FREQUENCY ANALYSIS OF KOROTKOV BLOOD - PDF document

1 SAM-'TR-66-8 "FREQUENCY ANALYSIS OF KORO
SAM-'TR-66-8 "FREQUENCY ANALYSIS OF KOROTKOV BLOOD PRESSURE SOUNDS USING, THlE FOURIER TRANSFORM. .::, ~THOMAS IRALITI'M'IRK, M.S.E.h. JOiHN 1. '1',TZ, .8,E4X.E JAY W. FICKES, B.S. CL. E A K-I NG H 0 USE YIOR F A SCJo TIVIC AND TiMCUNMCAL INI6'O1UL ATION 'ýebiuaiy 1966 11SAF' School of Aerospace Medicine Aerospace Medical Division (AVSC) Brooks Air Force Base, Texas Qualified requesters may obtain copies of this report from I)I)C. Orders will be expedited if placed through the librarian or other person designated to request documents from D)I)C. When U. S. (Government drawings, mpecifications, or other data are used for any purpose other than a definitely related government procurement operation, the government thereby incurs no responsibility nor any obligation whatsoever; and the fact that the government aiay have formulated, furnished, or in any way supplied the said drawings, upecificationw, or other data is not to be regarded by Implication or otherwist-, as li

2 any manner licensing the holder or any o
any manner licensing the holder or any other person or corpora. tion, or conveying any rights or permission to manufacture, use, or sell any patented Invention that may in any way be related thereto. l)istribution of this document Is unlimited. FREQUENCY ANALYSIS OF KOROTKOV BLOOD PRESSURE SOUNDS USING THE FOURIER TRANSFORM THOMAS iRAIUT1I4KUS. M.S.E.E. JOHN F. FEILTZ, H.S.E.E, JAY W. FIlKES, B.S. i ) _ --~~-". 4UUVUIA! I, -~ ..w -' . FOREWORD Thi. report was prepared at the Systems Research Laboratories, Inc., Dayton, Ohio, under Air Force contract AF 41(609)..2753 and task No. 793003, The report was submitted for publication on 29 November 1965. The data-collection phase of this study represents a cooperative effort between the USAF School of Aerospace Medicine and Systems Research Laboratories, Ine. The authors wish to acknowledge the contributions made by Dv, Sidney Leverett, Major Clay Gammon, and Charles Martin of the USAF School of Aerospace Medicine. They are

3 also grateful for use of the facilities
also grateful for use of the facilities at the USAF School of Aerospace Medi- vine and to CG&ptain John Alexander for securing', permission for use of the digital computer and data display facilities of tne WJSAir Aerospace Medical Research Labora- tories. This report has boon reviewed and is apprwoved. HAROLD V. ELL17 O Colonel, MC, USAF' Commnander -i t- A BSTIRACT The purpose of this investigation was to determine the frequency content of the sound signals (Korotkov sounds) obtained from the microphone located in the arm cuff of an automatic blood pressure measuring instrument. Korotkov sound recordings were made for five subjects in five experimental situations: rest, postexercise, passive tilting, centrifuge rides, and flights in NF-100 aircraft. The frequency analysis was performed by using a digital computer to obtain the Fourier transforms of the sound signals. The Fourier t.,ansforms were displayed on the computer oscilloscope and photographed. The photograph

4 s were then arranged in a number of rect
s were then arranged in a number of rectangular arrays for convenient comparison of the frequency content of the Korotkov sounds as related to the several types of Korotkov sounds, experimental situations, and subjects, Initial study of the 240 average Fourier transforms contained in these arrays indi- cates no readily observable common characteristics except that most of the sound energy is almost always located below 50 cps. iii t A, FREQUENCY ANALYSIS OF KOROTKOV BLOOD PRESSURE SOUNDS USING THE FOURIER TRANSFORM S I. INTRODUCTION and diastolic pressures, respectively. When ' the technic is employed in an aircraft, both the Data collection and computer technics were Korotkov sound signal and a signal proportion- employed in analyzing the Korotkov sounds of al to the pressure in the arm cuff are five subjects under several noise and stress telemetered to a ground station, conditions. An underlying purpose of this investigation was to determine what sound A necessary

5 requirement for the successful character
requirement for the successful characteristics, if any, were common to all use of this technic for measuring blood pres- wsubjects under all the noise and stress con- sure is that the sounds used for initiating cuff- ditions imposed and, by so doing, to help pressure readings are truly Kor3tkov sounds (develop an optimum design for an electronic rather than extraneous noise. Two factors, device by which Korotkov sounds can be recog- however, make accurate recognition difficult. nized. The first is environmental noise, which is especially troublesome under aircraft flight  IA technic has been developed for auto- conditiwns, and the second is the changing matically measuring blood pressure in a noisy nature of the Korotkov sounds themselves, A environment such as that which occurs in Kortkov sound, like many physiologic phenom- S flight. This technic employs an arm. pressure ena, is typified by a complex waveform. This cfhuff, which is automatically Inflated and de- wa

6 veform may change in period, amplitude,
veform may change in period, amplitude, S flated in a preprogramed manner. A sensitive and spectral content from individual to in- midronhanprprograted manner. Ah senfsioetiet dividual and for the same individual when the pulse sounds (called Korotkov sounds) which occur as the blood pulses through the Ware and Kahn (1) reported a device for brachial artery. The cuff pressure is de- measuring blood pressure as described above. creased from above systolic pressure to below By using band-pass filters with center fre- diastolic pressure at a rate of 5 mm./sec. The quencles at 40 cps and 150 cps they found that duration of this deflation period is appioAi- valid Korotkov sounds could be recognized by , mately 30 seconds and the period between suc- an electronic coincidence technic, A similar cessive inflations Is approximately 1 minute. device developed by Systems Research Labora- tori(s in 1964 also employs the coincidence Korotkov sounds are complex sound pulses technIi

7 c but places the filter center frequenci
c but places the filter center frequencies "which occur at a repetition rate equal to the at 40 cps a1d 100 ups. pulse rate during the time the cuff pressure is less than the systolic pressure and greater A brief survey is given of papers dealing than the diastolic pressure. The occurrence of with the qualitative characteristics of Korot- each Korotkov sound causes the cuff pressure kov sounds. to be read and recorded. The first and last cuff pressures recorded, corresponding to the As at qualitative description, the work of first and last. Korotkov sounds, are the systolic Collins et. al. (2) is noteworthy. Collins divided, J1 the train of Korotkov sound pulses which occur quencies in the 40 to 55 cps, range. The "between systole and diastole into 4 audible frequency and intensity (lecrease somewhat 4 phases and a fifth silent phase. As the pres- during phase 2. With the arrival of phase 3 sure in the cuff was released, just enough to sounds, the frequency aild inten

8 sity increase permit a jet of blood to p
sity increase permit a jet of blood to pass the restrictive quite rapidly. The principal frequencies lie in point, a "sharp light thud" was detected. This the 60 to 70 cps range and the intensity is sound could be heard over a range of about about twice that observed at the transition V 10 mm. Hg and was termed a phase 1 sound, from phase 1 to phase 2. With the occurrence As the pressure in the cuff continued to de- of phase 4 sounds, the amplitude and frequency crease, the sounds changed in quality and drop rapidly. The ampiitu(de goes to zero or intensity. Collins described these phase 2 near zero and the frequency components drop sounds as soft but rather inconsistent murmurs to 25 to 35 cps. It must be noted, however, which persisted over a pressure range of that the frequency response of the recording 10 mm. Hg. Loud, long, and clear sounds instrument utilized by Geddes fell cff ap- characterized phase 3. This type of sound preciably above 70 cps. persisted for 1

9 5 to 22 mm. Hg in the normal individual.
5 to 22 mm. Hg in the normal individual. Muffled and dull sounds were ob- In addition to revealing a ldivergence of served during phase 4, which wits associated opinions as to the spectral content of Kortkov with a pressure spread of 11 rmm. 11g. The sounds, this research has revealed no previous largest change in the intensity of the Korotkov studies of Korotkov sounds under condit,:ions sounds occurred at the transition from phase 3 ot high environmental noise and subject stress. to phase 4. H!. ANALYTIC METHODS Many attempts have been made to quantify characteristics of Korotkov sounds. One of The Fourier transform is the analytic t(,ol the first concerted attempts (1914) was made which was eml)hoyed to perform thle frequency by Hooker and Southworth (3), who used a analysis of the Korotkov sounds. The power- carbon-button microphone and a Lippman capil- spectrum technic is probably the most widely lary electrometer. They measured 15 cps its applied sophisticated s

10 ignal analysis technic the predominant f
ignal analysis technic the predominant frequency. In 1925, Braim- employed by medical researchers and wits era- well and Hickson (4) used a Frank capsule ployed at the outset of this study. It wits sooln and recorded sounds under the cuff. They abandoned, however, because t'le nature of the recorded frequencies as high as 100 to 150 cps, Korotkov sound sigials made the technic UnL- but they did not detect an appreciable dif- applicable. ference in the 4 phases. The following year, I{orns (5) used a membrane manometer and The power-spectrum approach is profitably observed frequencies in the range of 40 to applied when the signal consists of recurring 256 cps. In 1943, Groedel and Miller (6) used p~henomena Plus additive random noise. The a heart-sound pickup with a photographic autocorrelation step in the power-spectrum recording instrument to find prominent fre- technic then. serves to negate the random com- quencies in the range of 45 to 60 cps. The ponents of the si

11 gnal. The Fourier transform following ye
gnal. The Fourier transform following year, Rappaport and Lutisada (7) ()f the resultant autocor:.elation function yields utilized a pressure.equalized microphone its a the relative power density of the periodic com- pickup and a photographic ECG' instrument its l)onients of the original signal. The noise-abate- a recorder. Their findings show 25 to 55 cps ment feature of this technic nial(es it. attractive ias the predominant. frequency range. for the purpose 0Of our study; Korotkov sounds are not amenable to this method, however, be- One of the more currenit (luantifiAive cause they are transient and changeable in descriptions of the Korotkov sounds was pre- fori and (ldu not, nocessarily recur at a fixed selited in 1963 by Geddes et al. (8). According rate. This variation in rate was caused by the to Geddes, phase I soUt11(1s have principal fre- sligrht hut significant short-term changes in 2 It arterial pulse rate and was thus the major 2. A new analysis approach

12 may be adopt- reason for rejecting the p
may be adopt- reason for rejecting the power-spectrum ap- ed which would rely on the summation of the proach. Unless the sounds are stationary, the Fourier transforms of many Korotkov sounds autocorrelation step cannot reveal any tran- to cause any low-amplitude, high-frequency sient period components (such as damped sound components to rise above the level of the sinumoidal vibrations) that might occur in random noise. At least one investigation (12) several successive Korotkov sounds, involved an approach similar to the latter choice to obtain information about the high- In the approach employed in this study, the frequency content of Korotkov sounds. When autocorrelation step was eliminated and the the objective is to reveal the significant fre- Fourier transform was applied directly to in- quency similarities, however, or dissimilarities dividual Korctkov sounds. This approach ex- which occur for several subjects in several poses the frequency distribution of the

13 sound experimental situations, more valu
sound experimental situations, more valuable informa- energy of individual sounds without regard to tion can be derived by analyzing individual stationary sounds or presumed similarities be- Korotkov sounds and concentrating this anal- tween adjacent sounds. This makes it pos- ysis on the narrow frequency range containing sible, for instance, to compare the frequency the major portion of the sound energy. distributions of the 4 Korotkov sound phases, phase 1 soum(ls of the same subject obtained during ,ee'al different situations, and dif- Il. EQUIPMENT AND METHODS ferent subjects obtained during the same situa- tion. Experimental situations Direct application of the Fourier transform has one (lrawback--.-it includes no provision for Each of five volunteers was subjected to distinguishing the signal from noise. This is five different experimental situations: rest, only a minor problemn in experimental situa- exericse, passive tilting, centrifuge rides, and tions which

14 produce sound-signal levels that flights
produce sound-signal levels that flights in NF-100 aircraft. In each instance, are high compared to the baseline noise level, pressure-cuff inflations were performed and III situations such as centrifuge rides and tape recordings were made of cLIff pressure, flights in NF-100 aircraft, however, the Korotkov sounds, voice, ECG, respiration, and potentially high-noise level can make dis- G level (when appropriate). crimination *between signal and noise quite difficult. This difficulty was avoided for the Si.lce the "rest" and "exercise" sitiiatiO)is most part by limiting the range of the fre- were pursued in close succession, the cuff in- quency analysis to a maximum of 160 cps. flation profiles for both experimental situa- Since most of Ohe energy in ii Korotkov sound tions are shown joined (fig. la). During the can be expected to occur in the low au(.io- "rest" situation the pressure cuff was inflated frequency range, restricting the frequvncy twenty times for each su

15 bject. The :first. to analysis to the re
bject. The :first. to analysis to the region where the signal-to-noise inflations wore performed with the subject ratio is highest maximizes the probability that seated and relaxed; the last ten were per- the frequency distribution obtained is largely forIme(d with the subject seated bUt t ensillg the Comi)sed of sound energy. Urm wearing the cuff. During the "exercise" situation each subject exercised three thines Extending the frequency cutoff to the by runiiing in place foe 20 secon(ls. Cuff in- higher audio range, say 2000 cps, preseints two flat ions were noof attempted during the exercise choices: periods bek-ause motion artifact, was high; instead, they were performed between exercise .1. A large band of frequency information periods with the subject seated. Three conl - may be collecte(d in which noise and signal are secuit ive inflat.ions folhowod each exercise indist.ingu ishable, period. -3l Exercise, running ' I in place . ,,,Ret., am Rest,. arm! !!-I1-

16 1! ;- ~~relaxed | tensud [ 0 0 t m e 2
1! ;- ~~relaxed | tensud [ 0 0 t m e 21 @)2.3 2462z6 272&93 Cuff inflation number Figure la. Rest/exercise profile. Head elevated I : -- ,I Horizontal .0 TimeO ~ Head lovered Cuff inflation number Figure lb. Passive tilting profile. mTim I I I II I 1 I Cuff inflation number Figure Ic. Centrifuge profile. I-IA -m----~'Time ,-WEIA1 iD~~ 9 0 12 1 CU f inflation number Figure Id. NF-1O0 flight profile. FIGURE 1 Maneuver profiles employed in the four experinntanul Hit t.ioti durilg .whichl cuff!influt~itonni were performnd and Korotkov sounds recorded. Inflation ,ruimuers circuntseribed with Hyinbols are those that were sMelected an representative of each imaneuver. IY'ourir tranii/orins obtained fromn soundH with similar sHymbola wevre averaged. "4 The experimental profile for the "passive of an automatic blood pressure measuring in- tilting" situation is shown ik figure lb. The strument similar to that demcribed in ref- first four cuff iiiflations were performed wi

17 th erence 9. Proper use of the instrumen
th erence 9. Proper use of the instrument the subject lying horizontidly on a tilt table, requires that care be taken when placing the Then the subject was tilted head up 70", and pressure cuff and microphone. The procPdure five inflations were performed. Subsequently, involved locating the microphone over the the subject was returned to horizontal for two brachial artery about 3 inches above the elbow inflations before being tilted head (town 10" for and locating the pressure cuff over the two inflations. The seventeen-inflation profile microphone s;o that the microphone was located wis completed witli five more inflations in about one fourth the cuff width from the the horizontil position. bottom of the (,uff. Then, with the subject's arm relaxed, the cuff tension was adjusted so The experimental profile for the centrifuge that small sound signals were detected when rides (fig. 1c) consisted `" three consecutive the cuff pressure was above systolic. Subjects runs at

18 2,5 G, :.0 G, ant( 3.5 G, respectively,
2,5 G, :.0 G, ant( 3.5 G, respectively, with large biceps were instrunted to minimize G's were applied in the *-I-G, (longitudinal) urm flexure beLcause al increase in bicep direction, Three inflations were performed at diameter might. push the cuff d(own from its each G level, and one inflation wits performed optinimum locattion, at the 1 G level while the subject was heing rested between runs, Each run lasted ap- For all the experimental situations except proximately 3 minutes, Three inflationm prior the lntrifuge ides te daital signals were von- to the first centztfuge ridoe and t'wo( inflations h n'igerdst dtsinswreo- tolhe winthe last ride completed the thIrteen- (ltioned, toloem(tered, and recorded In the fol- following prof i de lowing manmer. They were amplified by Mennen (1reatbach (model 621 A) amplifiers alnd The experimental l)roffle for flights In NF. converted to FM by Bendix model TOE-305 tOO aircraft is shown in figure 1A. Following subtea'tier' oscil

19 lators, ,these FM sigals then three infl
lators, ,these FM sigals then three inflations during level flight (1 G), a were Ilultplexe(l ttid trailsirlitted on a c1ri'lei0 maneuver sequlence of l 0, zero (, 1 G was freoquency of 232,9 ,Mc,/se(, After d(mo(dlih. pursued three tim(es in succossionh One tn'h- t Ion at the telwrnetry ground station, the data lion was performed alt each G level in the siglads were i'ecoI'(lld In the FM mode o(n a ,W(1uence, The -3 (3 levels were obtained in Sigam) niomoel 4700 tape rocorder i elittOd at spiral diving turns lasting IipJ)roxillittely WV¾, I i./sec. 'T'ils recordo, has al FM centel' 1. minute each ; zero G levels wvere obtained by frequency of 6,75 kc./sec, and a double band- the plarabolic arc method and last(d alpproxl- width of 1250 eps, Since this recorder can mately one-half minute each, Two in'flht,ions i'e(cor' and I)laY back s(Iultaneously, lproper during level flight, completed the thIrtoeen. iecQo)ilig wats ve'if(ed by obsoi'ing 1the} dat a Inflation profile

20 , wri tUtn 1ii a lnautanously on ati Off
, wri tUtn 1ii a lnautanously on ati Offnro' strip chart, POcordet', T110he UI )l)0 i3 (1B frequency otf tli Data Collection Ko'otkov sound (h1tiliel Was lII1itted to 160 cps by loading ti h Men1n1n (krca tlach almpli )Ile' Although multiple data signals were tape with a 0.1 Id, caliipacItor, Also, tihe subcnrrliea' recorde(I (cuff l.)ressUre, F.o1otkov SOuiLIIS, (si min' I nla1tor for' the K)'rot(kov So)und(1 channel voice, resp)iration, and ( level whell appro- wits prov'lded with at I'llter to 11i1mit tihe Upl)(w' priate) ; only the Kor'otkov sounds were to bit :i 1 H freq ucaw toV 1(),0 ( ')S. analyzed in this stludy, The cuff pressur'e and voice signals were employed as convenient Dlltt from the con trifuge experiments were "keys" for identifying the location of the Korotkov sound( (latea on the recordings, not. telemeterd ; lnstedl, they were tranlsniitt ed via slip rings 1i)1d signa leadIs. 'ihe (l1111a The Korot.kov sound signals %oere obtained signals were am

21 pl)liflied by Taber m,)(del 2026-4 from
pl)liflied by Taber m,)(del 2026-4 from a imicrophone located in the )ressturce cuff ampl)lilfiers HIM reTC0(ded il tihe FM mnode on ai S ment Corp.) prograrned to collect, the digit ized Korotkov sound( signatis. The input to the Philbrick amplifier w;as In.. cou~pled (1.5 cps low-frequency cutoff) to eliminate any di.e. comiponent from the t ape-recor(Iet outtplut. The (digital compu~ter wits promramedl to co), lect 19-second segments of digitized SOtIl~ld signals at at rite of 400 samples per second onl m118"u111l command fr-om the operator. (Th i s nu~mber was selected because it is near the minimu~m ratte reqluiredl to resolve freqUenIcy compllonen3ts ats high its 160 cps-h hi-f- (lLtflICY cutoff Of the i'CcordC( sound~ (latai, The samp~led interval of 1 9 secondsl -wats chosen its FIGURE 2that wats sufficient time for the cuff p~ressure FIGURI~ 2to (lecrease from above systolic to below Colmputer outhode-rull tusbe displa)/1 of typiectl M.l diastolic.) By monito

22 ring the Cttff-pIe(SStii' Rodfl som~~t f
ring the Cttff-pIe(SStii' Rodfl som~~t f ~flOtOV oud V'(Ord R'OO(I channel with an oscilloscope 'onlnectedl to the starPts at upper /tilt and is iont'intwunly d15playe(d tLin eodrotuteo~rao 0Ct( o' u R,5-sc~ond uoom~cn tw pr lhine. Al ova bb'iitIeia shown~~~ inpst o o~ xrueti (Ij n. ind'ividual Nood of the som id-chain tel analog record which coil- taltied Korotkov 5OU L~IM5. W"hOU thle 0SClllOScop ilndilcte~l thait a It C Ch~ft 1fltoul jIrOfl Wits bV- Penio prtabe i~3trmonatio reor~iV ~ glimui jg (fig. .) , thle operalti r (lopressod a Perted prat l 7~imitr/ en.theio FM center opt- SWItch Oil thle comjIul~lr CouISOICle ad the 1`ol- erited tt 1711-n./se, Te IM coter re- lowing l9.*secotid segn-10ti. of thle sounld-channiel quency of this tape recorder is 3.37 lw,/soc.. and. rocoi'(l Wits ii Ittonni tICIIIily collect00d II (a d1igi i the double bandwidth is 625 eps. Prior to 1e-0- fra n trdol111I0.vti cordilng, the Korotkov somid s4ignal wats flltMend Iomtaa trdonniue.

23 I 10 by an SRL model 221.A Vap) filter o
I 10 by an SRL model 221.A Vap) filter operated inl Thel socouid(l dtiti-hlitlkidillg 14id) ext racted the low-p~ass mnode whaor the band-pass filter individual Iot'ot l(o\'oud SO I (1 fom the 1 9-second is 5 to 640 ('ps. At thle complotioui of ouc record conitaItilt ung JIIl the ,Korot kov sou tuds contriftige (Jxpek'flnmt the taple recordedl (hita V Ici0curd(ltrItgt n ltoTht tis wits wore vla,?ed back Into an Offtor strli)-chart; licoipl ished by fCirst dis.phylayig the I 9-secoild record%.to verify thle quiality of thle recordinigs, dgtiel51 tlt0't( t ~t 'ttit t' cathlode-ray tuibe ( fig.. 2). 1'I'lti tho oporilt ot Datil handling visually becat ed 4 sm~l ids that. wvere rupreoseti .i t ivu of' Koi'ot~kov son ads ot' phasos 1, '2, 3, midu Sover-al dattal-handlinig steps wrom reqJItU 'If V4J-rspect I V0lY, atic I extracted 01"'.1 fr'oml tihe to redunco thle tavl).t'Uecoildod Hotil 1 datat to at trecordl byN M)t' til posi tig ai mlovable ed it litie formIl Stit Itb 1

24 1 01' o III)IIINIH1 itoti(f the F01.11-i
1 01' o III)IIINIH1 itoti(f the F01.11-i01 befO Core thoiel augOf 1t C h son idI(. Thei COnII- tranlsform by it digi tad conmpu ter. 'rile first plit or I)1'UM'Pltl I hoti stilt ablyv labeledot ech Illii- step i nvolve~d Conivertinag the Ilittuil(g hte1.01)!rcot'(- (l II VIti ,oo (IovI s0-H on tOIId I~tic 0tiSCrIhed It, (itt higs to at digitill formallt. T[he facilitlo used 15 (1 01 digittub iagtuo Ic 111) ill a fotntit cottI'ltl pt tiNo to accomp~lish this stolu I inuded : (11.) an Am- Wit htilhe IBJM 709)4 comptii i ter whii ch \\its to( 1'POX FR 1200 tape recot'deu to pla'y back thle cI'l Ic iII I (v tile F trleot rio trlis iot'nts, antalog tapes, (2) at Phi Ibrick oj~eratioiial am11- 011lfier to act as at buffer stage between the Illi addition to thle editInig technlic, which Miecoz'der and the A/Il (convort~et, (3) at Rauy- re(Itice(. the ttttnubol of Korot~kov so.ittidls to theoit AD-5OA anitlog-to-dIigittll cotiverter, 1111(1 oily 4 from citch I n flat ion, til

25 he mass ofC record- (4) at I' ) -1 d :gi
he mass ofC record- (4) at I' ) -1 d :git al coiniputar (Di.)gi tal 11, Ij- 01 (ltit -its WiLS P(1cod fit t'ther by selecting 6 representative inflations from the multi-Infla- tion profiles followed in the several experi- mental situations. Inflations chosen as representative of each of the maneuvers which occurred in each experimental situation are identified in the experimental profiles (fig. 1). Following these two data handling steps, which enabled selection and preparation of the Korotkov sounds, the Fourier transforms of each sound were cnomputed by an IBM 7094 computer. 75 cps Sinusoidal Signal Results disxplay format A digital tap)e containing the Fourier trans- form depiction of the frequency distribution of each Korotkov sound wits returned from the 7094 computer facility to the PDP-1 computer facility so that. the transforms could be dis- played on the computer cathode-ray tube and photographed, Figure 3 indicates the graphic form of this display technic by

26 illustrating , how the Fourier transfo
illustrating , how the Fourier transform describes the fre. queny distribution of it sinusoidal signal. IV. tH 141 [i'rs i II 0 40 80 IZO 160 The accumulation of analog Korotkov sound data wits revor'(ied (luring 3(65 lressuro-cuff Frciuent-y tops) inflations, (.Cornimssinig five subjects, four Ox- i)erimential situations for onch subject, and FlIURIUi¢ O three maneuverts In erch situmtion, This large/ t111( oLIlt. of d a s iEt Iwiits' made mov m alUgeablo by fhv 'riqueelt dts ttitn ol t Haratko, mound. Upper soloctIng 4 sounids, reilrsomtitattvI e of the t.ri,,v IN t. di,,di 7#e t'7A e Ip N li Dt1141oid l N11/111t0 N11150,' trtet' 4 Korotkov sound p)hases, from the 30-odd Ii tihe, mli itHll U, frvotltt,/ p/'01 for Ihis HI/pall soun 1ds which OccCurP'Od d uLIrLng each inflattion, tibtalowd t /ith th1, ''furihr trtlit mfo / l, IlIh trutl(,fH 4IIt, The (hat Wer'e (doe'rvitsed futri}thet' by choosing pholograplr l III ./ the JlNli'r f~lh hrlt t I/, i ll, the Korotlhiv

27 sounds from only 2 to 4 r tpre- s.iitat
sounds from only 2 to 4 r tpre- s.iitatlyVe inflations from the groups of 6 to S i11lat loimis perf'orme(I o each subLect. dtiri 'ng This tnl nl)tur of nornnIiaked l0it't rh trants- eatrh nuetuiver, In this manner, the hida were 14,twmm thten wits reduced from 700 It) 2,)t 'v reducedI to 4 sMoun(ds MCa)ch flrom1 175 in flhtios -- averaging the Fourier tl'razsfornis obt elidnld 700 individual Korot(kov sounds. The Fourier t'rom rpelt it )us IH lat it.,-e~g., the thiree transfornms of these 700 son nds were then i)haLso I Ki)rotkov sonds f'm- smubiect, I (Wiring Com)LItued And ind ivhhtdilly normalized by its- the 3.5 ( ImaneOulver oin tOhn Uo tt i UfUge, The signing the lhrgest frequency conmponent In cticishimi to averi'ge the triamsfornis of' sini haly each transform the iarbit rary magnitude of genlerated Kurotl ov suttic(is rel)rselsttit c )ll . 10.0, ipromiise, betw('lw two diesm irale but c miflic t-ing 7 4v= ends. On the one hand, it was desirable to retain Ind

28 ividual transforms of every Korotkov sou
ividual transforms of every Korotkov sound to reveal any variations which might have occurred within the groups of like soundsm obtained from any one subject In any one situation. On the other hand, the total ac- cumulation of frequency-distribution data had to be limited to a reasonable amount If human11 4 analysts were to be able to read it and derive A meaningful Information from it. It wats also desirable to aver'age transforms obtained under Identical experimental conditions to negate the 0 Influence of random noise. Z Attempts were not made to decrease the " number of Fourier transforms below 240, be- cause to do so would require destroying one or more of the three par-ameters the datat were Intendod to reflect-Korotkov tiound phasem, subjects, arid experimental maneuvers. For 4 sound phiweg, five subjects, and twelve MIf. ferent maneuvers (three In each of Lour ex- perimental situations), the minimum number of Fourier transforms Is, necemssarily, 4 x5 X 12 =240

29 . Rectingular arrays of photographs of t
. Rectingular arrays of photographs of the40 810 16 Individual Fourier transiforms were constriietwl to dimplay the datat in it mariner which would rqec(ps facilitate assessment of the effects of oach of l~,j~~ cs the three experimental p~arameters, The 240 hotoraphs have been arranged twice, ii~ui' Sc) that at comparison ofV 4 Korotkov sounId ~H~ ~~~rI 111 rH 1 IKVDlV phases its they appeared for each experimental Ho)l Did (upe Di)Da)'7 l') (In OIiIN jol111 eiij'blultp "l(81i'' ( lon'ii maneuver (vortical columns of photographs) t '-I'/!I- Te low m , o w1v 1 'l IIvempined ji/l'II I'uil'l (I I'llP and the effects the throe mwieu vuz's hild 011 DD 111" ()fDtlDdf/Nitdv iDii 'JINlt'U b1'11(11- each of the 4 Kuroikov Mound phases (horizontal 111111h rows of photograiphs), `Vhe third experimental parameter, sub)ject mnumber, is the '"running"' HoundI-lihas nutnII10'n takesC o11 the role0 Of the parameter (page-to-page) for the twenty l)).gos pitge-to-pageo parameter. requi

30 red to displaty fWigure 5 t~o '24. The s
red to displaty fWigure 5 t~o '24. The samie niagn11tMI uevCIrsus fruqtwn01CY in-f Trhe second form of arranging the 240 Four- formation coitainlied lin the 240 p~hotographs icr transform lphotogruphs is shown lit figures waIs also0 11)'0i10d III ti-ahulili' fot'() to Peormilt 21 to 40. This arrangement makes i teasier to more 0exactin g exaitii io of.. I h da ta I" h' compare different, subjects during the same 240 pages of' computer print-out. reqjuiredl Nvoro lexperimental maneuver. Trhe transform photo.- 1CII( MI 411 Incldl as a1. 1epratly od appcfl1101iX. graphs for the five subjects aire arranged fin vertical columns for each of the three muneu- The datat obtained during111 the "paSSiVII tilt - vers. In the sixteen pages required for thiti lug" an,1d "rest,/0xeCIseI" OXl)eri-1ilent111 MitUa- 0dternate form of datat display, the Korotkov tloIIs wOre leas1t affected by aidditive rand1om 8 noise, Although an exacting determination of sounds the energy tends to be m

31 ore broadly how much noise was present c
ore broadly how much noise was present could not be made, distributed in the 0 to 50 cps range. an indication was obtained by applying the Fourier transform to sections of "baseline"- The prominent frequencies noted in these the microphone output (luring the quiescent conclusions do not closely agree with those puriod folowing a cuff Inflation. Figure 4 found by other investigators. Most others contains unnormalized Fourier transforms of a found some prominent frequencies above 50 cps plhase '1 Korot~kov sounLd and the baseline fol- (refer to Introduction). The chief reason for lowing the inflation during which the sound this discrepancy constitutes a significant fea. occurred, This Iprticular example, taken from ture of the computational methods used in this the d(ata obliained during "passive tilting," Is study. typical of most of the data obtained during "passive tilting" and "rest/exercise." It will Prominent frequency components higher be noted that. for subjects

32 2 and 5, however, than about 50 cps are
2 and 5, however, than about 50 cps are not discernible in the (luring "rest/exercise" large 60-cycle compo. Fourier transform photographs because the nents were present (and also the 120 cps amplitudes of higher components are so small ha, moenlo), that they are indistinguishable from random noise, If each photograph was an average of Data from the NF-100 flights are quite a large number of sounds, however, rather similhr in quality to that obtained during pas. than Just 2 to 4 sounds, the frequency distribu. sive tilting (with the exception of the data for tion of the noise would become "flatter," and subject 5), but the centrifuge data are generally low-level signal frequencies would become more p)oorOr III (lulility, For this experimental situa- identifiable, To confirm this hypothesis, a tion only the data from subject 5 appeal, to be small statistical study was performed on a relatively unblemished by noise, select group of Fourier transforms from the "passive

33 tilting" situation, Although the low. fr
tilting" situation, Although the low. frequency components were still most prom. V. CONCIATMIONS inent, noticeable amounts of sound energy were located at 46 cps, 104 ,,ps, and 150 cps, These Although caretfu stucly will be reouh'ed to frequiencies coincide closely with the three fre. i'uvoil all the c~ha,.'ctwtristics of this Korotkov (juen('es which Ware and Kahn (1) employed sound data (f 'Is, 5 to 40), Initial study h111 In thoir Indirect blood pressure Instrument- p'odlucod(l soverlal gWonerald conchluslols: namely, 40 cps, 00 eps, and 150 cps. 1, Except f'or the (liMshytilhiltles caused by Hence, the data in this study can be orn. dlfr'erenlos in the noime environment, the fre. ployed to indicate which of the Korotkov sound (ulen'y (listribution for all the experimental characteristics are discernible in Individual situations are quite similar, sounds, By additional statistical manipula. tions, the data may also indicate what general 2, No prominent differences

34 are con- characteristics are discernible
are con- characteristics are discernible by averaging sistently apllat'e)t when comn)arlsons are madce large groups of sounds, For it study designed imi)ong the several experimental rynaneuveO's or to determine diagnostic feasibility, the wide. amn)ong tho several subjocts. band characteristics attainable by averaging would be more suitable, When reliable recogni. :, Most. of the sound energy Is generally tion of the occurrence of Korotkov sounds is located below 50 cps with the most prominent the objective, however, data which express the l)tjal(s occulrrihi below 20 eps. characteristics of individual sounds are more valuable. Any device designed to recognize 4. For phases 1. and 4 sounds the sound Korotkov sounds must do so on an individual energy is often heavily localized below 20 cps, sound, real time basis; the device cannot re. whereas for plhase 2 and, especially, phase 3 spond to the average components of a sequence 9 ..................4 W I Rest, arm relaxed

35 Rest, arm tensed After exercise 40 8 10
Rest, arm tensed After exercise 40 8 1016 0 80 10 160 2 40 1O 0 1 GQ FREQUENCY (CPS) qetuej da1iu Iru tff/t'H~t~tl itti/i ll f~i$I l~ I hltit; /tl~.14l ii M 11s e//i01ntt1ti'rNu 10 Rest, arm relaxed Rest, arm tensed After exercise INN -ý I-- IpJi "" Rest, arm relaxed Rest, arm tensed After exercise Ij. 2 40 10 16b I 20 O 1 1 2 40 NO If K OW FREQUENCY (CPS) 10 101 U It 0 7 Ntorlflau id uiv'e'e l o -?'ot ier' trt'(lfn sltN e/ Iof tKrvtkove m wdI lH of subJeoet 8 obti.hl-td duhri/n th/i'e,', r ft' / (piell tIT lotI .-I "u- il ' O i-N 12 I UP Rest, arm relaxed Rest, arm tensed After exercise 1! 3 4 I-- FREQUENCY (CPS) E!€IV(tl'tli if 1'1(1If I('1*1' II o1, ~ *g b 77U~~o Rest artAfter exercise '-st, arm relaxed1 -tiar tensedc IIY 6 so0 It 0b2 960 tO o 1602 d f 6 FREQUENCY (CPS) Nomlle il vurago P'ourie',* tranafurtma of Korot kov No umbe of NWubJeel 5 o btain td during thrPee rent1/ 14 iovl m('l voN .no Huiot1Head 70 0up Head 100 down LiJ �~ -- 2 4 80 12 02 40 60

36 dO1606S 4086D 40 40 FREQUENCY (CPS) Nor
dO1606S 4086D 40 40 FREQUENCY (CPS) Normalized averape. Ii'ou~r'h' tansforinn of Korolkoi' mou-IsS of Nub jeet 1 obtainhed during t/uc-v tilt-tab/ti man~eu'Ierm, 15W Horizontal Head 700 up Head 100 down ! -' 7 2408020160240 80 120 160 2 40 0 120 160 FREQUENCY (CPS) FIGURE 11 Normalized ave~rage Fourier traftsorma of Korotkov Hounds of subject 2 obtained during three tilt.tablv r,:aneuvera. 16 Hori zontal Head 70 up Head 100 down 0. -- I- N a 0? 2 40 80 120 160 2 40 80 20 160 2 4 60 120 160 FREQUENCY (CPS) FIGURE4 12 Normalize'd ai'grage Fouri'r t'an.R, fortna of hKorotkfot sounds o' RUb joot * obtczinod during thiroJ# tjilt..ta, h'. .mane.E ,,,'er . 17 p , J Horizontal Head 700 up I-lead 100 down ii--- 2 40 60 120 160'2 40 60 120 146O-24l w 01C F~REQUENCY (CPS) 'Pad IH! * 4 p ~ ~ -, Horizontal Head 70 upHa10 0 down 80 ~ ~ u 2ea 16 019 S 0 down40ti tio I G 2. 5G 35G 4,Y II ii I' II 2 4080 120 160 2 40 0160 2 406012 160 FREQUENCY (CPS) F,"I (" U It E,: .15r i-ai ud al

37 wo-o/c FoueIrier frenNforuim of Ko'rolkn
wo-o/c FoueIrier frenNforuim of Ko'rolkni' ,.tuvidRii f ,mubjecut I oblubtedi~g durini fthree 20 ... ....... ..,, ........ .. L 102. 5Q 3.50 1 3 ~ N 22 i o lb2 40 68d *4 0 12 60 d 160 FREOUENCY (CPS) FIGURE I10 Normnalized average Fourier erans/orma of Korotkov sounda of subject I obtainqed during throe centrifuge maneuvers. 21 -- 4I IG2. 5G 3. 5G I-- FREQUENCY (CPS)1 VI(LtRN4 17 No~rm~nalized noitirttgn Fourier tru7Plflariti of IKorotkov vo~uInd, ofI mubjqeOt -I oblbted dwinqe ltfljg 22 I0 G. 3.5SG 2 40O80120Ot02 40 S0 0 160 24090g10 160 FREQUENCY (CPS) NorttaliAt'd at'srage Pocurlor 1.-atannortit of Iuioukthm voundn of Nub jeot 4 obtaiinod during thrve 23 =3ommp mo w jw 2 40 80 1 1W 2 40 00 120 160 2 9 FREQUENCY (CPS) NVormtalized avi'ravo P'oirior tratlfut/oIa o/~Uti 1'lkn mouandii o/ Hubjef't 6 oittini~ud durbip I'hrove 24 * 4. ~ ~ .,. * * ... foi 2 00120 1602 40 80 120 160O2 40690120 160 FREQUENCY (CPS) FIG1URE 20 Nor-onalirv'd oiurage Pourlotr trano~forms of Kor

38 olkot, moumtde of atsbjoet I oblaitiod d
olkot, moumtde of atsbjoet I oblaitiod duritip tlreog Ilig/t nimmeistuverm hi, N'- 100 aircraft. 2LS --,OUR 101 280 1 20 16 4 0 12 6 I---NC (PS FI tM2 N rai ze v-u ore rnfrwo oolm o umo ujo balddrn he fipt i--ntvrmi h.oo(rwt 4.-- 26I OG IG 3G 0 0C4 D ,J w �0 I-2 z- j , w -! CL 2 40 8'0 120 160 '2 .40 80 120 160'2 0 0 120 160 FREQUENCY (CPS) FIGURE 22 Nor'malize.d average; F ourier; tra~nnfortni of Korotkov sounds of subject .1 obtained durilng three,, flight puaneuveriv in NF-100 air'craft. 27 w! 2408 1201 0240 80121602 40802016 FREQUENY (CPS FIGU1 2 Norraiicd v~rge~~uur~r ranfora o Kootko ~ondaof ubjct 1 obaind drin tlreE fligt nLflv'1)v~H i N~v~1OOaircaft OG IG 3G ---- w 0 w z LAJ ui 2 4 02 40 80 120 1602 80 120 160 FREQUENCY (CPS) FIGURE 23 Nor'lizvd at-erytvue Fourie;r tranmforrnt of Ko'oth'o, soundA of Hubject , obta ined during Ihre, flight mciii eu c's int .NIn'.-100 aircra ft, 28 OG IG 3G a- ! 0 CL& �0 I"2 0. LsJ w 2 40 80 120 8602 40 80 120 16

39 0 2 40 60 20 1SO FREQUENCY (CPS) FIGURE
0 2 40 60 20 1SO FREQUENCY (CPS) FIGURE 24 ' Normalized average Fourier tratns forms of Korotlcoi sounds of subject 5 obtc'ined during three flight maneuvers in NF-100 airotaf/t. 2. 229 Rest, arm relaecd Rest, arm tensed After exercise goo I-- II 2 40 ob do 66 k 0 doI do 6k 4'0 do do i6 21 FREQUENCY (CPS) FIGURE 25 Normalized average Fourier transforms of phase 1 Korotkov sounds of five subjects obtained during three rest/exercise maneuvers. Figures 25 to 40: Vertical columns allow subject.to-subject comparison of the frequency dis- tributior of the sound energy for each maneuver; horizontal rows allow maneuver-to-maneuver comparison for each of the five subjects. 30 Rest, arm relaxed Rest, arm tensed After exercise Iii CY w 1' 2 40 eb Ido i k0 4o io ,do m6 k 4bo bdo to aed 22 FREQUENCY (CPS) FIGURE 26 Normalized average Fourier' transfor'm.8 of phaiw 2 Korotkov sounds of five subjects obtained durin.g thJrec restl/exercise mnaiteuvers. 31 Rest, arm relaxed Rest, arm

40 tensed After exercise 04 23 FREOUENCY (C
tensed After exercise 04 23 FREOUENCY (CPS) FIGURE 27 Normal/ized average Fourier transformas of phaws, .1 Koirotkov Hoitnds o~f live subjeets obtaived during three rest/exercise maneuvers. 32 Rest, arm relaxed Rest, arm tensed After exercise N LAJ I-- uii 2 4o io ik mI62 4o do I4o i62 4o So I o I6O 24 FREQUENCY (CPS) FIGIURE 128 ,\'rmIlizid Ut'vtrrotlp .' F rirIt' troo PflPHN ot' ph1/ ( , K. rott,'P v ( uvdm of five' Nuhij(,cts Uhbtlin-cd (II#PII!/ Ph IiC pit .'C 'i(iS DltIt I't' , Hori, onta I HetadI 70~ u 1) 1Id Ic)f 1 INWIE-t fi'U-s %Ir~f~ o ib ,flf 'bi..Hoi~i' dir l ilcJ il-ii l I I ft" os Horizontal Head 70* up Head 100 down w I.- UJ 4 - -J 2 4osogOoIieo66 40 ogO gO16 k40 do JO led FREQUENCY (CPS) FIGURE :30 Nor'malized aragcet'l€ Fourier tranimfornts of phame 9j Korotkov mounds o flive subject# obtained d( rinlf/ thr'e tilt-tfble naneu , 'ern. 35 Horizontal Head 700 tp Head 10° duwn w0 Ze N w � w 2 40 80 4o0 160 2 40 do 62 1602i 40 d0 do0 e60 "27 FREQUE

41 NCY (CPS) Nor'malivzd aver'o'I'uie Fo~ur
NCY (CPS) Nor'malivzd aver'o'I'uie Fo~urier tran -?foms of phiut, ;1 Kwo l,' Iol ,r.Ilids of Ove 411b iJects ohrib iml d durby/ threi tiltl-fale WI M IOMhICe 8. :,6 Horizontal Head 700 up Head 100 down N cy i 2 46 io lo 16 0 do do 16 k 40 do do id 26 FREQUENCY (CPS) FI(URX :12 Notmalized avweragle F'ourir trrum4fornm ,o.f ph eaC 4 Korotl .ot, Huflndo of jive Rubjoctv obtained luriilltr three tilt-tabhl mal-Ut,'CrS. 37 IG 4 5G 3. 5G 'In N U , lid '-i 2 4b so d~o s66 46 do 4o 1e66 4o do do ied 29 FREQUENCY (CPS) FIG~URE' 33 Normalize'd aveIralge Fourier transfortmH of phuimt, I I'orotkov xotindn of fiive, seubjeetN obtititld du'krinuj thrence (entrifuge, fltu'nlUvCzerti. 38 I G 2. 5G 3. 5G w-r 2 , U IxI 2 40 eo i0o ie62 4o do i4o 6 k 460 do do ied 30 FREQUENCY (CPS) FIGUJRi' :14 Nornmalized avra jp F.'ourier f r.,1f.ortim of pha .H)E 2 vrotkoe' uounidi of /of fie mubjnects obrt lcd ultring thr/ee' ectitrfug/ la't,'c'rN. 39 IG 2. 5G 3. 5G 404 N° 2 40 So ilo I6 2 4o

42 0o 4o m6 k 46 do do ed 31 FREQUENCY (CPS
0o 4o m6 k 46 do do ed 31 FREQUENCY (CPS) i.0 I'i't' I' ' t )'fl,' i If jJiiiirn , I¢' tl, NL t t~f t ie'u ,f h jt' l tN ui, l uh l 40 L 10 2.5G 3.5SG --o k- -0 so -o -b -b 4op61 0d o 614 t 31FEUNC CS N41'R 3 I- 11pilx' vrfeF u irta ootea iae41io o on so ieiu jfttolie f Urpp trl-eitrfii aiu em I'E 141 00 IG 3G UJ w 1 1 2 40 SO 40 16612 4o do d~o 1662 40 io 4o0 60 33 FREQUENCY (CPS) '1111 P~~~t)' hi~~fl up ?I~~jI~I I(1 11 It I0, fDD1 11,11 NIII ' ie Hdj'I D7I i 4'2 OG IG 3G o u wp--I u,-L K UJC, 0 -J 'Ir. 2 40 iO 40O 166 2 40 JO 4~0 6 k 4`0 1O 4~0 FREQUENCY (CPS) e Niiii i ted averamp (h/A 119 999 'l')N 119l fols fVi t 111( 1 fol 0 Ho-flli9'De9I'Ioffil, 14J,0 lbro1, 01111' 1# Nh'.Ii OG IG 3G i~ 2 40 6o10 166 k~ 4o do0 do0 m6k 40 io do 16 FREQUENCY (CPS) N 111.1111lll4t., dI f I'v ol , l "ll ivict I,' , i' tI(I 1HI'llp'11IH fit l l ic .I Jh,1, 0-0111-1', 1,' , V , , I Hill JO N 0 ' I'll h# Hi t H 11 JI-1M fl t 1011,(i flitrNo I/ Ill I /Ive / fl IllhI man l (it Id.orI i

43 , Ito 1 l o) to 0, "44 L N 36 FREQUENCY
, Ito 1 l o) to 0, "44 L N 36 FREQUENCY (CPS) ~g i~p~g,/~ b r uiiii ntibhirme elf 11;1(181. 1, I011,011kiii moile,1r14 oy li iubjiertH obtaii id 45~ of sounds. Therefore, it might be concluded shown, however, that a more exacting "finger- from the data in this study that recognition print" can be determined by extending efforts of Korotkov sounds could be achieved more to analyze all three of the parameters of a reliably by locating a single-filter center fre- transient-frequency, amplitude, and time quency at a very low frequency, say 10 cps, (fig. 2) Although this study has shown the where energy always occurs. Use of a high- frequency spectrums of the 4 Korotkov sound detection threshold for this single, low-fre- phases to be grossly similar, figure 2 shows quency filter might well prove more reliable marked differences in the waveforms of the than a coincidence technic with several low- recorded Korotkov souLnds. Evidently, then, threshold filters placed in the h!

44 gher frequency the time sequence of occu
gher frequency the time sequence of occurrence of the various range where components occur on the average, frequency components is noticeably different for each of the 4 Korotkov sound phases. VI. RECOMMENDATIONS It is recommended, therefore, that future analysis efforts use computer technics which The results determined in this study do not can also provide time-domain information. represent an end but a strong beginning toward Such analysis technics, although necessarily a complete description of Korotkov sound quite sophisticated, will produce a more precise characteristics, The frequency analysis tech- fingerprint of Korotkov sounds. With this nic employed reveals a "fingerprint" which is added Information, it is reasonable to presume useful for devising filtering methods of recog- that a simple and more reliable device can be nizing Korotkov sounds, This study has also devised for recognizing Korotkov soundm, REFERENCES 1. Ware, It. W., and A. It, Kahn, Automatic

45 Indirect for the estlmatlon of blood pr
Indirect for the estlmatlon of blood prosutv'. .1. L th. blood pressure determination in flight, J. Appl, Clin, Med, 29:0138-0.51 (19,14). Physiol, 18:210.214 (100$), 8, GuddoN, L. A., W, A. Speinco:, and II. H,, 11off, 2, Collins, V, J., and F, Malora, Sphysmomonom,, Grap)hic roeordlnlJ of tho Korot, kofr mounds, try: The indirect measurement of blood pres. Amger, IHeart J, 57:31I1 (1969), mure, Ane.th, Aniig., 422:443.462 (19010), 3, Hooker, 1), H., and J, 1), Southworth. Interpruta. 1), John~on, It, A, Model 1I, automtti bhod preA m. tion of the aumcultatory blood proomure sounds. Dayton, m hio, Tit 5)11.411, p U, 1,,4F, l'w, I 591), Arch, Intern, Mod, 131384.389 (1014), 4, Br'amwell, J, C., and S. K. H1 likeon, Phono-artorl. Io, G(iford, 8. It., and II. P. Brolda, NIS rilport :ii101, ol(rlaMg, .J. PhYNhol, 00:xxxl (102'5), roJjot 1 204-20.5512, Plhymlologicil monftorlnW kitl l)pmlf. ro' anintmtho"l alud oth14 ' o ther um, It) 5, Kornx, I., M, rho nat1ure and time

46 relations of May 1064. the coIrn )1v{io
relations of May 1064. the coIrn )1v{ioni moundm of KorIoIov In muan, Ameir. J, Phyihol. 7M111247-204 (1020), II, Edlnignr, IL., and M. 8l'rlnw, A utonitl ohe(,tronlie blood apl)puratu., Amor, J. Mld, .I'Elc, 2illI.1 42 0i, Groodol, 1., M., and M. Miller. Graphic study of (011.111), auscultatory blood prounuru measurulnunl,, ,Hxpor, Mod, Hulg. 10149.102 (104;1). 12. Talkafaki, 11., et. il, F"requenc'y arumlysls of vas. cular Nound lroducod at the blood premouret 7, Rappaport, M, H,, and A, A. Lulsada. A phylieal meas0uruMUnt. Pocowdings 6 5th 1 tvrntitlukoil iiid physioloilc analysis and a now procedure C(3'ongrmss of Angiolohlcta, Pari, toi~l, 40 Security Classification DOCUMENT CONTROL DATA .R&D (Security claesslication of title, body of abstract and indexing annotation muat be entered whan the overall report In claesified) I. RIGINATING ACTIVITY (Corporate author) 2a. REPORT SECURITY C LASSIFICATION Systems Research Laboratories, Inc. Unclassified Dayton, Ohio 2

47 b anoup 3. fEPORT TITLE FREQUENCY ANALYS
b anoup 3. fEPORT TITLE FREQUENCY ANALYSIS OF KOROTKOV BLOOD PRESSURE SOUNDS USING THE FOURIER TRANSFORM 4. DESCRIPTIVE NOTES (Typo of report and Inclusive date&) _Z4inal Report Sý AUTHOR(S) (Last name, Iirst name. Initial) Rauterkus, Thomas Fickes, Jay W. Feltz, John F. 6, REPO RT DATE 7a, TOTAL NO. O PAGES 7 b, NO, OF REP, S Feb. 1966 -46 12 e. CONTRACT OR GRANT NO. AF 141(609)-2753 t, ORIGINAToR0' REPORT NUMSER(S) b, PROJCT NO, SAM-TR-66-8 ° Task No. 793003 Sb. OR,1yPORT NOS) (Any other nu.mb.e that may be .aoI;.d d. 10. AVA IL ABIL.ITY/LIMITAI ION NOTICESI Distribution of this document is unlimited, II, SUPPL EMENTARY NOT'S 13, SPONSORING MIL.ITARY ACTIVITY USAF School of Aerospace Medicine Aerospace Medical Division (AFSC) Brooks Air Force Base, Texas 13 ABSTRACT The purpose of this investigation was to determine the frequency content of' the sound signals (Korotkov sounds) obtained from the microphone located in the arm cuff of an automatic blood pressure measur

48 ing instrument. Korotkov sound recording
ing instrument. Korotkov sound recordings were made for five subjects in five experimental situations: rest, postexeroise, passive tilting, centrifuge rides, and flights in NF-100 aircraft. The frequency anaLlysis was performed by using a digital computer to obtain the Fourier transforms of the sound signals. The Fourier transforms were displayed on the computer oscilloscope and photographed. The photographs were then arranged in a number of rectangular arrays for convenient comparison of the frequency con- tent of the Korotkov sounds as related to the several types of Korotkov sounds, experimental tLtuations, and subjects. Initial study of the 240 average Fourier transforms contained in these arrays indicates no readily observable common characteristics except that most of the sound energy is almoot always located below 50 cps. DD IJAN41473 ...-Unclassified DD IeAN64 1 -j Security Classification ...._ 14. LINK A LINK , LINK C KEY WORDS ROLE WT ROLE WT ROLE WT Biodyna

49 mz cs / Korotkov blood pressure sounds S
mz cs / Korotkov blood pressure sounds SFrequency analysis (blood pressure sounds) Fourier transform (blood pressure analysis) In-flight monitoring (blood pressure sounds) * Monitoring, in-flight Blood pressure sounds INSTRUCTIONS 1. ORIGINATING ACTIVITY: Enter the name and address imposed by security classification, using standard statements of the contractor, subcontractor, grantee, Department of De- such as: fensee activity or other organization (corporate author) issuing (1) "Qualified requesters may obtain copies of this the rport.report from DDC." 2a. REPORT SECURTY CLASSIFICATION: Enter the over- (2) "Foreign announcement and dissemination of this all security clasaificttion of the report. Indicate whether report by DDC is not authorizedm" "diRestricted Data" is included. Marking is to be in accord- (3) " y S. ioe nm t auencied m o S ance with appropriate security regulations. (3) "U. S. Government agencies may obtain copies of wt o this report directly from DD

50 C. Other qualified DDC 2b. GROUP: Automa
C. Other qualified DDC 2b. GROUP: Automatic downgrading is specified in DoD Di- users shall request through rective 5200. 10 and Armed Forces Industrial Manual. Enter "_userssha___requestthrough the group number. Also, when applicable, show that optional markings have been used for Group 3 and Group 4 as author- (4) "U. S. military agencies may obtain copies of this lized, report directly from DDC. Other qualified users 3. REPORT TITLE: Enter the complete report title in all shall request through capital letters. Titles in all cases should be unclassified, ,1 If a meaningful title cannot be selected without classifica- tion, show titl* classification in all capitals in parenthesis (5) "All distribution of this report is controigled. hual. immediately following the title. ified DDC users shall request through 4, DESCRIPTIVE NOTES: If appropriate, enter the type of report, sag#, interim, progress, summary, annual, or final. If the report has been furnished to the Office

51 of TechniclI Give tbe'inclusive dates w
of TechniclI Give tbe'inclusive dates when a specific reporting period ii Services, Department of Commerce, for sale to the public, Indi- covered, cate this fact and enter the price, if known. 5. AUTHOR(S): Enter the name(b) of author(s) as shown on I1. SUPPLEMENTARY NOTES: Use for additionel explana- or in the report. E'ntet last name, first name, middle initial, tory notes. If military, show rank end branch of service, The name of the principal o.,thor W an absolute minimum requirement. 12, SPONSORING MILITARY ACTIVITY: Enter the name of the departmental project office or laboratory sponsoring (pay- 6. REPORT DATE.. Enter the date of the report as (day, Ing for) the research and 4evelopment. Inclhde address. month, year, or month, year. If more than one date appears on the report, use date of publication, 13. ABSTRACTi Enter en abstract givinga briefand factual summary of the document indicative of the report, even though 7s, TOTAL NUMBER OF PAGES: The tutal page c

52 ount it may also appear elhewhere in the
ount it may also appear elhewhere in the body of the teclhnical re- should follow normal pagination procedures, i.e., enter the port, If additional space im required, a continuation sheet shall number of pages containing informationm be attached, 7b, NUMBER OF REFERENCES Enter the total number of It Is highly desirable that the abstract of classified reports references cited in the report. he unclensilied. Each paragraph of the abstract shall end with e., CONTRACT OR GRANT NUMBER: If appropriate, enter an indication of tho military security classification of the in.. 4 the applicable number of the contract or grant under which formation in the paragraph, represented an (7s), (S), (c), or (U). the report was written. There is no limitation on the length of the abstract, How. 8b, S, Ia Id. PROJECT NUMBER: Enter the appropriate ever, the suggested length is from I10 to 225 words, ubprojact number, system numbers, tuask number, etc. 14. KEY WORDS: Key words are technicall

53 y meaningful terms or short phrases that
y meaningful terms or short phrases that characterize a report and may be used as 9e. ORIGINATOR'S REPORT NUMSER(S)-, Enter the offi- index entries for rataloplng the report, Key words must be vial report number by which the document will be identified selected -o that no mecurity ulamnifiration is required. Identi. and controlled by the originating activity. This numb&,r must flora, muc-h as equipment model desinastion, trade name, military be unique to this report. project code name, erigraphic location, may be used as key 9b. OTHER REPORT NUMIMR(S); If the report has been words hut will be to low,.d by an indication of technical con. assigned any other report numbers (either by the orlginator tuxt, Thn assignment of links, rulen, and wnights is optional, or by the sponsor), also enter this number(s), 10. ,'.VAILABILITY/LIMITATION NOTICES: Enter any lim. itatlions on farthor dissemination of the report, other than those Unclassified "ANN ,Security Claslification 4 4