VTl notat 38A-2000
Heart rate measures as
drowsiness indicators
c
c
o
N
<
no
no
H G H O =Author
Jan Tornros, Bjorn Peters &
Joakim éstlund
Research division Traffic and Road User Behaviour
Project number
40353
Project name
Connections between heart rate
measures and EEG
Sponsor
Autoliv Research
Distribution
Free
Swedish National Road and
'Transport Research Institute
Foreword
The data analysed in the present report was generously provided by Dr Alain
Muzet, CNRS CEPA, Strasbourg, France.
AutoliV Research commissioned the study. Linkoping i augusti 2000
Jan Tomms
Contents
Page
1 Introduction 1
2 Method 1
3 Results 3
3.1 Connections between heart rate measures and EEG 3
3.1.1 Mean Inter-Beat Interval 3
3.1.2 Standard Deviation of Inter-Beat Interval 4
3.1.3 Standard deviation of Beat-to-Beat Interval 4
3.2 Connections between heart rate measures
and driving time 5
3.2.1 Mean Inter-Beat Interval 5
3.2.2 Standard Deviation of Inter-Beat Interval 6
3.2.3 Standard deviation of Beat-to-Beat Interval 6
4 Discussion 7
5 References 9
Appendices:
Appendix 1 EEG and driving time
Appendix 2 Mean Inter Beat Interval and EEG
Appendix 3 Standard Deviation of Inter-Beat Interval and EEG Appendix 4 Standard deviation of Beat-to Beat Interval and EEG Appendix 5 Mean Inter Beat Interval and driving time
Appendix 6 Standard Deviation of Inter Beat Interval and driving time Appendix 7 Standard deviation of Beat-to-Beat Interval and driving time
1
Introduction
Various attempts have beenmade by different researchers to find connections between heart rate measures and drowsiness. This issue has been studied in pro longed monotonous test sessions, such as laboratory vigilance tasks or long-distance driving.
As summarised by Bohman (1997) heart rate (HR) defined as number of
beats per time unit has in various studies been found to decrease gradually with distance or hours driven (Lisper et.al., 1973; Fagerstrom & Lisper, 1977). Riersmaa et.al. (1977) who used another related measure mean inter-beat interval also found a connection with driving time; this measure increased as a
function of driving time. Egelund (1982), however, did not find any connection
between HR and distance driven in highway driving. Torsvall & Akerstedt (1987) found that HR decreased during driving, whereas no connections were found between heart rate and EEG measures of drowsiness. Since HR is very sensitive to effects of physical activity and stress it has not been considered a very suitable drowsiness indicator (Egelund, 1982).
A different measure, heart rate variability defined either as the standard deviation of number of beats per time unit or as the standard deviation of inter beat interval - seems to hold more promise as a measure of drowsiness. Heart rate variability has been found to increase gradually during a vigilance task and in
prolonged driving (O Hanlon & Beatty, 1977; Mackie & O Hanlon, 1977), even though Egelund (1982) failed to find a connection between this measure and dri
ving distance. Apart from the overall variability, another variability index has also
been used, beat to beat variability. Bader (1995) claims to have found a
correla-tion between this measure and degree of wakefulness.
The aim of the present study was to investigate if heart rate measures could be used as indicators to detect drowsiness while driving. To this end, connections
were analysed between such measures and drowsiness, measured via EEG waves.
For comparative reasons connections between heart rate measures and driving time were also analysed.
2
Method
Available heart rate data and EEG data for eight partially sleep deprived subjects were analysed. The data had been collected during a test session in a driving simulator, measured continuously during a two hours test drive (see Muzet et.al., 1998, for further details). The driving task was a monotonous drive on an empty freeway and subjects were instructed to comply with the speed limits and to stop at any time if needed.
The EEG measure of drowsiness used is based on the established fact that dozing off during a task is associated with increased alpha and theta power
density (Torsvall & Akerstedt, 1987). The number of seconds per 10 seconds
epoch, where either alpha and/or theta waves were present on at least two of the
three EEG recordings (F3, C3, 01), was calculated as a measure of drowsiness.
Values consequently varied from 0 (no EEG signs of sleepiness) to 10 (continuous alpha and/or theta rhythms). So the following EEG measure was used:
E = number of seconds in the j:th ten seconds epoch where alpha/theta waves are present on at least two of the three analysed EEG recordings
For heart rate, three different measures were calculated (le = length of izth inter
beat interval in the jzth epoch):
a. mean inter-beat interval for each ten seconds epoch
I l
ZR.
UIBIMEANj =
b. standard deviation of inter beat interval for each ten seconds epoch n Ian E2
IBISDJ- = g n )
c. standard deviation of beat-to-beat interval for each ten seconds epoch:
2
11 1(Ri' Ri+ ')
331513,. =
J
n l1
Intra individual correlation coefficients were calculated. For each subject
Spearman rank-order correlation coefficients (rs) were calculated between each of the three heart rate measures and the EEG drowsiness score (Guilford & Fruchter,
1973). In a similar fashion, product-moment correlation coefficients (r) were
calculated for the heart rate measures in relation to driving time.
3
Results
Table 1 shows the maximal EEG drowsiness score and the percentage of driving time where EEG drowsiness signs appeared (EEG score > 0) for each of the eight subjects. Two of them exhibited pronounced drowsiness signs whereas three subjects showed very few drowsiness signs with the remaining three subjects scoring in between.
Appendix 1 shows the individual EEG signs of drowsiness as a function of driving time. The Spearman rank order correlation coefficient (rs) was significant (p<.05) for four subjects. The strongest individual connection was rS=.41.
Table 1 Maximal drowsiness and percentage of driving time with EEG
drowsi-ness Signs
Subject Maximal Percentage of driving EEG score time with EEG score > 0
Subject 1 5 3.9 Subject 2 2 0.7 Subject 3 3 4.9 Subject 4 1 0.3 Subject 5 2 0.8 Subject 6 7 30.2 Subject 7 5 9.2 Subject 8 9 29.6
3.1
Connections between heart rate measures and EEG
3.1.1 Mean Inter-Beat Interval
Table 2 Correlation between Mean Inter-Beat Interval and EEG score Subject Correlation coef cient (rs) Significance level
Subject 1 .060 .110 Subject 2 .095 .011 Subject 3 .105 .005 Subject 4 -.033 .381 Subject 5 -.001 .972 Subject 6 .1 12 .003 Subject 7 .040 .280 Subject 8 .243 .000 Average .05 .226
Table 2 shows the correlation between EEG drowsiness scores and mean inter
beat interval (IBIMEAN). There was a significant correlation for 4 of the 8 sub jects, three positive and one negative. The correlations were, however, with the
exception for the person who showed more drowsiness signs than the remaining subjects (r=.24) very low, below r =. 12.
Appendix 2 shows the individual results for this comparison.
For analysing the average intra-individual correlation coefficient, individual
correlation coefficients were transformed into Fisher 2 scores (Guilford &
Fruchter, 1973). Average 2 was calculated across subjects and the deviation from
zero was tested (t test). The result was t7 = 1.33; p =.226. That is, there was no
ge-neral connection between mean inter beat interval and EEG scores. Average 2 was re-transformed to yield an average intra subject correlation coefficient (rS = .05).
3.1.2 Standard Deviation of Inter-Beat Interval
Table 3 Correlation between Standard Deviation ofInter-Beat Interval and EEG score
Subject Correlation coefficient (rs) Significance level
Subject 1 .034 .359 Subject 2 -.013 .727 Subject 3 .038 .309 Subject 4 .001 .983 Subject 5 .031 .411 Subject 6 .248 .000 Subject 7 .092 .013 Subject 8 .007 .848 Average .05 .1 1 1
Table 3 shows the correlation between EEG drowsiness scores and standard deviation of inter beat interval (IBISD). There was a significant correlation for 2 of the 8 subjects. They were both positive.
Appendix 3 shows the individual results for this comparison.
For analysing the average intra individual correlation coefficient, individual correlation coefficients were transformed into Fisher 2 scores. Average 2 was calculated across subjects and the deviation from zero was tested. The result was
t7: 1.82; p=.111. That is, there was no general connection between standard
deviation of inter beat interval and EEG scores. Average 2 was re transformed to yield an average intra subject correlation coefficient (rS = .05).
3.1.3 Standard deviation of Beat-to-Beat Interval
Table 4 Correlation between Standard Deviation of Beat-to-Beat Interval and
EEG score
Subject Correlation coefficient (rs) Significance level
Subject 1 .061 .103 Subject 2 .044 .241 Subject 3 .046 .220 Subject 4 -.024 .517 Subject 5 .064 .089 Subject 6 .091 .015 Subject 7 .094 .012 Subject 8 .002 .968 Average .03 .145
Table 4 shows the correlation between EEG drowsiness scores and standard deviation of beat to-beat interval (BBISD). Again, there was a significant correla tion for 2 of the 8 subjects (the same two subjects who showed a significant corre lation for the other variability measure presented in table 3).
Appendix 4 shows the individual results for this comparison.
For analysing the average intra individual correlation coefficient, individual correlation coefficients were transformed into Fisher 2 scores. Average 2 was calculated across subjects and the deviation from zero was tested. The result was
t7 = 1.64; p=.l45. That is, there was no general connection between standard
deviation of beat to beat interval and EEG scores. Average 2 was re transformed to yield an average intra subject correlation coefficient (rS = .03).
3.2
Connections between heart rate measures and
driving time
3.2.1 Mean Inter-Beat Interval
Table 5 Correlation between Mean Inter-Beat Interval and driving time
Subject Correlation coefficient (r) Significance level
Subject 1 .288 .000 Subject 2 .542 .000 Subject 3 .276 .000 Subject 4 .097 .009 Subject 5 .200 .000 Subject 6 .090 .016 Subject 7 .523 .000 Subject 8 .169 .000 Average .24 .034
Table 5 shows the correlation between driving time and mean inter-beat interval (IBIMEAN). There was a significant correlation for all 8 subjects, six positive and two negative.
Appendix 5 shows the individual results for this comparison.
For analysing the average intra individual correlation coefficient, individual correlation coefficients were transformed into Fisher 2 scores. Average 2 was calculated across subjects and the deviation from zero was tested. The result was
t7 = 2.62; p=.034. That is, there was a connection between mean inter beat interval
and driving time. Average 2 was re transformed to yield an average intra subject
correlation coefficient (r = .24).
3.2.2 Standard Deviation of Inter-Beat Interval
Table 6 Correlation between Standard Deviation of Inter-Beat Interval and driving time
Subject Correlation coefficient (r) Significance level
Subject 1 .186 .000 Subject 2 .274 .000 Subject 3 -.002 .950 Subject 4 .209 .000 Subject 5 .108 .004 Subject 6 .388 .000 Subject 7 .287 .000 Subject 8 .029 .443 Average .19 .006
Table 6 shows the correlation between driving time and standard deviation of
inter beat interval (IBISD). There was a significant correlation for 6 of the 8
subjects. They were all positive.
Appendix 6 shows the individual results for this comparison.
For analysing the average intra individual correlation coefficient, individual correlation coefficients were transformed into Fisher 2 scores. Average 2 was calculated across subjects and the deviation from zero was tested. The result was t7: 3.84; p=.006. That is, there was a connection between standard deviation of inter beat interval and driving time. Average 2 was re transformed to yield an average intra subj ect correlation coefficient (r = .19).
3.2.3 Standard deviation of Beat-to-Beat Interval
Table 7 Correlation between Standard Deviation of Beat-to-Beat Interval and driving time
Subject Correlation coefficient (r) Significance level
Subject 1 .064 .088 Subject 2 .440 .000 Subject 3 .042 .256 Subject 4 .182 .000 Subject 5 .029 .432 Subject 6 .070 .061 Subject 7 .323 .000 Subject 8 -.081 .030 Average . 14 .067
Table 7 shows the correlation between driving time and standard deviation of beat to beat interval (BBISD). There was a significant correlation for 4 of the 8 subjects, three positive and one negative.
Appendix 7 shows the individual results for this comparison.
For analysing the average intra individual correlation coefficient, individual correlation coefficients were transformed into Fisher 2 scores. Average 2 was
culated across subjects and the deviation from zero was tested. The result was t7 = 2.17; p=.067. That is, there was no general connection between standard deviation of beat to-beat interval and EEG. Average 2 was re-transformed to yield an average intra subject correlation coefficient (r = .14).
4
Discussion
There was no general connection between any of the three heart rate measures investigated and EEG signs of drowsiness.
For some subjects there were, however, connections. For mean inter beat inter
val there was a comparatively strong (although quite weak) positive connection for the person with the most pronounced drowsiness signs; inter-beat intervals became longer heart rate slowed down with increasing drowsiness. For three other subjects much weaker connections were found; two were positive and one negative, which of course is not very promising for this measure.
The connections were even weaker for the two variability measures; only two subjects showed a connection between EEG scores and the variability measures. They were positive but weak; variability increased with increasing drowsiness. For the person with the most pronounced drowsiness signs there was no connec-tion at all.
So in the present study the connection between EEG scores and the selected heart rate measures was larger for mean heart rate than for any of the two varia bility measures, even though it was very weak also in that case.
It is somewhat uncertain how drowsy the subjects actually were. Drowsiness certainly varied considerably between subjects some of them showed very few signs of drowsiness but none of them fell asleep (Muzet; personal communica-tion). It is quite possible that stronger connections might have been found had the subjects been drowsier.
The results were rather different for connections between heart rate measures and driving time instead of EEG scores. At the group level the connection was positive for mean inter beat interval and for standard deviation of beat
inter-val, thus supporting earlier findings (Riersmaa et.al., 1977; Lisper et.al., 1973; Fagerstrom & Lisper, 1977; O Hanlon & Beatty, 1977; Mackie & O Hanlon,
1977), whereas there was no connection at the group level for standard deviation of beat-to-beat interval.
All eight subjects showed a connection between driving time and mean inter-beat interval, although two of them had a negative correlation instead of positive. Six and four subjects showed a connection for the two variability measures res-pectively; for inter beat interval it was positive for all six subjects, whereas for beat-to-beat interval it was positive for three of the four subjects.
So the outcome was more favourable when heart rate measures were related to
driving time than to EEG measures. However, the low or non existent connections
with EEG drowsiness signs give very little support to the hypothesis that there would be a relationship of notable size between the heart rate measures studied
here and drowsiness, a result also found by Torsvall & Akerstedt (1987) for mean
heart rate.
It might be stressed that the present analyses were performed on data collected under quite favourable circumstances. The test situation was very well controlled, and the recording of the EEG and heart rate data were very precise. In less
controlled situations one might expect to find even lower connections between the measures analysed in the present report.
The results of the present analyses consequently do not seem promising for the heart rate measures investigated as drowsiness indicators. However, no safe conclusions regarding this issue can be made. It would be very advisable to try to seek more definitive answers by performing a study where the heart rate and EEG of a number of subjects were recorded while driving until and well beyond - the point of falling asleep. This point would preferably be defined by a combination of observer ratings and driving behaviour measures which may very well be the most reliable means to objectively define these critical circumstances. In this way it would be possible to evaluate heart rate measures with respect to their capacity to detect the occurrence of serious drowsiness. Especially relevant from a traffic safety point of view is that by performing such a study one would be able to evaluate heart rate measures as predictors of the occurrence of dangerous loss of control by the driver caused by drowsiness or sleep.
5
References
Bader, G: Method and apparatus for monitoring and estimating the
awake-ness of a person. International Patent, Patent Number WO 95/33403, 1995. Bohman, K: Drowsy driver a literature survey. Autoliv research, Vargarda,
Sweden, 1997.
Egelund, N: Spectral analysis of heart rate variability as an indicator of
driver fatigue. Ergonomics, 25, 7, pp 663-672, 1982
Fagerstrom, K-O & Lisper, H-O: Effects of listening to car radio, experience,
and personality of the driver on subsidiary reaction time and heart rate in a long term driving task. In Mackie RR (Ed.) Vigilance, New York: Plenum Press, pp 73 85, 1977.
Guilford, JP & Fruchter, B: Fundamental statistics in psychology and
educa-tion. Fifth edieduca-tion. McGraw Hill, London, 1973.
Lisper, H-O & Laurel], H & Stening, G: Effects of experience of the driver on heart rate, respiration rate, and subsidiary reaction time in a three hours continuous driving task. Ergonomics, 16, pp 501-506, 1973.
Mackie, RR & O Hanlon, IF: A study of the combined effects of extended dri-ving and heat stress on driver arousal and performance. In Mackie RR
(Ed.) Vigilance, New York: Plenum Press, pp 537 558, 1977.
Muzet, A & Muzet, V & Roge, J & Thierry, P: Does partial sleep deprivation have an effect on low vigilance state occurring while driving in early afternoon? Paper presented at the 9th World Congress of Psychophysiology,
Toarmina, September 14 18, 1998.
O Hanlon, JF & Beatty, J: Concurrence of electroencephalographic and per-formance changes during a simultaneous radar watch and some implica-tions for the arousal theory of vigilance. In Mackie RR (Ed.) Vigilance, New York: Plenum Press, pp 189 202, 1977.
Riersmaa, JBJ & Sanders, AF & Wildervanck, C & Gaillard, AW: Performance
decrement during prolonged night driving. In Mackie RR (Ed.) Vigilance,
New York: Plenum Press, pp 189 202, 1977.
Torsvall, L & Akerstedt, T: Sleepiness on the job: Continuously measured EEG changes in train drivers. Electroencephalography and Clinical Neuro physiology, 66, pp 502 511,1987.
Appendix 1 Page 1 (4)
EEG and driving time
Relationship between EEG scores (0-10) and driving time in seconds
Subject 1
rS=.162 p=.OOO
1O 9 8... E EG 1- {I I?! K?! E! 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 2 rs=.059 p=.112 10 9... 8.. E E G o 800 1600 2400 32'00 4o'oo 48'00 5600 64b0 72'00 8000 TIME VTl notat 38A-2000Appendix 1 Page 2 (4) Subject 3
rs=.138 p=.OOO
E E G 3- £3 2- :3 m m 5: m 1- Willi 3 3 WEEK} Em m 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 4rS=.OZ4 p=.524
E E G 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMEAppendix 2 Page 3 (4) Subject 5 rS= .O69 p=.065 E E G 1 {if} n m 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 6 rs=.408 p=.000 5- m 53:: mt: mamas EE G 4 - {3 E531 EZ : 33 3mm i3 DRE 3- {3 53W mm mm 2" {3 £31?! GQWHWWUWBWWWHW £223 1- F3 Ema mm Etiri tié RE E3 33 WEED
0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000
TME
Appendix 1 Page 4 (4) Subject 7 (92.058 p=.117 E E G 2- 13 am 22:: :2 2322mm
1- mum mmwuma mm: [mm {imam
O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TIME Subject 8
rs=.083 p=.026
1O E E G 3- mm 3:: 2" nmmmamm mmm mnnnammmmm 1- wnummm W {SW W E! {13$ {1333!} W O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMEAppendix 2 Page 1 (4)
Mean Inter-Beat Interval and EEG
Relationship between mean inter-beat interval in milliseconds and EEG
scores (0-10) Subject 1 rs=.060 p=.110 1200 1100- 1000- 900-$2 m m W W W M i {I a 800- 700-IB I M E A N EEG Subject 2
rS=.095 p=.011
1200 1100' 1000 900' 800- 700- 600-IE I M E A NAppendix 2 Page 2 (4) Subject 3 rs=. 105 p=.005 1200 1100' 1000 900' 800- 700- 600-I B I M E AN 500- 400- 300- 200- 100-0 -L0 Q0 L0 20 30 40 30 a0 10 &0 30 10K) EEG Subject 4 rS= .033 p=.381 1200 1100 1000 900' 800- 700- 600-1B I M EA N 500- 400- 300- 200-
Appendix 2 Page 3 (4) Subject 5
rS= .001 p=.972
1200 1100 1000' 900 800- 700- 600-IB I M E AN 500- 400- 300- 200- 100-0 Subject 6rS= .112 p=.003
1200 1100' 1000 900'{V I E W {3 m : (m
800- 700- 600-1B I M E A N 500-400 300 200- 100-O
Appendix 2 Page 4 (4) Subject 7
rS=.040 p=.280
1200 1100- w 1000-900' $3 33 33 CK } EE K} {3 800- 700- 600-IE I M E A N 500-400 300- 200- 100-O -1 0 1 2 3 4 5 6 7 8 9 1O EEG Subject 8rs=.243 p=.000
1200 1100 000' 900' m m [ m m H LE {3800-
700 I B I M E A N 600' 500- 400-I 300-200~ 100-EEG
Appendix 3 Page 1 (4)
Standard Deviation of Inter-Beat Interval and EEG
Relationship between standard deviation of inter-beat interval in milli-seconds and EEG scores (0-10)
Subjectl rs=.034 p=.359 250 225 200' 175-£3 150- m 00 P t pg 125- g ) 1 . 100-75~ g (I: 50- m
iii: a g;
. 25- f a E: O l I 1 O 1 2 3 4 5 6 7 8 9 10 EEG SubjethrS=-.013 p=.727
250 225 200' 175- 150- 125-IB IS D 100- 75- 50-Appendix 3 Page 2 (4) Subject
rS=.O38 p=.309
IB IS D 250 225' 200 175- 150 -ES E! 125- 100-75- 8a 50- g a 25- 3 EEG Subject 4 rs=.001 p=.983 IB IS D 250 225' 200* i} 175 a 150- 125- 100- 75- 50-Appendix 3 Page 3 (4) Subject 5 rs=.031 p=.411 250 225 200 175- 150- 125-IB IS D 100- 75- 50-25- {£33 3 5 3 E" Subject 6
rS=.248 p=.OOO
250 225 200 175 125-IB IS D 100-75* : G W I3 {. 3 E! 50-GE ! 13 25-m a : W 5 3 5 ! a n m I} 00 A 01 o). \l CD (D 10 EEGIB IS D 150-125~ 100* 75* 50- 25-200" 175 EEG gm yzzs ms aus 1m m £3 {£3 {3 W W W W m :3 m m m m m m £3 Q O £1 El i} [X E 1O
VTI notat 38A-2000
rs rs
.092 p=.013
250 225 200 175- 150- 125- 100-754 50- 25-250 225' Subject 7 Subject 8.007 p=.848
:a=:::1 :=z= ::x= : 15 5: 3»: sz {zm zm m g m Si} EEG 10 Appendix 3 Page 4 (4)Appendix 4 Page 1 (4)
Standard deviation of Beat-to-Beat Interval and EEG
Relationship between standard deviation of beat-to-beat interval in milli-seconds and EEG scores (0-10)
Subject 1
rS= .061 p=.103
300 275 250 225' 200- 175- 150-BB IS D 125 100- 75- 50- 25-O J N W {13 133 351 5 m m C R 13 -1 <5 1' i 3 4 5 6 7 8 9 10 Subject 2rs=.044 p=.241
300 275 250' 225 200* 175- 150-BB IS D 125-100 75- 50-Subject 3
rs=.046 p=.220
BB IS D Subject 4rS=-.O24 p=.517
BB IS D Appendix 4 Page 2 (4) 300 275 250 225' 200- 175- 150- 125- 100-75 50~ 25-m m m EEG 300 275' 250' 225' 200- 175- 150- 125- 100-EEGAppendix 4 Page 3 (4) Subject 5
rs=.064 p=.089
300 275 225 200- 175- 150-BBIS D 125- 100- 75-50~ "3 Subject 6 rs=.O91 p=.015 300 275' 250 225 175- 150-BB IS D 125 100- 75- 50-a m m um m £3C H E W E3 {3 EEG
Appendix 4 Page 4 (4) Subject 7
rs=.094 p=.012
BB IS D 300 275 250' 225 200- 175- 150-EE G 125- 100-75-EEG Subject 8
rs=.002 p=.968
BB IS D 300 275' 250 CE [2 E} 225' 200-175 150- 125-U 100-If? 75- 50-35 5 m m 323 531 21HES EEG
Appendix 5
Page 1 (4)
Mean Inter-Beat Interval and driving time
Relationship between mean inter-beat interval in milliseconds and driving time in seconds Subject 1 r=.288 p=.000 1200 1100 1000 3 900 700-600 - E3 I B IME A N 5002 400- 300- 200- 100-0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 2
r=.542 p=.000
1200 1100 1000 900 800- 700- 600-I B I M E A N 500- 400- 300- 200- 100-0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME VT| notat SSA-2000Appendix 5 Page 2 (4) Subject 3
r=.276 p=.000
1200 1100' 1000' 500* I B I M E A N 400-300~ 200- 100-0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 4r= .O97 p=.009
1200 1100' 1000 900~ 800- 700- 600- 500-I B I M E A N 400- 300- 200- 100-O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMEAppendix 5 Page 3 (4) Subject 5 r=.200 p=.000 1200 1100 1000 900 800- 700- 600-IB I M E AN 500- 400- 300- 200-100 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 6
r= .090 p=.016
1200 1100' 1000 900 800- 700- 600-500' IE I M E A N 400- 300- 200-100 o 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMEAppendix 5 Page 4 (4) Subject 7
r=.523 p=.000
1200 1100' 1000' 900 800- 700- 600- 500-1B I M E A N 400- 300-200 100-0 O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 8 r=. 169 p=.000 1200 1100' 1000 900 800* 700- 600- 500-IE I M E A N 400- 300-200 100 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMEAppendix 6 Page 1 (4)
Standard Deviation of Inter-Beat Interval and driving time
Relationship between standard deviation of inter-beat interval in milli-seconds and driving time in milli-seconds
Subject 1 r=. 186 p=.000 300 280' 260' 240' 220-200 180- 160-140* IB IS D 80-O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 2
r=.274 p=.000
300 280' 260' 240 220* 200* IB IS D o 800 1600 2400 32b0 40.00 4800 5600 6400 72.00 8000 TMESubject 3 Appendix 6 Page 2 (4)
r= .002 p=.950
IB IS D Subject 4 300 280' 260 240' 220- 200- 180- 160- 140- 120- 100- 80- 60-40 20-0 800 1600 2400 3200 4000 4800 5600 6400 7200 TMEr=.209 p=.000
IB IS D 8000 300 280- 260- 240- 220-200~ 180- 160- 140- 120- 100- 80- 60- 4o- 20-0 800 1600 2400 3200 4000 4800 5600 6400 7200 TME 8000Appendix 6 Page 3 (4) Subject 5
r=.108 p=.OO4
IB IS D 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 6 r=.388 p=.000 IB IS D 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMEAppendix 6 Page 4 (4) Subject 7
r=.287 p=.000
IB IS D 0 800 1600 24.00 3200 4000 4800 5600 64.00 7200 8000 TME Subject 8r=.029 p=.443
IB IS D 300 280 260 240 220 -200 - m 180 - I :1 h m 160 - n 140 -120 ~ 100 ~ 80 so 40 20 -O 800 1600 2400 3200 4000 4800 5600 64.00 7200 8000 TMEAppendix 7 Page 1 (4)
Standard deviation of Beat-to-Beat Interval and driving time
Relationship between standard deviation of beat-to-beat interval in milli-seconds and driving time in milli-seconds
Subject 1
r=.064 p=.088
300 280' 260 240 220- 200-180* BB ISD 140-O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 2r=.440 p=.OOO
260 BB IS D 0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TMESubject 3 Appendix 7 Page 2 (4) r=.042 p=.256 BB IS D Subject 4 300 280' 260' 240' 220-200 180- 160- 140- 120- 100- 80- 60- 40- 20-800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME r=. 182 p=.000 BB IS D 300 280 260' 240' 220 200' 180' 1601 140 120 100 80- 60-401 20-0 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME VT| notat 38A-2000
Appendix 7 Page 3 (4) Subject 5
r=.029 p=.432
300 280 -260 q 240 220 200 180 160 14o 120 100 80 so 4o 20 -BB IS D O 800 1600 2400 3200 4000 4800 5600 6400 7200 8000 TME Subject 6 r=.070 p=.061 300 280 260 240' 220- 200- 180- 160- 140- 120- 100- 80-60- §1.7 40- g 20 -BB IS D -m :3 ' 3%" o 800 1600 2400 3200 4000 4800 5600 6400 72b0 8000 TMESubject 7 Appendix 7 Page 4 (4)