Comparison of oscillometric pulse amplitude envelopes recorded from the
- 格式：pdf
- 大小：613.06 KB
- 文档页数：7
Medical Engineering &Physics 32 (2010) 1124–1130Contents lists available at ScienceDirectMedical Engineering &Physicsj o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /m e d e n g p hyComparison of oscillometric pulse amplitude envelopes recorded from the locally compressed radial arteriesR.Raamat ∗,J.Talts,K.Jagomägi,J.KivastikDepartment of Physiology,University of Tartu,Tartu,Estoniaa r t i c l e i n f o Article history:Received 7March 2010Received in revised form 1August 2010Accepted 7August 2010Keywords:Oscillometric envelope Oscillometric pattern OscillogramRadial non-invasive blood pressure Photoplethysmographic Pneumoplethysmographica b s t r a c tSimultaneously recorded oscillometric envelopes,obtained pneumo-and photoplethysmographically from a small local pad-type pneumatic cuff on the left and from a larger cuff on the right radial artery,were compared in 17healthy subjects.For oscillometric estimation,specific fixed ratios based on evidence in the literature were used.The obtained envelopes for each person were shifted and aligned at the point of upper arm mean arterial pressure for this person,thus eliminating the brachial-to-radial mean blood pres-sure gradient and possible left–right difference.In this way,the shape of differently recorded envelopes as a determinant of the accuracy of oscillometric estimation was studied.Results showed an advan-tage of photoplethysmographically compared to pneumoplethysmographically recorded envelopes.For a smaller cuff (diameter 40mm),the mean difference in mm Hg ‘oscillometric estimate minus auscul-tatory reference’and standard deviation were in the case of photo recording for systolic and diastolic pressures −0.6(6.3)and 1.2(3.4),respectively.In the case of pneumo recording,these parameters were considerably larger,being 12.1(11.9)and −6.2(10.9),respectively.For a larger cuff the same tendency was revealed.Photo recording was found to be less sensitive to alterations in the cuff size and characteristic ratios.© 2010 IPEM. Published by Elsevier Ltd. All rights reserved.1.IntroductionThe oscillometric method is increasingly used in blood pres-sure (BP)measurement.Due to the simplicity and reliability of this method,the oscillometric principle is employed in the majority of present-day automatic and semiautomatic non-invasive blood pressure monitors.Oscillometric blood pressures are typically determined from the envelope of successive oscillometric pulse amplitudes obtained from the occlusive cuff during its inflation or deflation.The highest point of the envelope curve is generally regarded as the mean arte-rial pressure (MAP)[1,2].This is consistent with the oscillometric maximum criterion described by Marey more than 130years ago [3].Two general types of criteria have been used to estimate systolic (SBP)and diastolic (DBP)blood pressures:height-based criteria and slope-based criteria [4].In the height-based approach the systolic and diastolic pressures are determined using special fractions of the maximum oscillation amplitude [5,6].These fractions are known as characteristic ratios (systolic and diastolic,respectively).In the slope-based approach the cuff pressure at which the oscillometric pulse amplitude increases rapidly is taken as the systolic pressure,∗Corresponding author.Tel.:+3727374987.E-mail address:raamat@ut.ee (R.Raamat).while that at which the amplitude decreases rapidly is taken as the diastolic pressure.Mathematically,the latter criterion defines the points at which the first derivative of the envelope curve is maximum or minimum.There are no definitive correct values for the systolic and dias-tolic characteristic ratios,but rather that different manufactures use different values.Theoretical and experimental work has sug-gested the following values for characteristic ratios:(a)systolic from 0.46to 0.64and diastolic from 0.59to 0.80(theo-retical work by Ursino and Cristalli [7]),(b)systolic 0.593and diastolic 0.717(theoretical work byDrzewiecki et al.[6]),(c)systolic from 0.45to 0.57and diastolic from 0.69to 0.89(exper-imental work by Geddes et al.[5]),(d)systolic (mean and SD)equal to 0.49(0.11)and diastolic equalto 0.72(0.12)(experimental work by Amoore et al.[8]).Specific studies have demonstrated an effect of the shapes of oscillometric pulse amplitude envelopes on the differences between auscultatory and oscillometric BP measurement [8,9].Subsequently,recording of oscillometric envelopes of appropriate shape and stability remains a topical issue for the oscillometric method.Oscillometric wrist monitors have shown lower reliability com-pared to upper arm devices.On the other hand,oscillometric wrist1350-4533/$–see front matter © 2010 IPEM. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.medengphy.2010.08.002R.Raamat et al./Medical Engineering&Physics32 (2010) 1124–11301125devices have small dimensions and are highly user-friendly.It is worth to mention that due to collateral circulation at the human wrist area,it is possible to compress only a single(radial)artery while another(ulnar)artery continues to supply blood to the hand [10,11].This means that repeated or long-term measurements do not cause any venous congestion and ischaemic pain at distal sites to the cuff.Traditionally,the arterial pulsations in oscillometric devices are picked up from the occlusing cuff by means of a manometer[4]. However,in some implications a photoelectric detection directly from the tissue(photoplethysmography)is also used[12].Other methods than pneumatic pulse sensing seem to become more pop-ular in the wrist BP devices[10,11,13].Cuff size has been shown to be an important determinant of the accuracy of BP measurement[14,15].Because of differences in anatomical structure,recommendation concerning the cuff size for upper arm cannot simply be applied for the wrist[11].In the present study we compare the shapes of pneumo-and photoplethysmographically recorded oscillometric envelopes obtained simultaneously from the radial arteries during partial compression of the wrist by local cuffs of different size.By apply-ing height-based criteria we assess which method of recording of pulse amplitude envelopes and which cuff size have a higher potential for accurate oscillometric wrist BP estimation.Certain fixed characteristic ratios,based on the data in literature,are used.2.Methods2.1.SubjectsA group of17volunteers,10females and7males,aged from18 to35,were studied.They all were normotensive,had no history of vascular disease and gave their informed consent to participate in the study.The study was approved by the Ethics Committee of the University of Tartu.2.2.Experimental design and protocolDuring the experiment a subject was lying comfortably on a couch with both hands resting at heart level.Systolic and dias-tolic reference BP was measured at the subject’s left upper arm employing the auscultatory method(Precisa N sphygmomanome-ter,Rudolf Riester GmbH&Co.Kg).Auscultatory SBP and DSP were taken twice before and twice after the recording of oscillometric envelopes.A pneumatic pad-type cuff(Cuff1)containing an elastic mem-brane was attached to the left wrist by aflexible Velcro strap and a U-shaped aluminium clip(Fig.1).The latter served to locally counter-balance the force exerted by an inflated cuff to the strap, in this way preventing the pressurisation of the ulnar artery. The photoplethysmographic sensor incorporated an infrared LED TSMG2700(820nm,10mW,Vishay Corp.)and a daylight protected photodiode BPW34FS(Siemens Corp.),both having small dimen-sions to be easilyfitted under the cuff.The light pathway was orthogonal to arterial axis and the used wavelength was typical of infrared photoplethysmographs.Before adjusting Cuff1(diameter40mm,thickness of the elas-tic vinyl membrane0.1mm),a photoplethysmographic sensor was attached over the radial artery by means of two adhesive strips. The correct place for the sensor was found by palpation.As a rule, it was located at the very distal end of the radius on the volar surface[10].The sensor was positioned in a way that the artery remained between the photodiode and LED.Optical components were adjusted at a15mm distance between centres,thus notaffect-Fig.1.Pad-type wrist cuff:1,photoplethysmographic sensor;2,pneumatic capsule with an elastic membrane;3,velcro strip;4,metallic cliping the correct pressure transfer from the elastic membrane to the artery.Another photoplethysmographic sensor was placed on the right radial artery in the same way as that on the left hand.Differently from the left hand,the right artery was locally compressed by a larger cuff(Cuff2)having a bladder of85mm×45mm(an infant cuff from the Babyphone set,Rudolf Riester GmbH&Co.Kg).Cuff2 was placed symmetrically over the right radial artery.Pressure in both applied cuffs was controlled by a common lin-ear deflation generator(Fig.2).The connecting narrow tubes were chosen to enablefilling/emptying of both pneumatic cuffs with an equal time constant of approximately0.7s.This was found to be a good compromise to precisely track the relatively low deflation rate of the linear deflation generator(2mm Hg/s),and to enable a satisfactory high-passfiltering of picked up pulsations(lower cut-off frequency of0.24Hz).A possible interaction between the cuffs was avoided by use of connecting tubes of narrow diameter and a low output impedance of the linear deflation generator.After an initial equilibrium period of5min,three cuff deflations were carried out with a30s interval between them.Pressure was rapidly raised to160mm Hg and then linearly lowered at a rate of 2mm Hg/s.The total time for the measurement session in each subject was approximately10min.2.3.Signal processing and data analysisPressure signals P1(t)and P2(t)from Cuff1and Cuff2, respectively,were recorded by two similar pressure transducers (0–200mm Hg,0–100Hz),optical transmittance signals I1(t)and I2(t)by two similar photoplethysmographs(Fig.2).The amplified and conditioned analog signals were digitised by an AD converter(16-bit,100Hz)and transferred to the computer like in our previous studies[16,17].Assuming that the Lambert–Beer’s law holds true in the tissues where the photoplethysmogram is measured[18–20],a logarith-mic photometer should be used to provide for a linear relationship between the recorded light transmittance changes and the real blood volume changes.Since the applied photoplethysmograph did not contain a logarithmic amplifier,this operation was executed by the computer software before extracting the oscillometric pulses.1126R.Raamat et al./Medical Engineering &Physics32 (2010) 1124–1130Fig.2.Block diagram of the experimental setup.In each of the 17persons,the measurements during three cuff deflations were included in the analysis.For further evaluation,the envelopes were smoothened by a moving window with a width of 9cardiac cycles.This was found to effectively suppress distortions from breathing while having a minimum influence on the shape of envelopes.Considering the heart rate of subjects and the used deflation rate,the corresponding smoothing window expressed in pressure units was approximately 16–20mm Hg.Thereafter the points of maximum oscillation were found.To assess the quality of obtained envelopes for oscillomet-ric estimation,considering the shape while suppressing variation from the location of points of maximum oscillation,all the envelopes related to one subject were shifted along the pres-sure axis and aligned at the point equal to the reference MAP for this subject.This operation was also expected to remove the brachial-to-radial pressure gradient as well as possible left–right pressure difference if it existed.The reference MAP was com-puted from the auscultatory SBP and DBP by the traditional formula [21]:MAP =DBP +(SBP −DBP)/3.It is known that this approximation is influenced by the shape of the arterial pres-sure pulse [22,23].Hence,taking into account a homogeneity of the test-group,the rest situation and supine position,variation in the pressure pulse shape index can be assumed to be relatively low.Besides the pressure-directional normalisation,an amplitude-directional normalisation was introduced by dividing the oscilla-tion amplitudes by their maximum value during every deflation.When executing oscillometric estimation,the systolic charac-teristic ratio 0.5and diastolic characteristic ratio 0.8were applied [5].Table 1Individual and group-averaged results of oscillometric estimation of the radial blood pressure compared to auscultatory reference for Cuff 1.SubjectAuscultatory (reference)Oscillometric—auscultatory SBP (mmHg)DBP (mmHg)MAP (mmHg)SBP (mmHg)DBP (mmHg)PhotoPneumo Photo Pneumo 1126.080.095.3−4.2 2.60.8−2.52114.066.082.0 6.121.50.9−3.63102.568.379.70.215.7−1.6−5.14118.874.088.9 2.4 6.90.9 1.55120.574.589.8−0.520.6 1.7−15.56124.867.086.3−9.419.9 2.1−0.87115.068.083.7−5.717.20.9−9.88123.569.387.3 3.627.7 2.7−17.89104.062.576.3−2.60.2 3.20.410110.864.379.8−5.00.2 3.4−4.011111.564.880.3−8.2−7.4 4.2−0.312127.570.889.711.714.8 2.9−30.413108.370.082.80.523.2−4.3−7.614123.364.884.30.611.7 5.6−2.61590.058.068.7−1.1 4.5−2.4−1.016100.058.072.0−3.9−2.5 2.6 1.517112.064.880.1 5.328.5−2.9−7.3Mean 113.767.382.8−0.6a 12.1c 1.2b −6.2c SD(10.4)(5.6)(6.7)(6.3)(11.9)(3.4)(10.9)a Not significant (p =0.65)calculated by Student’s paired test.b Not significant (p =0.08)calculated by Student’s paired test.cSignificant (p <0.01)calculated by Student’s paired test.R.Raamat et al./Medical Engineering&Physics32 (2010) 1124–11301127Fig.3.Example of normalised individual oscillometric envelopes A=f(CP),recorded simultaneously applying different recording and different cuffs on the radial artery of Subject3.Black lines denote Cuff1,grey lines Cuff2,solid lines photo and lines with circles pneumo recording,respectively.Levels for the amplitude-based oscillo-metric estimation of SBP and DBP as well as corresponding auscultatory references are shown.The normalised oscillometric amplitude A is dimensionless,cuff pressure CP is in mm Hg.To compare the shapes of envelopes obtained by applying differ-ent cuffs and different methods of recording over all the subjects, a slightly modified pressure-directional normalisation procedure was used:the normalised amplitude A was not plotted against the cuff pressure CP,but against the pressure P=P max−CP,where P max was the pressure coordinate for every envelope at which A was of maximum value.This allowed a simple comparison of the shape of multiple curves A=f(P).Before the construction of sets of envelopes A=f(P)and cal-culating the averaged envelopes,the individual envelope curves were linearly interpolated and sampled at equally spaced intervals 2mm Hg.2.4.StatisticsTo test for the presence of significant differences,Student’s paired t-test was used.To test all the hypotheses,a level of signifi-cance of0.05was applied.Data are expressed as mean and standard deviation(in parentheses).3.ResultsResults of oscillometric estimation of the radial artery blood pressure applying local compression by Cuff1are listed in Table1. Its left part presents averaged data over four auscultatory mea-surements on the brachial artery of each subject(as reference). The right part contains pressure differences‘oscillometric minus auscultatory’for pneumoplethysmographic as well as for photo-plethysmographic recording.The exposed data are differences for every subject averaged over three cuff deflations.Table2presents data for Cuff2analogous to Table1.As for some subjects the systolic criterion0.5was not applicable when pneu-moplethysmographic recording was applied,the related systolic pressures are missing in Table2.Fig.3is an example of normalised individual oscillometric envelopes A=f(CP),obtained by applying different methods of recording and different cuffs on the radial artery of Subject3.Levels for the amplitude-based oscillometric estimation of SBP and DBP as well as corresponding auscultatory references are shown.Figs.4and5expose normalised individual oscillomet-ric envelopes A=f(P)for Cuff1and Cuff2,respectively.Fig.4.Set of normalised individual oscillometric envelopes A=f(P)recorded from the locally compressed radial artery by Cuff1(n=17).(a)Envelopes recorded photoplethysmographically;(b)envelopes recorded pneumoplethysmographically. Ranges of estimated SBP and DBP for the whole group on the levels of0.5and0.8 are shown.The normalised oscillometric amplitude A is dimensionless,pressure P is in mmHg.Fig.5.Set of normalised individual oscillometric envelopes A=f(P)recorded from the locally compressed radial artery by Cuff2(n=17).The symbols and units are the same as in Fig.4.1128R.Raamat et al./Medical Engineering&Physics32 (2010) 1124–1130Table2Results analogous to those in Table1for Cuff2.Subject Auscultatory(reference)Oscillometric—auscultatorySBP(mmHg)DBP(mmHg)MAP(mmHg)SBP(mmHg)DBP(mmHg)Photo Pneumo Photo Pneumo1126.080.095.318.3 4.8−2.7 2114.066.082.07.536.5 4.5−8.5 3102.568.379.70.724.2 3.4−11.6 4118.874.088.9−9.8 2.9−9.6 5120.574.589.816.319.3 2.30.3 6124.867.086.3−6.5 6.3 1.8 7115.068.083.7−11.3 1.7−4.8 8123.569.387.3−10.2 4.1−5.9 9104.062.576.3−8.2 3.8 1.8 10110.864.379.80.5 3.5−0.5 11111.564.880.30.826.8 5.6 1.1 12127.570.889.7−11.3−5.8 6.4 6.4 13108.370.082.8−7.5 6.0 1.8−12.3 14123.364.884.3−13.07.5 1.5 1590.058.068.7−4.8−1.8−9.8 16100.058.072.0−7.5 1.0−13.017112.064.880.1−6.9 6.3−18.2Mean113.767.382.8−3.1a 3.8b−4.9b SD(10.4)(5.6)(6.7)(8.6)(2.7)(7.3)a Not significant(p=0.19)calculated by Student’s paired test.b Significant(p<0.01)calculated by Student’s paired test.Oscillograms are normalised in the way described in Section2.Eachfigure contains a set of17individual curves obtained photoelectrically as well as pneumatically.The group ranges of oscillometrically estimated SBP and DBP are indicated.For the sets of individual oscillograms presented in Figs.4and5, the group-averaged envelopes have been calculated(Fig.6).Con-siderable differences in these envelopes depending on the method of registration and size of the cuff can be seen.4.DiscussionThe mainfinding of the study was that photoelectrically vs. pneumatically recorded oscillometric envelopes,obtained from the locally compressed radial arteries,fitted better for oscillometric estimation having less variability and better agreement between estimates and references.The following aspects of comparison illustrate these assertions:(a)Tables1and2demonstrate a better accuracy and less variabilityof photo vs.pneumo recorded data.Fig.6.Normalised group-averaged oscillometric envelopes A=f(P)for four sets of individual curves presented in Figs.4and5.The denotation of lines is the same as in Fig.3.(b)Figs.4and5clearly show less variability of photo vs.pneumorecorded envelopes.(c)Figs.3and6demonstrate that photo recorded envelopes aresharper compared to pneumo recorded curves.Because of a higher slope of photo derived envelopes,possible variation of characteristic ratios causes a smaller error of estimation com-pared to that in the case of less steep pneumatically recorded curves.(d)Figs.3and6expose a lower dependence of photo vs.pneumoderived envelopes on the cuff size.The empirical nature of the choice of characteristic ratios is one of the limitations of the study:by varying the values offixed char-acteristic ratios,the higher or lower SBP or DBP estimates can be obtained.In this study we employed the same values of ratios for all the recorded envelopes,the systolic characteristic ratio being 0.5and the diastolic ratio0.8.The choice was based on evidence in the literature[5,8,15,24,25].Two more limitations can be noted.The range of ages considered in a young population allowed us to better demonstrate variabil-ity of envelopes resulting from the method of recording rather than from the individual haemodynamic parameters of subjects. Also a simplification was introduced when we shifted and aligned the envelopes for each person at the point of upper arm reference MAP,thus eliminating the brachial-to-radial mean blood pressure gradient and possible left–right difference.Thus,the shape of dif-ferently recorded envelopes as a determinant of the accuracy of oscillometric estimation was observed.Following data in Table1,one can see that for a smaller cuff(Cuff 1),the mean difference in mm Hg‘oscillometric estimate minus auscultatory reference’and standard deviation were in the case of photo recording for systolic and diastolic pressures−0.6(6.3) and1.2(3.4),respectively.In the case of pneumo recording,these parameters were considerably larger,being12.1(11.9)and−6.2 (10.9),respectively.When using local compression with a wider cuff(Cuff2)and photo recording,the mean difference in mm Hg‘oscillometric esti-mate minus auscultatory reference’and standard deviation were −3.1(8.6)and3.8(2.7)for SBP and DBP,respectively(Table2). Applying Cuff2and pneumo recording,the corresponding value forR.Raamat et al./Medical Engineering&Physics32 (2010) 1124–11301129DBP was−4.9(7.3),but for SBP the mean difference could not be estimated since the normalised amplitude of oscillations in a num-ber of subjects did not fall lower than the0.5level at suprasystolic pressures.Analysing mean differences in Tables1and2,we should keep in mind that auscultatory readings were taken from the brachial while oscillometric ones were obtained from the radial artery.It is known that due to the pulse reflection,SBP and DBP can vary from the brachial artery to the radial artery:the radial SBP can be higher and the radial DBP lower than corresponding brachial pres-sures.However,the pulse pressure amplification for young subjects with compliant arteries under afore described conditions can to be predicted small,less than12%[26].If assuming a potential pulse pressure amplification in the left hand(Cuff1),the pneumatically recorded higher radial SBP and lower radial DBP can be regarded as possible physiological changes,but the magnitude of the pulse pressure change of18.3mm Hg(59%)appears to be too big for a physiological change.Thus,the pulse amplification cannot be con-sidered as a source of large differences in Table1.The worse agreement in the case of pneumo recording com-pared to photo recording is graphically demonstrated in Fig.3, where a set of normalised oscillometric envelopes is shown for Sub-ject3.All the envelopes were recorded simultaneously,thus they reflect the same haemodynamic condition of the subject.For both cuffs,the envelopes labelled‘pneumo’were considerably broader compared to those labelled‘photo’.In oscillometric estimation this difference causes an overestimation of SBP and an underestimation of DBP(note auscultatory references below the envelopes).The different width of photo vs.pneumo derived patterns can be explained by the used method of recording and the anatomy of involved arterial segments:an infrared beam controls only a short superficial segment of the radial artery located at the very distal end of the radius on the volar surface,while the pneumatic cuff collects signals from a wider area,involving also a deeper placed segment between the tendon of the brachioradialis muscle and that of theflexor capri radialis muscle[11,13].It is understood that a broadening of the pneumatically recorded envelope can occur in the latter case.A lower dependence of photo vs.pneumo derived envelopes with respect to the cuff size(Figs.3and6),revealed in the present study can also be regarded as an outcome of the above described phenomenon:an optical method samples the artery only locally and is less influenced by the residual part of the cuff.Individual variability in the shape of oscillometric envelopes can be observed in more detail in Figs.4and5,where sets of normalised patterns and corresponding ranges of estimated SBP and DBP are shown for different cuffs and different recording.Thesefigures graphically illustrate results presented in Tables1and2,indicating a larger variability in the estimated blood pressure when pneumo-plethysmographic compared to photoplethysmographic recording was used.It can be seen in Fig.5b that for some subjects in the group the systolic criterion0.5was not applicable.In principle,the applied systolic characteristic ratio can be adapted for the pneumatically derived envelopes to enable them to match the auscultatory readings.However,even after introduc-ing of a new level for the systolic characteristic ratio,the accuracy of estimation remains low due to large variability of envelopes. It should be emphasised that working with broad envelopes with high levels of detection(in other words,closer to the plateau)makes estimation more dependent on several disturbing factors(including variation in characteristic ratios)[7].According to the recommendations of the American Heart Asso-ciation,an occluding cuff should have a bladder length that is80% and a width that is at least40%of the limb circumference[27]. These recommendations have been applied for the upper arm as well as for wrist measurement.However,several theoretical and experimental studies[10,11]have demonstrated that the criteria recommended for the upper arm cannot simply be applied for the wrist because of the difference in anatomical structure of these two sites.Lu et al.[11]simulated the pressure transmission from a local pad-type square cuff to the radial artery and found that the optimum bladder sidelength required for accurate blood pressure measurement was between one-third and one-half of the wrist diameter provided that the cuff was placed over the site at which the radial artery crosses the most protuberant spot on the volar aspect of the distal end of the radius.An applied experimental pad-type cuff with a bladder of sidelength of27mm effectively pressurised the radial artery if the arterial volume oscillations were detected by an embedded photoelectric plethysmograph having components adjusted at a distance of about6mm.In experiments of the present study,a greater distance between photoelectric components was chosen—15mm.It is our opinion that even when the local cuff is adjusted precisely in relation to the artery,the photoelectric components and radial artery can change their positions during compression.This results in two drawbacks:first,the optical measurement is distorted and,second,the rigid photoelectrical components above or close to the artery affect the pressure transfer to the critical segment of the artery.A greater distance between the components is believed to reduce these dis-advantages noticeably.Due to a greater distance between the emitter and receiver,the sampling area of the photoplethysmograph in our experiments was expected to be wider than in the referred studies and,subsequently, a diameter of40mm was chosen for Cuff1.The group-averaged photoelectrically recorded envelopes in Fig.6indicate that there still exists a possibility to improve the effectiveness of occlusion of the artery by slightly increasing the area of compression of Cuff1 (compare the curves marked‘Cuff1,photo’and‘Cuff2,photo’in the region of negative pressures P).In the present study we demonstrated an advantage of pho-toplethysmographic vs.pneumoplethysmographic recording of oscillometric envelopes from the locally compressed radial arteries of young subjects.Location of peaks of oscillometric patterns was not observed.Further work is required based on aged population with arterial mechanics different from that in the present study and considering the brachial-to-radial bias.5.ConclusionsPhotoplethysmographically compared to pneumoplethysmo-graphically recorded pulse amplitude envelopesfitted better for the height-based oscillometric estimation of the radial blood pres-sure when a localised compression of the wrist was applied.The mean differences‘oscillometric estimate minus auscultatory refer-ence’and variability were considerably smaller for systolic as well as for diastolic pressure in the case of photo recording.In addition, photoplethysmographically registered data were less sensitive to alterations in the cuff size and characteristic ratios.AcknowledgmentsThis work was supported by Project SF0180148s08and Grant 7723from the Estonian Science Foundation.Conflicts of interestNone declared.References[1]Mauck G,Smith C,Geddes L,Bourland J.The meaning of the point of max-imum oscillations in cuff pressure in the indirect measurement of blood pressure—Part II.J Biomech Eng1980;102:28–33.。
