120 130 41 42 43 44 45 46 12-Apr-17 10:53:03.341 AM Skipped beat in arterial pressure
- 21. 21 Things to consider (continued) • What kind of analysis will I do? How long does each recording period need to be? • e.g. spectral analysis to look at low frequency oscillations may need several minutes of stable data • There is no right or wrong answer that applies to everyone • It always comes down to Your experiment, Your data and Your study questions.
- 22. Automating file saving Cloud storage
- 23. 23 Why automate your file saving? • Files can be automatically named including date and time • Stop files getting too big to open and analyze • Consistent file sizes and time periods makes analysis easier support@kahasciences.com
- 24. 24 How do we do it? • Macros • e.g. save data and start new file every 12 hours • Setup LabChart file as you normally would • Units conversion, time mode, sampling rates etc • Setup basic Data Pad columns that you know you will need e.g. means, heart rate etc • Don’t setup complicated channels to calculate or analyze as you go (e.g. Blood Pressure module) • Append LabChart settings file with macro
- 25. 25 How do we do it? NEW !!! • Edit macro • choose base filename template (date and time will be appended to name for each file) • Set location for file storage • Set recording period (default 12 hours)
- 26. 26 Basic macro steps 1. Save file with date and time appended 2. Start sampling and run for 43200 seconds 3. Save the file 4. Select all data 5. Clear the data Repeats until stopped
- 27. 27 What do we see? • Add comments as usual • Keeps running and saving new files until manually stopped
- 28. 28 Output • Files named with date and time recording started • 3 animals with 3 channels each = 9 channels • 12 hours x 9 channels @ 1kHz = ~320 MB
- 29. 29 Things to think about • If recording video, you need to run a slightly different macro to acknowledge deletion of video data • See the Kaha website • How long should you record for? • Keep file size manageable (well under 1GB) • You can only see the data in the latest file (while recording) • Review your data files regularly and/or track mean levels by analyzing data as you go • Think about where you are storing your files • Local PC drive is good for day to day recording but make sure you back them up
- 30. Audience Poll
- 31. Using macros to automate basic analysis
- 32. 32 Analyzing your data files • Automate repetitive tasks • Example: Calculate hourly averages of all the data stored in a single directory • Macro: • Asks for folder location • Opens each file in turn • Adds date and time column to Data Pad (because I forgot to add them during setup) • Sets time mode to Time of Day • Run Multiple add to data pad: 1 hour (3600 seconds) every 1 hour (3600 seconds) • Saves Data Pad as text file (.txt) • Closes LabChart file • Says “No” to “Do you want to Save?” • Requires calling a .vbs file which must be copied to your PC
- 33. 33 Analyzing your data files (continued) • Use batch file to combine individual text files • Open in Excel and sort by Date and Time columns to make sure data is in correct order • Plot and further analyze your data
- 34. 34 Write/Record your own macro • Press record • Do the steps you want to automate • Press stop • Edit the macro text
- 35. 35 A word of caution • Don’t over automate your analysis • e.g. hourly averages combined to daily averages in Excel rather than 12hour average of each file • Make sure to review your data and make sure it is sensible • For very long-term experiments (e.g. months) I analyze data in one- week batches as I go to see long-term trends emerging • Still check signals (& animals) every day
- 36. 36 Summary • Always considering your study outcomes • What data should you collect? • What will you do with your data? • Automate your file saving wherever possible • Setup your files correctly from the start • Calibrations, comments, time mode, Data Pad columns • Use macros to automate basic analysis tasks but always keep in mind what those macros are doing and what your data is showing you • Backup your data !!!!!
- 37. 37 • Example macros on the Kaha website • https://www.kahasciences.com/download/labchart-macros/ • Kaha Knowledge Centre • https://www.kahasciences.com/knowledge-centre/ • General enquiries and pricing • sales@kahasciences.com • sales@scintica.com • Technical and scientific enquiries • support@kahasciences.com • Webinar feedback/suggestions • support@kahasciences.com Questions? ?
- 38. To ask a question, click the Q&A Button, type your question and click send. Any questions that are not addressed during the live webinar will be answered following the event. Please indicate whom you want to address your question. Thank you for participating! Q&A SESSION: Sarah-Jane Guild, Ph.D. Chief Scientific Officer Kaha Sciences