Sunteți pe pagina 1din 8

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325693609

Qualitative and quantitative evaluation of a new wearable device for ECG and
respiratory Holter monitoring

Article  in  International Journal of Cardiology · June 2018


DOI: 10.1016/j.ijcard.2018.06.044

CITATION READS

1 221

12 authors, including:

Antonio Sarmento Stefania Paolillo


Universidade Federal do Rio Grande do Norte University of Naples Federico II
26 PUBLICATIONS   18 CITATIONS    105 PUBLICATIONS   1,474 CITATIONS   

SEE PROFILE SEE PROFILE

Carolina Lombardi Massimo Mapelli


I.R.C.C.S. Istituto Auxologico Italiano Centro Cardiologico Monzino
157 PUBLICATIONS   2,329 CITATIONS    49 PUBLICATIONS   81 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Chest Sonography View project

Sleep and stroke View project

All content following this page was uploaded by Antonio Sarmento on 01 September 2018.

The user has requested enhancement of the downloaded file.


IJCA-26593; No of Pages 7
International Journal of Cardiology xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Cardiology

journal homepage: www.elsevier.com/locate/ijcard

Qualitative and quantitative evaluation of a new wearable device for ECG


and respiratory Holter monitoring
Antonio Sarmento a,1, Carlo Vignati b,c,1, Stefania Paolillo d, Carolina Lombardi e, Alessandra Scoccia b,
Flavia Nicoli b, Massimo Mapelli b, Alessandra Leonardi f, Dario Ossola f, Rudy Rigoni a,f,
Piergiuseppe Agostoni b,c,⁎, Andrea Aliverti a
a
Dipartimento di Elettronica, Informazione e Bioingegneria, Politecnico di Milano, Milano, Italy
b
Centro Cardiologico Monzino, IRCCS, Milano, Italy
c
Dept. of Clinical Sciences and Community Health, Cardiovascular Section, University of Milano, Milano, Italy
d
IRCCS SDN, Napoli, Italy
e
Sleep Medicine Center, S. Luca Hospital, IRCCS, Istituto Auxologico Italiano, Milan, Italy
f
L.I.F.E. Corporation S.A., Luxembourg

a r t i c l e i n f o a b s t r a c t

Article history: Background: Recent advances in wearable technology make continuous cardiorespiratory monitoring possible,
Received 22 December 2017 with potential applications in assessment of cardiopulmonary patients, healthy subjects and athletes. The aim
Received in revised form 8 June 2018 of the present study was to qualitatively and quantitatively evaluate a new wearable device (Learn Inspire
Accepted 11 June 2018 Free Entertain = L.I.F.E.) by embedding in a compression shirt a 12 lead ECG system and 5 respiratory sensors.
Available online xxxx
Methods: Thirty cardiorespiratory patients and ten healthy subjects were studied for 24 h during their usual life
activities. In 8 healthy subjects, simultaneous measurements of the device and of an ergo-spirometer were per-
Keywords:
ECG monitoring
formed during different levels of ventilation in five different body positions. The quality of ECG signals in terms
Respiratory monitoring, wearable device of measurability of heart rate, P wave, QRS complex and ST segment, was analyzed by four expert cardiologists/
Apnea respiratory physiologists using an arbitrary 1–5 scale. The sum of the respiratory signals was used to calculate the
respiratory rate, inspiratory time and relative changes of tidal volume. These parameters were compared to ergo-
spirometer measurements.
Results: Median quality value was N3 for heart rate, QRS complex, ST segment and P wave (except in L3, aVL, aVF,
V1 and V2 leads). Median quality of respiratory traces was N4 in patients and between 3 and 4 in healthy subjects.
The respiratory monitoring of respiratory rate and inspiratory time was accurate in all body positions. Tidal vol-
umes were underestimated due to a high level of ventilation.
Conclusions: The L.I.F.E. device provides an accurate continuous monitoring of cardiorespiratory signals during
the 24 h both in normal subjects and cardiorespiratory patients.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction continuous recording of 1 to 12 ECG leads done for 24 h up to seven


days. The recording of ECG traces is usually obtained by electrodes
Cardiorespiratory monitoring is used to assess athletes, normal sub- and an external recorder fastened together by cables or by a Wireless
jects, and patients with different diseases [1]. It has a pivotal role in pa- technology [3]. In the real world setting, the loss of the signal or poor
tients' management and frequently allows to underscore unexpected signal quality is experienced frequently. Respiratory monitoring can de-
clinical conditions or to explain unclear and frightening symptoms [2]. tect hypopnea and apnea which is obtained by a nasal flow recorder
The monitoring is based on two major aspects: a) ECG monitoring and combined with respiratory movement recorders usually located both
b) respiratory monitoring. Both include day-time and night-time mon- at the chest and abdominal level. As for ECG recording, electrodes and
itoring as well as rest and exercise. ECG monitoring is performed by cables are required and often lead to failed signal connection and vari-
able signal quality [4, 5].
Learn Inspire Free Entertain (L.I.F.E.) device has been built to simul-
⁎ Corresponding author at: Centro Cardiologico Monzino, IRCCS, Dept. of Clinical taneously record ECG and respiratory movement through different sen-
Sciences and Community Health, Cardiovascular Section, University of Milano, Milano,
Via Parea, 20138 Milano, Italy.
sor elements woven into the fabric of a compressible, self-wearable and
E-mail address: piergiuseppe.agostoni@unimi.it (P. Agostoni). washable garment which is even suitable for a residential laundry ma-
1
Both authors equally contributed to the present work. chine. With a single recording, both ECG Holter and polysomnography

https://doi.org/10.1016/j.ijcard.2018.06.044
0167-5273/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
2 A. Sarmento et al. / International Journal of Cardiology xxx (2017) xxx–xxx

data are simultaneously registered with certain advantages. Specifically, wore an implantable cardioverter/Cardiac resynchronization pacemakers. All were
asked to perform their usual daily life activities while wearing the L.I.F.E. garment. Sub-
apnea/hypopnea induced arrhythmias [6–8] are detectable. In a com-
jects were allowed to discontinue recordings only during shower.
prehensive comparability study, several cardiorespiratory monitoring Raw ECG and respiratory data were extracted from the Secure Digital card located in
devices have been proposed but none have the characteristics of the recording module, stored in a dedicated PC and successively analyzed by two different
L.I.F.E. shirt. experts in cardiology and respiratory physiology, respectively. Every 30 min, 1 min was
The following study was therefore undertaken to assess the overall analyzed by each expert by giving grades that ranged from 0 to 5, i.e. 0 = no signal and
impossible interpretation; 1 = poor signal and difficult interpretation; 2 = insufficient
quality of the ECG and respiratory monitoring by L.I.F.E. shirt. Our as- signal and questionable interpretation; 3 = sufficient signal and reliable interpretation;
sessment, by two independent cardiologists, studies a 24 h period of 4 = good signal and reliable interpretation; 5 = excellent signal and reliable interpreta-
the quality of the ECG traces as it pertains to heart rate and specific tion. The experts in cardiology independently analyzed the quality of the signal as a mea-
ECG waves in 12 leads. Additionally, five respiratory traces (plus one sure of heart rate, recording quality of P wave, QRS complex and ST segment. Similarly, the
respiratory physiology experts analyzed the quality of the five respiratory signals and their
sum-signal which combines all the respiratory traces) were assessed
sum. In both cases, data is the average between the two readings.
by two respiratory physiologists in 10 healthy subjects and 30 cardiore-
spiratory patients. In the healthy volunteers, a comparative analysis
between respiratory parameters, assessed by a commercial ergo- 2.3.2. Part 2: validation study
spirometer used as gold standard, and L.I.F.E. shirt was performed in dif- Simultaneous measurements of ventilatory parameters were performed in 8 of the
10 healthy subjects (6 males and 2 females; age 34 ± 15.5 years, weight 71 ± 9.4 kg,
ferent body positions and during different stages of hyperpnea. Simu- height 176 ± 5 cm, body mass index 22.9 ± 3.7 kg/m2) by a commercial ergo-
lated obstructive and central apneas were analyzed as well. spirometer and L.I.F.E. system. After wearing the garment and positioning the spirometer
mouthpiece, recordings were collected during quiet breathing and during five different
2. Materials and methods respiratory maneuvers in which the subjects were asked to increase the spontaneous
tidal volume by ~1.5, ~2 and ~2.5 times, and to emphasize thoracic and abdominal contri-
2.1. Subjects bution to breathing. In addition, two simulations of apneas were performed:
a) obstructive apnea, by externally occluding the spirometer for about 10 s with the sub-
This is a cross-sectional study performed within the confines of the World Medical ject actively trying to breathe; b) central apnea, by asking the subjects to hold the breath
Association Declaration of Helsinki. All subjects were recruited from the Centro for about 10 s at functional residual capacity. Each respiratory maneuver was followed by
Cardiologico Monzino (Milan, Italy), from January 2016 to February 2017, and signed an an interval of quiet spontaneous breathing. The entire sequence of respiratory maneuvers
Inform Consent Form. The study was approved by Centro Cardiologico Monzino Ethical was repeated in five different body positions, namely supine, prone, seated, and right/left
Committee (CCM 448). Inclusion criteria were: age 20–80 years and stable clinical condi- leaning.
tions. L.I.F.E. In this laboratory evaluation, respiratory traces were continuously transmitted
through a Bluetooth connection allowing real-time display and storage into a dedicated
2.2. Measurements computer.
The five respiratory traces were totaled up to provide a single global trace. This was
L.I.F.E.'s medical compression garment (CE approval n. 0068/QPR-DM/068-2016) was low-pass filtered (cut frequency 1.5 Hz), and used to calculate the following parameters
composed of 12 ink-based dry electrodes that lead to ECG monitoring, 5 respiratory strain on a breath-by-breath basis: a) respiratory rate (RR, in breath per min); b) inspiratory
sensors and 1 accelerometer (embedded in the control unit for monitoring the activity and time (Ti, in seconds); c) relative tidal volume, obtained by dividing each signal breath am-
position of the subjects). plitude by the average of the first ten breaths recorded during spontaneous breathing in
The ECG electrodes were located on the internal surface of the garment following the each position.
configuration adopted during any regular exercise test (Fig. 1, supplementary material). After proper alignment of L.I.F.E. and ergo-spirometer signals, all the above respiratory
The ECG electrodes included a layer of adhesive, a layer of conductive ink having conduc- parameters measured by L.I.F.E. were compared to ergo-spirometer measurements,
tive particles, a binder, a solvent, a thickener and a gradient region between the conductive expressed in either absolute (RR and Ti) or relative (tidal volume) values.
ink and the adhesive, comprising a nonhomogeneous mixture of the conductive ink and
the adhesive. Electrodes were connected to a sensor module interface via a stitched zig-
zag connector formed on a separate piece of compression fabric attached to the garment 3. Statistical evaluation
body. In order to minimize the relative movement between the ECG dry electrodes and
skin, each ECG electrode was equipped with an expandable support, shaped to control
the direction of expansion in the axis perpendicular to the electrode and held between All data are expressed as mean ± standard deviation unless other-
the sensor present on the inner, wearer-facing surface of the garment, and the outer sur- wise stated. Differences between ventilatory parameters obtained by
face, less flexible second fabric integrated with the expandable support. The outer fabric L.I.F.E. and ergo-spirometer in each maneuver and position were
was integrated into the garment body and adapted to include elastic straps (Compression expressed in relative values (i.e. data obtained with L.I.F.E garment
Garments Having Stretchable And Conductive Ink, US Patent n. 8948839, Feb 3 2015).
minus data obtained during ergo-espirometer divided by data obtained
The shirt included five respiratory sensors, positioned on the anterior external surface
of the garment as follows: two thoracic, at the level of the manubrium on the left and right
during ergo-espirometer). Data has also been described by means of
sides; one central, at the level of the xiphoid process and two abdominal, between box plot. Intergroup and intragroup data, regarding respiratory traces
the lower costal margin and the umbilical level on the left and right sides, respectively and ECG signals, were compared using Mann-Whitney and Kruskal-
(supplementary material, Fig. 1). Respiratory sensors were made of an elastic ribbon im- Wallis tests, respectively. For the later, in the event of a statistically sig-
pregnated with conductive ink, similar to that used for the ECG electrodes, an electrical
nificant difference, Dunn's post hoc test with adjusted p values was ap-
connector at each end of the elastic ribbon, and a cover made of compression fabric. The
strain respiratory sensors provided electrical resistance variations measured at a sampling plied to identify the difference signals. Cohen's kappa coefficient (k) was
rate of 50 Hz. used to study inter-rater agreement and the levels of agreement
were determined, as proposed by Landis & Koch [9], with k values
2.2.1. Ergo-spirometer 0.00–0.20 considered as slight agreement; 0.21–0.40, fair agreement;
A flow-sensing V-max 229D Spectra metabolic cart, (Sensor Medics, Yorba Linda,
0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement,
California), previously calibrated with a 3-liter syringe in BTPS conditions, was used to
compare the ventilatory parameters obtained by the L.I.F.E. garment. and 0.81–1.00, almost perfect agreement.
Signal quality was also compared between groups of different dis-
2.3. Protocol eases (chronic obstructive pulmonary disease, heart failure and coro-
nary artery disease patients). In addition, patients were also divided
The study protocol was divided into two different parts. The first consisted of a
into those presenting BMI b 30 and ≥30 kg/m2 in order to know if obe-
24-hour continuous recording during daily life activities either in healthy subjects
and patients. The second part was conducted only in healthy subjects and performed sity would interfere in the ECG and respiratory data collection. The for-
in a laboratory environment. mer analysis was performed using Kruskal-Wallis with Dunn's post hoc
test (in case of statistical significance) and the latter was performed by
2.3.1. Part 1: 24-hour test applying the Mann-Whitney test.
Thirty cardiorespiratory patients (10 chronic obstructive pulmonary disease, 10 heart
failure and 10 coronary artery disease patients, 26 males and 4 females, age 68 ± 12,
The statistical computations were carried out using SPSS 23.0 statis-
height 172 ± 8 cm, weight 80 ± 16 kg), and 10 healthy subjects (8 males and 2 fe- tical software (IBM Corp.). For all statistical analyses, a significance level
males, age 39 ± 16, height 177 ± 6 cm, weight 71 ± kg) were recruited. Six patients of b0.05 was adopted.

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
A. Sarmento et al. / International Journal of Cardiology xxx (2017) xxx–xxx 3

4. Results generally lower than in healthy subjects, both in terms of median values
and interquartile ranges. Median score values, however, were statisti-
As shown in Supplementary Table 1, an overall substantial agree- cally greater in all derivations for heart rate, but V1 and V2 (p b 0.01).
ment inter-rater reliability was observed regarding the quality of the re- Regarding QRS complex, V1, V2 and V3 derivation grades were signifi-
spiratory and ECG signals. In the ECG signal, only the QRS complex cantly lower (p b 0.001) in patients. ST segment and overall quality
showed a moderate level of agreement (k = 0.424, p b 0.05). had similar lower quality in patients when compared with healthy
In Fig. 1, a representative example of all 12-lead ECG and the five re- (Fig. 2). For P-wave, median values of the grades in patients were statis-
spiratory traces (with the sum signal on the bottom right) recorded by tically lower in all derivates but L1 (p b 0.001) when compared with
L.I.F.E. device is shown. The figure is a screenshot of the monitoring soft- healthy. Two patients were excluded from P wave analysis due to per-
ware taken during a 30 s time interval of the validation study. manent atrial fibrillation.
Fig. 2 shows the results of intergroup analysis regarding the quality Fig. 3 shows the results of quality assessment, regarding intergroup
assessment in terms of measurability of heart rate, P wave, QRS com- analysis, of all respiratory traces, including the sum signal, obtained in
plex, ST segment, and overall ECG signal from the 12 derivations, ob- both healthy subjects and patients. Adjusted p values for intragroup
tained from expert cardiologists in both healthy subjects and patients. analysis are shown in Supplementary table 2. In terms of median values
Adjusted p values for intragroup analysis are shown in Supplementary of the grades, the quality of the signals in patients was slightly better
table 2. The quality of most of the ECG signals was evaluated in the than in healthy subjects (p b 0.05) for left (median of 4.50 versus
vast majority with grades above 4. As expected in healthy subjects, 3.97) and right (median of 4.55 versus 3.82) abdominal sensors. For
aVL and aVF derivations resulted to be the worst compared to all the left and right thoracic sensors, respectively, median values were healthy
other derivations (p b 0.01). This is shown in supplementary material (median of 4.23 and 4.26) were slightly better than in patients (median
table 3, where the percentage of the intervals evaluated with grades of 4.62 and 4.58). Median score values were always higher than 4 for all
≥ 3 is reported for all derivations. The signals quality in patients was sensors. In healthy subjects the left and right abdominal sensors

Fig. 1. Example of all 12-lead ECG and the five respiratory traces recorded by L.I.F.E. device.

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
4 A. Sarmento et al. / International Journal of Cardiology xxx (2017) xxx–xxx

Fig. 2. Intergroup quality assessment in terms of measurability of heart rate, P wave, QRS complex, ST segment, and overall ECG signal from the 12 derivations. Quality grades are shown in
the Y-axis. Intergroup comparisons: † p b 0.05, †† p b 0.01 and ††† p b 0.001,

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
A. Sarmento et al. / International Journal of Cardiology xxx (2017) xxx–xxx 5

Fig. 3. Intergroup quality assessment in terms of measurability of respiratory traces. Quality grades are shown in the Y-axis. Intergroup comparisons: † p b 0.05.

provided median signal quality (3.97 and 3.82, respectively) between 3 24 h either in normal subjects and cardiorespiratory patients. For respi-
and 4 (p b 0.01) (Supplementary Table 2). Total sum between healthy ratory monitoring, this is not true for high tidal volumes. To the best of
(median of 4.78) and patients (median of 4.85) was not significantly dif- our knowledge, L.I.F.E. device is the first cable-less simultaneous 12 ECG
ferent. In the supplementary material, table 4 reports the percentage of leads and respiratory movement recorder, which uses different sensor
the intervals evaluated with grades above or equal than 3. Fig. 4, reports elements embedded in the fabric of a compressible, easy self-wearable
an example of central (B) and obstructive (C) apneas recorded by garment.
chance in 2 patients and one healthy subjects (A). The quality of tracings Quality of ECG and respiratory traces were analyzed using arbitrary
was excellent and data was clearly diagnostic. units. The sampling of traces was random and done regardless of
The quality of the respiratory signals was not different between pa- subject's body position or activity. Overall quality for ECG was good
tients. However, when analyzing ECG signals, heart failure patients (N3) which means good quality with easy interpretation of the data. In
showed a lower quality (p = 0.01) in V5 (median of 4.50 versus 4.86) healthy subjects aVL and aVF leads showed the highest quality variabil-
and V6 (median of 4.44 versus 4.86) derivations of the QRS complex ity for all the analyzed parameters. This is expected because in such
when compared with chronic obstructive pulmonary disease patients. leads in normal subjects' P wave and QRS are isodifasic, i.e. close to 0
Although significantly differences were found, quality of the signals amplitude, thus difficult to evaluate even in a standard ECG. Conse-
was still considered good (4 – reliable interpretation). Seven patients quently, as in ECG in normal subjects, heart rate cannot be easily mea-
presented BMI ≥ 30 kg/m2 and no differences were observed regarding sured while ST segment is characterized by a flat T wave, again
the quality of both the ECG and respiratory signals when compared with difficult to evaluate. As expected, patients recordings showed more
those presenting BMI b 30 kg/m2 (data not shown). quality variability. Indeed heart rate detection, albeit very good in all
A representative example (supplementary material, Fig. 2) of the leads, showed the greatest variability in V1 and V2 because in many
overall signal obtained by totaling the signals of the five respiratory sen- subjects QRS is isodifasic or flat and more subject to respiratory varia-
sors during the different phases of the validation study protocol (top tion in such leads. The same applies to P wave and QRS, from the latter
panel) and the corresponding breath-by-breath values of relative tidal heart rate is measured, extending the limitation to aVL and aVF for the
volume, obtained after alignment of L.I.F.E. and ergo-spirometer signals same reasons. ST segment analysis showed a very good quality from
(bottom panel) is described. During the periods of spontaneous quiet V3 to V6 with the greatest variability in the inferior leads and in V1
breathing, increased tidal volume (~1.5 and ~2 times) and abdominal and V2, but again this is usual during activity and with different body
breathing, relative variations of tidal volume were in good agreement. positions.
Conversely, during the periods of increased tidal volume (~2.5 times) Overall quality of median values of respiratory traces was very good
and, mainly, thoracic breathing, relative variations of tidal volume mea- with a slightly better overall grade in patients (Table 4, supplementary
sured by L.I.F.E. device underestimated those measured by the ergo- material). It must be recognized that in normal subjects when com-
spirometer. Central and obstructive apneas were simulated by all sub- pared to tidal volume measured by a standard ergo-spirometer, with
jects and recorded as such by L.I.F.E. shirt. high volumes and particularly with thoracic ventilation, L.I.F.E. device
The median relative difference values of RR, Ti and tidal volume, as tidal volume measurement was lower than needed for a reliable evalu-
well as their interquartile ranges, during the different phases of the ation. This is likely due to the reaching of the maximum stretching
study protocol and the different body positions are reported in Fig. 3 possibility on the thoracic volume transducer (Figs. 2 and 3 in supple-
(supplementary material) in left and right panel, respectively. The me- mentary material). In daily activity, a tidal volume increase of
dian values obtained for RR was around 0% with slightly interquartile N2 times is rarely achieved except during strenuous exercise.
range above this point and for Ti was always below 10%for all the con- From a clinical point of view, the above reported data opens a new
sidered maneuvers and body positions. The median relative difference frontier in ambulatory cardiorespiratory monitoring, allowing simulta-
between tidal volume measured by L.I.F.E. and ergo-spirometer was neous recording of ECG and respiratory traces. Moreover, as shown in
lower, particularly during the periods of increased tidal volume and Fig. 4, respiratory traces allowed a reliable identification of simulated
both thoracic and abdominal breathing. apneas and were very precise when analyzing normal or slightly in-
creased tidal volumes in all body position. This was obtained without
5. Discussion any nasal flow cannula which frequently is not well tolerated by
subjects.
This study shows that ECG and respiratory movement monitoring Finally, it is recognized that regardless of the above reported results,
can be performed with high-quality recordings by L.I.F.E. device up to more technical developments and further studies are needed to

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
6 A. Sarmento et al. / International Journal of Cardiology xxx (2017) xxx–xxx

Fig. 4. Example of simulated obstructive and central apneas recorded in a healthy subject (A), and central (B) and obstructive (C) apneas recorded by chance in 2 patients who participated
to the present study.

improve and confirm the clinical applicability of L.I.F.E. device. Specifi- Conflict of interests
cally, further needed are a) a different respiratory sensor position and
shape for high tidal volumes recording, b) EEG tracing for sleep quality Alessandra Leonardi, Dario Ossola, Rudy Rigoni are L.I.F.E. em-
definition, and c) body position and activity recording by inertial sen- ployees. None participated in data reading.
sors such as multi-axis accelerometers [10]. Moreover, a quantitative Andrea Aliverti received a consultation fee from L.I.F.E.
comparison between standard polysomnographic monitoring and
L.I.F.E. device in patients with documented sleep apnea/hypopnea is Acknowledgments
needed.
In conclusion, it seems to us that L.I.F.E. multiparametric recording Authors contributions: PA and AA had full access to all of the data in
technology is a step forward in ambulatory subjects monitoring. the study and takes responsibility for the integrity of the data and the
Supplementary data to this article can be found online at https://doi. accuracy of the data analysis, including and especially any adverse ef-
org/10.1016/j.ijcard.2018.06.044. fects. AS, CV, SP, CL, AS, FN, MM, PA, and AA contributed substantially

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
A. Sarmento et al. / International Journal of Cardiology xxx (2017) xxx–xxx 7

to the study design, data analysis and interpretation, and the writing of force of the American Academy of sleep medicine, J. Clin. Sleep Med. 8 (2012)
597–619.
the manuscript. AL, DO, RR contributed substantially to study design, [5] V.K. Kapur, D.H. Auckley, S. Chowdhuri, D.C. Kuhlmann, R. Mehra, K. Ramar, et al.,
data interpretation, statistical analysis and writing of the manuscript. Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea:
an American Academy of sleep medicine clinical practice guideline, J. Clin. Sleep
Med. 13 (2017) 479–504.
References [6] A.M. May, D.R. Van Wagoner, R. Mehra, OSA and cardiac arrhythmogenesis: mech-
anistic insights, Chest 151 (2017) 225–241.
[1] E. Piotrowicz, A. Jasionowska, M. Banaszak-Bednarczyk, J. Gwilkowska, R.
[7] G. Parati, C. Lombardi, F. Castagna, P. Mattaliano, P.P. Filardi, P. Agostoni, Heart fail-
Piotrowicz, ECG telemonitoring during home-based cardiac rehabilitation in heart
ure and sleep disorders, Nat. Rev. Cardiol. 13 (2016) 389–403.
failure patients, J. Telemed. Telecare 18 (2012) 193–197.
[8] G. Parati, C. Lombardi, K. Narkiewicz, Sleep apnea: epidemiology, pathophysiology,
[2] C.A. Goudis, D.G. Ketikoglou, Obstructive sleep and atrial fibrillation: pathophysio-
and relation to cardiovascular risk, Am. J. Phys. Regul. Integr. Comp. Phys. 293
logical mechanisms and therapeutic implications, Int. J. Cardiol. 230 (2017)
(2007) R1671–R1683.
293–300.
[9] J.R. Landis, G.G. Koch, The measurement of observer agreement for categorical data,
[3] J.S. Steinberg, N. Varma, I. Cygankiewicz, P. Aziz, P. Balsam, A. Baranchuk, et al., 2017
Biometrics 33 (1977) 159e174.
ISHNE-HRS expert consensus statement on ambulatory ECG and external cardiac
[10] A. Aliverti, Wearable technology: role in respiratory health and disease, Breathe 13
monitoring/telemetry, Heart Rhythm. 14 (2017), e55-e96.
(2017), e27-e36. https://doi.org/10.1183/20734735.008417.
[4] R.B. Berry, R. Budhiraja, D.J. Gottlieb, D. Gozal, C. Iber, V.K. Kapur, et al., Rules for
scoring respiratory events in sleep: update of the 2007 AASM manual for the scoring
of sleep and associated events. Deliberations of the sleep apnea definitions task

Please cite this article as: A. Sarmento, et al., Qualitative and quantitative evaluation of a new wearable device for ECG and respiratory Holter
monitoring, Int J Cardiol (2017), https://doi.org/10.1016/j.ijcard.2018.06.044
View publication stats

S-ar putea să vă placă și