Psychophysiology tips

By Ilinca Popovici (Fall 2006)

General GSR tips for experimental setup (startle & emotion):
      (sources: handbook, Ch 8)
      Monitor sources of artifact (events such as coughing, sneezing, sighing, deep
       breathing, moving etc) by flagging during the experiment (For emotion we will
       have a video feed, so we can flag afterwards. For startle I can include the startling
       wave and map it on top of all the other measurements; real-time flagging might be
       hard to do in practice.) (Christina – use the E-prime software)
      (biopac) wait 5 minutes after gel application (Christina - wait 10mins)
     Use “GEL101” (isotonic, 0.5% saline in a neutral base). This should closely
      resemble sweat in its salinity. (checked on google – it does) Otherwise, high
      levels of NaCl inflate measures.
     Wash hands with nonabrasive soap before attaching the electrodes. Do not use
      alcohol since that changes the natural resistance/conductance properties of the
      skin. (Christina - wipe with alcohol)
     Make sure that a consistent skin area comes in contact with the paste, since it is
      the contact area, and not the size of the electrode which affects the conductance
      values. (Christina – use 8mm white colar, apply gel in cavity, wipe off surplus,
      peel colar cover, stick to skin)
     Use the nondominant hand, volar surfaces of distal phalanges (tips of pointing and
      middle finger, on the palm side). If it has too many cuts/calluses, use volar
      surfaces on middle phalanges (middle segment of pointing and middle finger on
      the palm side).
     Startle: what to expect – a common minimum response amplitude is
      0.05microsiemens, and the latency window is 1-4 seconds.
     Compute subjects’ startle range (minimum SCL during rest, maximum when
      blows balloon to bursting -!! Then express SCL as a proportion of the subject’s
      range: (SCL-SCLmin)/(SCLmax-SCLmin)
     STARTLE tone used: 1-sec tone of approx. 75dB
     Since stress is a big factor for GSR, the experimenter has to be as unintrusive as
      possible (for both startle and emotion)
     Clean electrodes after use
     Christina – Do not ground! (Biopac mistake)
     Christina – for startle, use 200samples/sec, gain of 20, filter of 1Hz, DC, DC
     More Biopac Tips (pending)

General emotion tips for experimental setup:
     (sources: handbook, Ch22)
     Incorporate verbal reports of emotion
     Include neutral stimuli (relaxation stage? Clear mind?)
     Use emotions from three groups: high-arousal pleasant, high-arousal unpleasant,
      and neutral (low arousal). Otherwise it will be hard to disassociate effects of
      stimulus arousal from effects of stimulus valence on physiological and behavioral
     Include at least two emotions at each level of valence (2 pleasant, 2 unpleasant) –
      this is for study of discrete emotions
     Minimize our contact with the subjects (may influence their emotions, “emotional
     Use a questionnaire to determine: gender, cognitive style, temperament, and +ve/-
      ve emotionality (p. 632); RSP - Ask subjects about age, smoking, asthma or other
      lung diseases.
     Also ask subjects how they felt during each phase afterwards (validation)
     (sources: “increased feelings with increased body signals” EGG paper):
     Ask the participants to complete a “Positive and Negative Affect Schedule
      questionnaire” which assesses the background mood when the subject starts the

General startle tips for experimental setup:

     (sources: handbook, Ch22)
     Most commonly used startle-eliciting stimulus in human research is a brief (50-
      msec) burst of white noise at around 95 decibels.
     Startle response picture to keep in mind on page 631
     (sources: “Expressive suppression during an acoustic startle” paper)
     If subjects are told to inhibit expressive behaviour, there is increased sympathetic
      activation. Idea for experiment: be consistent, use only random, unanticipated
      startles, maybe also tell subjects not to inhibit response?)
     They focus on a window of 5s before and 6s after the startle presentation
     Use questionnaire for: consent form, health checklist, current emotional state
     Subjects feel self-conscious with facial EMG
     Should film subjects, and exclude artifacts from data afterwards.

     More Biopac Tips (pending) (i.e. use twice the expected sampling frequency for
      each measurement, value TBD )

General EMG tips for experimental setup (startle & emotion):
     (sources: handbook, Ch 7)
     Should monitor whole muscle regions (which is usually the case with surface
      electrodes anyway)
     Careful grounding (lots of interference)
     0.5cm diameter detection surfaces and 1.5cm diameter housings are used
      commonly for limb and trunk EMG, as opposed to face (0.25 cm diameter
      detection surface and 0.5 or 1cm housing)  we can use the 0.4cm diameter
      detection surface electrodes in the lab.
     1cm inter-detector surface spacing recommended
     Clean skin and abrade gently to lower inter-electrode impedances
     Filtering: a pass-band from 10Hz to 500Hz is satisfactory for most psycho-
      physiological recording situations
     Upper-trapezius muscle activity connected to the level of activation, and
      attention-related activity was observed over this muscle as well (frontalis and
      upper trapezius)
     (from “standardizing surface EMG recordings for assessment of activity and
      fatigue in the human upper trapezius muscle): inter-electrode distance suggested
      is 20mm.
   (from Hanah’s paper on face EMG) Used Biopac amps, signals amplified 5,000,
      digitized at 1000Hz, artifact monitored through videofeed, then 15Hz high-pass
      filter to reduce movement and blink-related artifact, then full-rectified. Data then
      visually inspected, and data with further artifact excluded from analysis. To
      correct the positive skew inherent to EMG data, all data was subjected to a
      square-root transformation. Then EMG reactivity was measured as the difference
      between activity during the 6s stimulus period and 1s prior to stimulus onset.
   (from Biopac and Hanah)
   Use 1Hz – 5000Hz and 100Hz off setting to get unfiltered data (can filter
      afterwards, using a better program i.e. Matlab). I observed higher peaks for 1Hz –
      5000Hz than 5Hz-500Hz, and peaks almost gone if 100HzHP is on.
   Use 2 shielded electrodes (black part is the shield) and one unshielded for ground
      (plug in first)
   Skin abrasion by rubbing with alcohol strips (Sterets)
   Attaching the electrode: put on double sided tape, put gel in cavity with cue tip
      (careful not to get bubbles), peel off tape cover, and place on skin.
   Check impedance before and after the experiment by plugging in the + and –
      electrodes in the Mastech, using the Ω setting. <50kΩ is a good value for face.
   When done, wash electrodes with cue tip under running warm water, and leave
      overnight before reuse.
   More tips:
   Use Christina’s startle template to test EMG and Biopac template.
   Exact placement of electrodes:
   from Recording Techniques, use the acromion and spine of 7th cervical vertebrae,
  spaced around the middle of the line connecting the two. Test by lifting the shoulder
  (head turned slightly towards the opposite side.
   From “Short communication. Surface Electrode placement and Upper Trapezius”,
  this placement actually produces activity in middle trapezius instead.

General RSP tips for experimental setup (startle & emotion):
     (sources: handbook, Ch 10)
     RSA=oscillatory influence of respiration on heart rate (noninvasive index of vagal
     Few placement guidelines of the strain gauges exist
     Provides time-dependent respiratory parameters: breathing frequency, time of
      inspiration Ti and time of expiration Te)
      (source: Andre?)
     use TWO strain gauges (one for abdomen, and one for chest)

General NIBP tips for experimental setup (startle & emotion):
     (sources: Biopac template, website, and Norm)
     Keep transducer plugged in; press test to initialize before experiment
     Plug into analog channel in UIM100C; set up channel in Biopac by testing
      (collect test-pulses) and scaling top of square wave to 100Hgmm and bottom to
     Put on wrist so that sensor is on outside of wrist, on vein (cable towards body on
      left hand), strap on tightly.
     Subject must be very still during data collection
     HR can be calculated by counting peaks
     Systolic BP calculated by finding maximum peak
     Diastolic BP calculated by finding the min first peak after max peak
     Use 200 samples/sec

General EGG tips for experimental setup (startle & emotion):
     (sources: handbook, Ch 11)
     Look for a 3cpm sinusoidal wave (activity of stomach lowest when empty. When
      food enters stomach, incl. water, incr. in amplitude occurs at 2.2-2.5cpm, then
      after 10-15min goes back up to 3cpm.) Tachigastria is 4-9cpm (related to nausea),
      bradygastria is 0-2.25cpm.
     Subjects food intake: for most experiments, they are instructed to fast for at least
      4hrs, for others they are asked to fast overnight and given a standard small
     Skin treatment: shave hair, and abrade area until pink (see next source)
     Lower amplitudes expected for negative moods like depression, and higher for
      positive moods like vigor.
     Electrodes: use Ag-AgCl
     Recording site: for single channel (what we have) greatest EGG amplitude will be
      obtained with one electrode on subject’s left side, approx. 6cm from the midline
      and just below the lowest rib, and the other on the midline, just above the
      umbilicus. (same as in “increased feelings with increased body signals” Crohn’s
      disease paper). The location affects the amplitude and waveform, but not the
     Measurement of greatest interest: frequency
     Filter out signals below 0.5cpm (to eliminate shifts in baseline due to DC
      potentials) and above 15cpm (to avoid domination of gastric signal by ECG)
     --> settings for us, for data recording: HP:0.005Hz (filters out <0.3cpm),
      LP:1Hz(filters out >60cpm), sampling rate: 200samples/sec (but can use anything
      >4), gain: 2000 - +/- 5mV. (Will apply more appropriate filters afterwards, and
      remove RSP artifacts)
     Should record respiration at the same time, and later remove respiration artifacts
      (its frequency range is approx. that of tachigastria)
     Use a sampling rate of >4samples/sec
     Use at least 4min data windows
     (sources: “comparison of multichannel EGG obtained with the use of three
      different electrode types”)
     They use a different electrode positioning (see diagram)
     Once I determine the right positions for electrodes for one subject, I should mark
      it down on a transparent foil so that I can reproduce it for repeated measurements.
     GEL: use normal EEG gel (GEL100)
     Electrode size: “Too small an electrode conductive area would bring about a
      weakening of the electrical signal received from the abdominal surface”
      (Mintchev et al, 1997)  with small electrode (0.78cm²) compared to large
      electrode (5.3178cm²) the small electrode stability of the dominant frequency was
      inferior, and the signal quality worsened with time.
      the small EMG face electrodes are 0.12cm², BUT we should use at least 2cm²
      (used in this paper, worked ok), so that is a radius of at least 0.8cm for the
      conductive surface. (NB: Found some that are 0.8cm diameter, so 0.8cm², will
      have to do for now.)
     (sources: “increased feelings with increased body signals” paper)
     They use BiopaK and AcqKnowledge, 200Hz sampling rate, electrode positioning
      like in HDBK.

General ECG tips for experimental setup (startle & emotion):
     (sources: handbook, Ch 9)
     Startle and emotional stress can induce rhythmic dysfunction, can see on ECG as
      irregularly timed, extra, or skipped beats.
     Use 3 unipolar leads on right and left arm, and left leg (Einthoven’s equilateral
      triangle principle)

     (sources:Biopac)
     Electrodes: use either
       disposable: EL500 series and LEAD110/110S (we have these in the lab)
       reusable: EL258 with adhesive disks(ADD208) and GEL100 (we have these in
      the lab)
     Should record normal, rather than R-wave mode (just shows a peak when an R-
      wave is detected, useful for calculating beats per minute BPM and inter beat
      interval IBI). Good diagram of expected waveform in the hardware guide, page 84.
     Should sample at 250samples/sec or faster (for R-wave detection)
     Leave HP switch off
     Use 3 leads (shielded Vin- on right arm, shielded Vin+ on left arm, and
      unshielded GND on right leg). Diagram on page 85 of hardware guide. Note:
      handbook specified left leg. Will use left leg.
     There exists an AcqKnowledge template; I will be using that to test out ECG

General PPG tips for experimental setup (startle & emotion):
   (sources: Biopac)
   Place on finger tips (Christie paper confirms)  best to place on ring finger of left
    hand if we’re doing GSR at the same time
   Peak measurement recorded indicates point of maximal blood density; can get
    indication of blood pressure if compared to the point of R-wave onset in the ECG
   Use Biopac template (PPG + GSR + RSP) for settings  HP =0.5Hz, gain = 20
    (1000mV peak to peak), HP2 = DC, sampling rate=200samples/sec.
   No gel for TSD200; just use strap to attach to finger
   Note: readings vary considerably depending on location on body and tension in
    the strap of the transducer.
   Leads: red-+Vsup; black-GND; blue/purple-input

Shared By: