Do en icula epola iza ion in e al a ios
depend on hea a e and should hey be a e-co ec ed?
Ca los Al a ado-Se ano1,2, Juan Ramos-Cas o2, and Ramon Pallàs-A eny2
1Depa men o Elec ical Enginee ing, CINVESTAV-IPN, México D.F., México
2Depa men o Elec onic Enginee ing, Technical Uni e si y o Ca alonia (UPC), Ba celona, Spain
Abs ac –QT in e al dec eases wi h inc easing hea
a e (HR), hence o de ine no mal alues o QT in e al o
di e en hea a es equi es he applica ion o a QT in e al
a e co ec ion o mula. Howe e , i ’s unknown he
in luence o HR on he a ios be ween se e al en icula
epola iza ion in e als, ha can be use ul as isk p edic o s
o malignan en icula a hy hmias. This s udy analyzes
he in luence o HR on QTp/QT, JTp/JT, Tpe/JTp, and
Tpe/JT a ios on a heal hy subjec du ing exe cise. The
in e als we e measu ed in V3 lead and he cha ac e is ic
poin s o QRS complex and T wa e we e de ec ed by
compu e algo i hms based on he wa ele ans o m (WT).
Exe cise s ess was g aded acco ding o he ou ini ial s ages
o he B uce p o ocol. The hea a e was di ided in o wo
anges: less han 100 bea s/min and om 100 bea s/min o
185 bea s/min. When HR exceeded 100 bea s/min, QTp/QT
and JTp/JT a ios inc eased whe eas Tpe/JTp and Tpe/JT
a ios dec eased o inc easing HR. Howe e , when HR was
below 100 bea s/min, hose same ou a ios we e
independen om HR. As opposed o he QT in e al, he e
is no need o co ec QTp/QT, JTp/JT, Tpe/JTp, and Tpe/JT
a ios when HR is below 100 bea s/min.
Key wo ds––QT in e al, exe cise ECG, hea a e
I. INTRODUCTION
QT p olonga ion in he 12-lead ECG is associa ed wi h a
high isk o en icula a hy hmias and sudden ca diac
dea h in a ious clinical condi ions, o example in pos -
myoca dial in a c ion (MI) pa ien s, howe e , i s
p edic i e alue is con o e sial [1,2]. The disc epancies
abou he p ognos ic alue can be a ibu ed o ha he QT
in e al has many sou ces o a ia ion [3]. One o hem a e
he measu emen me hods ha use di e en de ini ions o T
wa e end and di e en lead. O he sou ces o a ia ion a e
due o physiological ac o s, such as elec oly e diso de s,
d ugs, au onomous one, diseases, and, mainly, hea a e
(HR) [3,4]. This HR dependence on QT du a ion was
ea ly well es ablished and, consequen ly, o de ine no mal
anges o QT in e al, i s ac ual du a ion is co ec ed
acco ding o HR [5].
Figu e 1 shows o he en icula epola iza ion in e als
ha a e use ul in p edic ing malignan en icula
a hy hmias. P e ious epo s on HR dependence on hese
in e als in no mal subjec s ha e indica ed ha a es , JT,
QTpeak (QTp) and JTpeak (JTp) depend on HR bu
Tpeak-Tend (Tpe) does no [6-9]. Du ing exe cise, QTp
and JTp depend on HR and Tpe has a weak bu
s a is ically signi ican dependence on HR [6,8,10]. The e
a e no known epo s abou he dependence o HR on JT.
Fig. 1. Time in e als o en icula epola iza ion ac i i y.
Dynamics o he a ios be ween some en icula
epola iza ion in e als du ing exe cise in no mal subjec s
has also ecei ed some a en ion. I has shown ha
Tpe/QT inc eases [11] and QTp/QT dec eases [12] a
ele a ed HR, bu dynamics o JTp/JT, Tpe/JTp, and
Tpe/JT does no seem o ha e been s udied. The aim o
his wo k is o analyze he in luence o HR on QTp/QT,
JTp/JT, Tpe/JTp, and Tpe/JT du ing exe cise. These a ios
we e calcula ed o he V3 lead eco ded on a heal hy
subjec du ing he i s ou s ages o he B uce p o ocol,
in wo anges: HR below 100 bea s/min and HR om 100
bea s/min o 185 bea s/min. The cha ac e is ic poin s o
QRS and T wa es we e de ec ed by compu e algo i hms
based on he wa ele ans o m (WT).
II. METHODOLOGY
Subjec and da a acquisi ion
The subjec was a 45-yea old heal hy male. Exe cise es s
we e pe o med on a eadmill (Technogym Run Race
HC1200) ha can be p og ammed o ollow di e en
medical p o ocols and includes a wi eless hea a e me e
which we used o assess ha HR did no exceed 185
bea s/min. The eadmill was p og ammed acco ding o
he B uce p o ocol [13], which includes se en g aded
s eps las ing h ee minu es each and di e ing by an
inc eased loo ele a ion and speed. Signals we e acqui ed
by using he modi ied 12-lead ECG p oposed by Mason
and Lika [14] o exe cise es ing. The wel e
simul aneous signals we e ampli ied (p og ammable gain
QTp
JTp
Tpe
QRS JT
Q
T
0-7803-7789-3/03/$17.00 ©2003 IEEE 59 EMBC 2003
500–2500, bandwid h om 0.025 Hz o 500 Hz) by a
p op ie a y 16-channel high- esolu ion sys em [15]
connec ed o a digi izing plug-in PC ca d (Da a
T ansla ion DT21EZ, 12 b esolu ion) ha sampled hem
a 1 kHz a e. The analyzed signals we e 30 minu es
eco dings om he V3 lead co esponding o he i s ou
s ages o he B uce p o ocol.
Da a p ocessing and analysis
Al a ado [16] desc ibes he algo i hms used o de ec
cha ac e is ics poin s o he QRS complex and he T wa e
which a e based on WT. The alida ion p ocess o he
h ee designed algo i hms has been pe o med by using 25
eco ds o he well-anno a ed e e ence da abase o ECG
measu emen , he CSE mul i-lead measu emen da abase
[17], in which he golden s anda d had been de i ed by an
in e na ional g oup o ca diologis s who isually
de e mined he on- and –o se poin s o P, QRS and T
wa es [18]. In hose 15-lead eco ds, he ea lies onse
(QRS) and he la es o se s (QRS and T wa e) o each
eco d we e de ec ed, and he measu emen s we e wi hin
he ole ance limi s o de ia ions wi h espec o he
manual measu emen s de e mined by he CSE expe s
[19].
When he T and P wa es me ged (app oxima ely a hea
a es abo e 150 bea s/min), he T-wa e end was es ima ed
as desc ibed by Al a ado e al. [16]. The ECG in e als
measu ed om he V3 lead we e: QT, QTp, JT, JTp, Tpe,
and QRS (Figu e 1). Bea s displaying a oo small T-wa e
ampli ude, o ha we e se e ely co up ed by mo emen
a i ac s, we e excluded. The dependence be ween he
en icula epola iza ion in e al a ios conside ed and
HR we e analyzed by linea eg ession using he leas
squa es me hod o wo anges o HR: below 100
bea s/min and om 100 o 185 bea s/min.
III. RESULTS AND DISCUSSION
Figu es 2(a) and 2 (b) espec i ely show QTp/QT and
JTp/JT du ing exe cise. Table 1 sums up he esul s o he
eg ession analysis o hese a ios e sus HR o he wo
de ined HR anges.
(a)
(a)
(b)
Fig. 2. Dynamics o (a) QTp/QT and (b) JTp/JT du ing
exe cise es ing on a heal hy subjec .
Table 1 Rela ionship be ween QTp/QT, JTp/JT and HR
(m = slope, = co ela ion coe icien )
HR < 100 bea s/min HR > 100 bea s/min
m m R
QTp/QT 0 0.01 0.001 0.77
JTp/JT 0 – 0.07 0.001 0.71
When HR < 100 bea s/min, nei he QTp/QT no JTp/JT
depend on HR. When HR > 100 bea s/min, bo h QTp/QT
and JTp/JT inc ease wi h HR, and he co ela ion is high.
The inc ease o QTp/QT wi h HR disag ees wi h a
p e ious epo [12] desc ibing a dec ease o QTp/QT o
ele a ed HR, bu in ha s udy [12] me hod measu emen
o end o T wa e and HR ange measu ed is unknown,
he e o e i is no possible compa e bo h s udies in
ele a ed HR.
Figu es 3(a) and 3(b) espec i ely show Tpe/JT and
Tpe/JTp du ing exe cise. Table 2 sums up he
co esponding esul s o he eg ession analysis. When HR
< 100 bea s/min, nei he Tpe/JT no Tpe/JTp depend on
HR. When HR > 100 bea s/min, bo h Tpe/JT and Tpe/JTp
dec ease wi h HR, and he co ela ion is high.
(a)
60 80 100 120 140 160 180 200
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
QTp/QT
Hea a e
(
bea s/min
)
60 80 100 120 140 160 180 200
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
JTp/JT
Hea a e (bea s/min)
60 80 100 120 140 160 180 200
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
Tpe/JT
Hea a e
(
bea s/min
)
60
(b)
Fig. 3. Dynamics o (a) Tpe/JT and (b) Tpe/JTp du ing
exe cise es ing on a heal hy subjec .
Table 2. Rela ionship be ween Tpe/JT, Tpe/JTp and HR
(m = slope, = co ela ion coe icien )
HR < 100 bea s/min HR > 100 bea s/min
m m
Tpe/JT 0 0.07 –0.001 –0.71
Tpe/JTp 0 0.07 –0.002 –0.72
The e o e, as opposed o he QT in e al, he e is no need
o co ec QTp/QT, JTp/JT, Tpe/JTp, and Tpe/JT when
HR is below 100 bea s/min. This independence may
inc ease he impo ance o hose a ios as possible
p edic o s o malignan en icula a hy hmias and
sudden ca diac dea h. Fu u e wo k will ocus on
co obo a e hese indings in a la ge popula ion o pos -
MI pa ien s wi h and wi hou suscep ibili y o en icula
achya hy hmias.
ACKNOWLEDGEMENT
This wo k has been unded by a schola ship om he
CONACYT (México) o Ca los Al a ado and by p ojec
SAF98-0120 om he Spanish CYCIT. The au ho s a e
g a e ul o J. Amo ós o hei help in using he eadmill,
and o he olun ee s, pa icula ly B. Pé ez and M. Va gas.
REFERENCES
[1] B. Su awicz, S.B. Knoebel, “Long QT: good, bad o
indi e en ?,” J. Am. Coll. Ca diol., ol. 4, pp. 398–413, 1984.
[2] P. Da ey, “QT in e al and mo ali y om co ona y a e y
disease,” P og. Ca dio asc. Dis., ol. 42, pp. 359–384, 2000.
[3] C. Funck-B en ano, P. Jaillon, “Ra e-co ec ed QT
in e al: echniques and limi a ions,” Am. J. Ca diol.,
ol. 72, no. 6, pp 17B–22B, 1993.
[4] A.J. Moss, “Measu emen o he QT in e al and he isk
associa ed wi h QTc in e al p olonga ion: a e iew,”
Am. J. Ca diol., ol. 72, no. 6, pp. 23B–25B, 1993.
[5] H.C. Baze , “An analysis o he ime- ela ions o
elec oca diog ams,” Hea , ol. 7, pp. 353–370, 1920.
[6] J. O’Donnell, S.B. Knoebel, D.E. Lo elace, P.L. McHen y,
“Compu e quan i a ion o Q-T and e minal T wa e (aT-eT)
in e als du ing exe cise: me hodology and esul s in no mal
men,” Am. J. Ca diol., ol. 47, pp. 1168–1172, 1981.
[7] M. Me i, J. Benho in, M. Albe i, E. Loca i, A.J. Moss,
“Elec oca diog aphic quan i a ion o en icula epola iza ion,”
Ci cula ion, ol. 80, pp. 1301–1308, 1989.
[8] K.G. Lax, P.M. Okin, P. Klig ield, “Elec oca diog aphic
epola iza ion measu emen s a es and du ing exe cise in
no mal subjec s and in pa ien s wi h co ona y a e y disease,”
Am. Hea J. ol. 128, pp. 271–280, 1994.
[9] I. Sa elie a, S.B. Reddy, A.J. Camm, M. Malik, “Does
dispe sion o epola iza ion depend on ca diac cycle leng h and
should i be a e-co ec ed? Obse a ions in 1906 heal hy
subjec s (abs ac ),” J. Am. Coll. Ca diol., ol. 35, no. 2, Suppl.
A, pp. 143, 2000.
[10] P.P. Da ey, “QT in e al measu emen : Q o Tapex o Q o
Tend,” J. In . Med., 246, pp. 145–149, 1999.
[11] J. O'Donnell, D.E. Lo elace, S.B. Knoebel, P.L. McHen y,
“Beha io o he e minal T wa e du ing exe cise in no mal
subjec s, pa ien s wi h symp oma ic co ona y a e y disease and
appa en ly heal hy subjec s wi h abno mal ST segmen
dep ession,” J. Am. Coll. Ca diol., ol. 5, pp. 78–84, 1985.
[12] L. Zhang, G.M. Vincen , “Sympa he ic modula ion a ec s
epola iza ion dispa i y in LQTS and no mals: QTp/QTe
changes du ing exe cise and B-blockade (abs ac ),” J. Am. Coll.
Ca diol. (Feb. Special Issue), pp. 37A, 1994.
[13] R.A. B uce, F. Kusumi, D. Hosme , “Maximal oxygen in ake
and nomog aphic assessmen o unc ional ae obic impai men
in ca dio ascula disease,” Am. Hea J., ol. 85, pp. 546–562,
1973.
[14] R.E. Mason, I. Lika , “A new sys em o mul iple-lead exe cise
elec oca diog aphy,” Am. Hea J., ol. 71, no. 2, pp. 196–205,
1966.
[15] J. Ramos, “De ección de mic opo enciales au icula es de al a
ecuencia,” Ph.D. disse a ion, Technical Uni e si y o
Ca alonia, Ba celona, Spain, 1997.
[16] C. Al a ado, “Análisis de la a iabilidad de los in e alos de
iempo del ECG,” Ph.D. disse a ion, Technical Uni e si y o
Ca alonia, Ba celona, Spain, 2001.
[17] Willems J.L., P. A naud, J.H. Van Bemmel, R. Degani, P.W.
Mac a lane, C. Zywie z o he CSE Wo king Pa y, “Common
s anda ds o quan i a i e elec oca diog aphy: goals and main
esul s,” Me h. In o m. Med., ol 29, no. 263, 1990.
[18] J.L. Willems: Common s anda ds o quan i a i e
elec oca diog aphy. CSE mul ilead a las, Measu emen esul s-
da a se 3. CSE P ojec . ACCO Publ., Leu en, Belgium, 1988.
[19] The CSE wo king pa y, “Recommenda ions o measu emen
s anda ds in quan i a i e elec oca diog aphy,” Eu . Hea J.,
ol. 6, pp. 815–825, 1985.
60 80 100 120 140 160 180 200
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Tpe/JTp
Hea a e (bea s/min)
61