CA2080297A1 - Method and apparatus for excluding artifacts from automatic blood pressure measurements - Google Patents

Abstract

A METHOD AND APPARATUS FOR EXCLUDING ARTIFACTS FROM
AUTOMATIC BLOOD PRESSURE MEASUREMENTS

Abstract of the Disclosure A method and system for detecting blood pressure pulses and screening out artifact-induced pulses when an automatic blood pressure monitor is provided. In preferred embodiments, the system detects the occurrence of an oscillometric pulse, determines the amplitude and rise time of the pulse, determines whether the amplitude and rise-times are within patient-specific screening limits and disregards pulses not within the limits, determines whether a pulse matches other pulses and saves the matching pulse data, determines the diastolic and systolic pressure based on the saved matched pulse data, and updates the patient-specific screening limits.
In a preferred embodiment, pulses match based on similar amplitude and rise time characteristics. The amplitude and rise time screening limits are updated based on average values of prior blood pressure measurements for the patient.

Description

p~cri~ o~

A Mh~OD A~1:1 AEtpARaTus FOR EXCL~ING ARTI~ACTS F~OP~
s AuToMa.TIc BLOOD PRESS~RE ~aS~REMEN~S

~i51 This invention relat:es to the autoD~atic ;~
measlLremen E: o~ blood pressllre, and more particularly to a ~ethod and apparatus ~or detecting blood pressu~e pulses and scraening out ~tifact~induc:ed pulses.

- Auto~atic bLood pressure monitors are com~only used to psri.odically ~eas~e the~ blood pressure of a patlent. In most automatic blood pressura monitors, a pr~ssu:re c~f~ is attached to a patient' 5 arss over the b~chlal ar'c~ry . The cu~i~ i5 f irst pressurizlad wi~ an applied pxess~ure that is high ~nough to substantially 2~ oc~lude the brachial arter he c~î pressure is then gradually reduced, either continuol:lsly or in increme~ts.
As ~he pressure is reducsd to sys~olic pressure, the flow o~ blo~d through the brachial arl:ery beneat~ the cuff ~creases su~tantially. l' When the~ blo~d f :Lows 'chrough th~ ~rac:hial art~ry f ollo~iI~g ~z~ch contraci:ion of ~ hQart, it i~nparts a pulsatile move~ent to tha w~ll of ~e artery. This pulsatile move~nent is coupled to a blood pressure cl~f~
extendiRg over the artery as minute chanqes in the cu~

3 0 prQssu3~e, which a: :a 3~own as o~cillometric pu15~ -P~uto~atic blood pressure monitors measure and record the alhplitude o~ the oscillometric puls~s at a number o~ ~uf~
pressures. Af'cer the bloc~d pressure mQasu~ement had b~en completed, a table contains tha oscillom~tric pulse amplitudes recarded at each GU~:e pre5SUre.
In theory, the systolic, diastolic, and mean art:erial blood pressures can then be deter~irled from the 2~2~

values in the table using empirical defini~ions o~ these pa~a~e~ers as a function o~ the amplitudes of thes~
oscilloma~ic pulses. However, blood pressure ~leasure~nen~s are often advexsely affectE2d by artifact~
5 generally produced by pztient movement. Motion-induced arti~acts can substantially alter the measured amplitude oi~ scillomet~ic pulses thus introduc:ing inaccuracies in the measurement o~ the patient ' s blood pressure .
Prior systems use vario~s techniques to min~ize 10 l:he 6~ ects of artifac:1:s. S<:~e prior ~ystems screen o~cillometric pulses bas~d on their amplitude. Pulses ~ith amplitude~ outside ~he screen are considered arti~act~induce~ The screen is generally a population--based screen and not speci~ic to any patien~. Some ~5 previous systems also compare s~ ential pulses to ensure thay are blood pres~sure induced. I~ two se~aential pulses hav~ similar amplitude3, the~ syste~s a~ me the puls6~ is blaod pres~;u~e inducsad. HOWQVer, thQSe screening and comparison teChni~UQs do not always produce acceptabl~
2a results. It would be des~rable to have a system in which the scr~ening and comparing of pulses mc~e accurately id~ntiî ies blood pressure induced pu:lses .

It is an object o~ the present invention to provide an improved me'chod and sy~tem ~or dQtecting blood pre~;sure induc:P~ pulses in an automatic blood pressure ~eastare~ent d~vice~
I~ is aalother ohject of the pre~ent invention ta 3 0 pro~ride a me'ehod ancl system in which oscillometric puls~s ar~ ~atched based on the amplitude and rise time characterl~;tics c~ the puLs~s.
It is another obj ec:t of the present inven'cion to provide a method and syste~ in whic:h arti~ac~:-induc:ed pulses can be detected by the use o~ patient-spec:iîic a3~plitude and rise time screening limits.

,.

: ;

2~8~97 It is another obj ec~ of the present invention to provide a ~e thod and syste~ in which the patiea~-specif ic ~creening limits are automaticaLly updated based on data ~at~ered durirl~ the blood pressure measurement proc~ss.
These and other o~j ects, which will become apparent as the invention is more fully described b~low, are obtained by an improved bl~cd pressure measurement d~vice. In pre~erred eD~odiments, th~ d~vice detects t21~
occurrenGe ~f an o~cillom~tric puls~, determines the amplitude and rise ti~e Q~ the pulse, deter3lines whel:her the amplitude and rise time are within patient-specif ic scraeninq li~llit S and disregards pulses nc~t within the limits, dete~ines whether a pulse matches ot~er pulses and saves the ~Qatching pulse data, determines the diastolic and sys~oliG pres~urs based on th~ sav~d matchad puls~ data, and updat~s patient-speciîic screQning l~mits.
In a pr~3f erred embc~diment, pulses ~atch basad oxs simila:~
amplitud~ and rise time chaxacteristics. The amplitllde and rise time scre@ning l~mits are updated based on Z3 averaqe values o~ prior blood pre~s~3ure measurements for ~e~ patient.

~igure 1 is a blocX diagram of an aul:c~matic blo~d pressure ~onit~9r ~or detectin~ blood press~e pulses and screenin~ out artifact-induced pulses.
Fi~ure 2 is a flow diagra:m oi~ an over~riew of 'che proce~:~ing o~ the present imrerltion.
Fig~res 3A, 3B, and 3c are a detailecl flow 30 diag~am of the procsssing routin~ of the prasent inventlon.
Figure 4 is a flow diagram o~ tha ~nt2rrupt routin~.

.. ,, ~ , ..... .

2 ~ 9 7 one embodim~nt of a system f or detecting blood pressure puls~s and screening out arti~act-induc~d pulses in an autol~atic blood pressure measuring systsm is illustrated in Figure 1. The syste~ 10 comprise~; a number oî hardware components, all of which are conventîonal.
T~e system includ6~s a conventional }3locd pres~;ure c~ 12 in ~luid co~nmunication with conduits 14 and 16, a con~rentional pump 18, a conventional valve 20, and a conventional pressure transd~cer 22. The pump 18 a~d valve ~ o arP operated by a cor;ventional microprocessor 3 o .
As explained in greater de~tail below, during the operation o~ the automatic: blood pressure Dleasurin~
sy~;tem, the blood pressure cuf~ 12 is ~n~lated to a pr6~ssure that is great~r than sys~ole as indicated by the pressuxe transducer 22. The valv~ 20 is then open~d, u~ually ~or a predetermined periocl, alt3;1ough it may be continu~usly opan to allow a slight lea3cage of air fro the blood pre sure cu~ 12. However! the valve 20 normally allows air to esc:ape ~rom the cuff 12 fairly rapidly in relatively small increm~ts. A^~ the pre~sure in the cu~ 12 is reduc~d, either gradually or incrementally, th~ pressure in the cuf ~ 12 is measured by the~ pr~ssura ~ansducer 22.
Th~ preR~;ure in th~ blood pre~;ure cu~ 12 Gonc~ists of two CODIpOnent5~ na~aely, a relati~ely con~tant~
or "DC, " component and a relatively variable, or '~At:, "
co~ponent~ Th~ relativ~ly cons~ant component i5 a ~unctio~a o~ th~ pressure in ~ blood pressure cuf~ 12.
3 0 Th6!~ r~lativQly variable com~onent is produced by the minute chanye in the pressure of the cu~f 12 ~ollowing each contrac:tion o~ the heart. Thus, the reLatively constant DC component of the pre~sure in the cuf ~ can be us~d as an indica~ion a~ cu~ pressure, while the 35 relativcly variable AC component of the pressure in the C~ 12 can be used as an indication of an oscillometrlc pulse .

Two signals are obtained from the pressure transducar 22. O~e set of circuitry 34 supplies a DC
cQmponent to an analo~-to-digital (~/D) converter 32.
~nther set of circuit~y 36 supplies an AC compone~t to the S A/D conv~rt~r 32. The signal supplied through the DC
cixcuitxy 34 is thus an indication of the cu~ pres~ure, while the signal suppli~d ~hrough the AC circuitry 36 is an indication o~ ~Ac oscillome~ric pulse. The A/D
conv~rter 32 digitizes the DC and the AC signals and outputs digital bytes indicative o~ ~heir values ~hrough a bus 3~ to the microprocessor 30.
As ~entianed above, the microprocessor 30 is o~
conventional vari~ty and, as is typical with such devices, is cann~cted to a random acc~ss memory 40 used for the storage o~ da~a, and to ei~her random access memory or read-only ~emor~ 42 that contains the so~tware ~or ope~ratinq the microprocessa~ 30. Operator controls 44, such as a keyboard or buttons, are al50 connect2d t :3 tha ~icroEaro~:~ss~r 30.
Althouqh tlle measur~g system 10 illustrat2d in Figure 1 u~ilizes a pressure transciucer 22 and separate cir :uitry ~or the ~C and tha ~C pres;sure signals, it will be understood that other ~plementations ~re possi~le.
Por examplla, a single circuit providin~ a signal corres~ndinq to both th~ stQady-~ta~e a~d the variable prass~uras; in the cu~f 12 can b~a supplied to tha analog-to-di~i~al converter 32. After the signal is digitized by analog-to-digi~al converter 32 and applied ~o tAe microproc~ssor 3 O, algori~s exel::uted by the 3 0 microproc:essor 3 0 can de~ect the ste~dy~state component of the cu~ pressure and the variable coDlponent va~ia~ions in the cuf f pre ~ure .
In a E~ref~rred eD~adi.ment, th~ pr~sent in~rention is used to perio~ically deter~ine the systolic and diastolic blood pressure of a patient. To determine the systolic and diastoliG blood pressure, the preferred ~ystQm of the present invention collects bload pressure 2 g 7 readings at several c~f f pressures, ref erred to as target prsssure levels. When the cuff is inflated or deflated to the target pressur~, the ~yst2m samples the output of the A/D converter 32 to deter~ine when an oscillome~ric pulse S occurs. The oc:ourrence o~ an oscillometric pulse is d~tected when the ~ nal applied tu tha A/l) converter 32 has predetarmined characl:eristics, as described in greater detail below. ~rhe syste~ th~an det~rmines whet:her these pulses are within pul~e amplitude and rise ~imR screening 10 li3sits. If an oscillo~etric pulse is ~ot wit~in the screening limits, it i5 disre~arded. Th~ screening limits ~or the patient are updated aft~r the systolic: and diastolic blood pressure are determined . This ei~f ects the automatic updating o~ patient-specif ic screening limits .
15 Th~3 syst:em collects data on puls~s at a particul~r tarqet pressure level until two ~Datching pulse~ are detected.
P~l~es match when they hav~ similar amplitude and rise ti~ characteristic~. When a ~atch is ~ound, tha system ~aves the ~atching puls~ data. The syste~ collects matching pulse data for various ta:rget pressure levels.
The system than evaluates the mat:rhin~ pulse data to de~ermine the systolic and diastolic pressure. Algorithms to determine the systolic and diastolic bload pressures ar~ w~ know~ as are various algorithms to reduc~ tha 25 Q~ e-:ts o~ arti~acts. An algoritbm to deée~mine systols and dia~ol~ is d~scribed in U. S. Paterrt Nl~ . 4, 7~5, 820 ~ntitled Method And Apparatu~ For Systolic Blood Pressure ~easurement , which is incorporated herein by ref erence .
Algorithms to reduce tha e~fects of artifacl:s are 30 desc:ribed in U~ S. Patent No. 4, 77'7, 959 entitlad Arti~act Detection Based On Heart Rate In A Method And Appara~us For Indirect Blood Pressure ~easurement and U. S. Pat~nt No. 5, 014, 714 entitled Method And Appaxa~us For Distinsuishin~ Between Ac~ ate l~d Inaccurate 3100d 35 Pressure Neasurements In The PresencP Of Arti~act, which are hereby incorporated by r eerence . I~ the system 2~8~

succ:essfully deter~ines sys~olic and diastolic blood yr~ssure, the system then updates the screening limits.
Figure 2 is a flow diagram o~ an over~riew o~ the processinq o~ the presant inventior~. In a pr~erred 5 embod~ment, the processin~ is ~mplemented on a computer pro~am, which ~xecutes on ~icroproc:essor 3 0 . In bloc3c 2 01, the system sel:s the cu~f pressure to th~ tar~et:
pressur2. In a pre~erred embod~ent, t~e target pressure is ~nitially set in exc~ss of thr anticipated systolic:
10 pressure. The target pressure is then decremented se~entially. In ~lcck 202, the systeDI waits until the interrupt rou~ine detes::ts a p~a}c in ~e oscillo~aetric data. A peak usually indicates t21at a pulse has occurred, al~ough a pea}c could be produced by an arti~act. In a 15 pref err~d em~odiment, a peak is d~tected when the AC
pxessur~ is increasing and passes through a "high trigger~' :L~v~l and than pas~s throuqh a "lo~ trigger" level while decreasing. In bloc~ 203, ~he system determines whether the detected pulse is within the screeninq l~its. In a pre~erred embodiment, th~ scree~ing limits are pati~nt-specific lLmits basQd on the amplitude and the rise tim~
of the puls~. In alternate e~hod'ments, th~ pulse ~all tLme, pulse ~idth, integral of the pulse, and derivatives ~ ~he pulse can b~ used ~s s~ee~ning limits (and ~or ~atching pulses a~ discussed ~low). I~ the puls~ data i5 within th~ screening li~its~ then the syste~ continu~s at bloc~ 204, ~lse the syste~ disregards the pulsQ data and lo~ps back to bl~ck 202 to await the next peak. The syætem ~aintains a Temporary Tabl~ ~f pulse data that is collectPd at a particular tarqet px~ssure and that is within the screening limits. In block 2~4, the system d~t~rmines whether the current puls~ data matches any of th~ data stored in the T~porary Table. In a preferred embodiment, a ~atch is d~tected bas~d on the amplitud~ and the rise tim~ of ~hQ pulse . If a matoh is detected with any a~ the puls~ data in the Te~porary Table, the system continues at blQck 206, else the system adds the current 2~297 pulse data to the Temporary Table ln blocX 205 and loops to bloc~ ~02 to detect ~he next peak. The system maintains an Oscillometric Table ~hat GOntains ~he DC
pressure, AC pressur~, rise time, and heart rate ass~ciated with each ~atched puls~. (The rise time is used to update the screening lLmits.) In block Z06, the system adds the average pulse data o~ the matched pulses to the Oscillometric Table. In block 207, the systam clears the Temporary ~a~le Ln preparation for the collecting of d~ta at th~ next target pressure. In block 208, the syst~m deter~ines whether there is enough data in the Oscillometric Table to per~o~ a blood pressure ~aluation. If there is e~ough data in the Oscillometric Table, the system continues at bloc~ 2~, el~e the system lo~ps to block 2al to collect ~ore data. In bloc~ 209, th~ system evaluates ~he Oscillom~tric Tabl~. In bloc~
210, the sy~tem determines if th~ evalua-tion was succ~ss~ul. If the ev~luation was ~uccessful, the sy~tem c~ntinues at bloc~ 211, elsa the syst:em loops to block 201 to s~t the c~f~ pressure at the ~ext target pressure. In block 211, the syste~ updates the screening limits ~or the patiQnt and the proc~ssLng for 1~e particular blo~d pre~sure reading is co~plete.
Figur2s 3A, 3~, and 3C are a detailed ~low diagra~ ~ the procassing routine o~ the presen~
in~ention. Th~ proc~dure that invokes this routine p~ormC~ a mlmb-ar of conventlonal tasks. These tasks ~ncluds, upon completion of a ~easurament, ope~ing the value 20 to fully de~ate the cuf~ 12, displayi~q or ~toring ~he results o~ the measuremsnt, and scheduling the next automatic measur~ent. Also, the in~oking proc~ure ~ ializ~s the scre~ning li~its be~ore Ln~kin~ the processing routine ~h~ first time for a particular patient. In a pxeer~ed embodLment, the am~litude and rise time screening limits are set to a level that will, ~ ect, allow any pulse to pass throu~h ~he screeniny test initialiy. The scre~nin~ limits will be .......

2~297 autolaatically adjusted to the patient ' s characteristics as described below. Referring now to Figure 3A, in block 301, the system perfonns initializatiorl for the subse~uent processing. The syste:m initializes thP Te~nporary ~able, S Oscillome~ric: Table, and vari~us ~lags. In block 302, the system detex~ines i~ ere is ~nou:~h data in ~e 05 illometric Table to calculate a blood pressure reading.
In a pref erred embodiment, thea~e is enou~h data in t:he O~cillo~aetric Table when 'chere are three entries. One 10 skilled in the art would appreciate that other criteria can be used to deters~ins whether there is enough data in the Oscillome~ic Table~ If there is enouqh data in the Oscillo~netric Table, the syste~n c~ntinues at block 330 in Figure 3C, else ~'le syste~ continues at bloc~ 3~3. In 15 blocks 303 'chrou~h 308, the sys~em sets the c~:E pressure to the target pressure and loops wa:Lting f or a peak to be detected. Iî a pea3c is detected the system continues at bloc~ 310 Ln FiqurQ 38~ In block 303, the system calculat~s a ~arqe~ pre~s~re. In a pre~erred embodL~en~, ~ha target prQssure is initially set higher than the anticipated systolic pressure. In a pre~e~red emb~diment, each pass through block 3û3, the ~ys~e~ decrements the target pressure by 8mm. ~owever, c~ne skilled in the art w~uld appreCiatQ that other ~thocls o~ calculating th~
target pressur~ would b~ acc~ptabl~. In blocX 304, the sy~tem sets th~ cuf~ pre~sure to th8 tarqet pres~ure.
Initially, th~ cu~ pressur~ st~rts off wall ~low the tar~et pressure. I~ the cuf~ pressure is below the target pressure, ~he system energizes the pump 18 to increase ~hQ
pre~sure. Con~ ely, i~ ~he target pressure is belaw ~he cu~ pressure, then tha system will release pressure ~rom the cuff 12 ~hrouqh the valve 20. In bloc~ 305, ~he system d~termin~ whethar a pQak has b~en detectQd. The int~rrupt routin~, a~ de~eribed in Figure 4, d~termines 3 5 whe~her a peak has be~n detected and sets an appropriate ~lag. I~ a peak is detected, then th sy~te~ continues at ~lock 310 o~ Figure 3B, else the system continues at block 3 07 . In bloc:}cs 3 07 and 3 08, the syste~Q checXs various ~lags that ar~ set by t~e intPrrupt routine. These fla~s are used to determine whether the time at the t~rget pres:~ura has been too long and whether the cuff pressure 5 is r~eaar enough to the target pressure . In blocX 3 07, if the t~me at the partic-llar targ~t pressure has been too lon~, then tha system loops to bloc~ 3 û3 to calculate a n~aw target pre~ re and cortinue proc~ssing, el~e th~
~ystem contintle~ at blocX 3û8~ In ~lock 308, the 5y5tem 10 deter3aines whether the cuf~ pressure is near the target pr~ssure . If t:he cuf f pressure is near t~e taxget pre~sure, then the system continues to wait ~or a peak by looping to bloc3c 305, else the system loops to bloc3c 304 to re~;at l:he ~f ~ pressure to the target press~Lre .
Re~e~in~ now to Figure 3B, in bloc3c~ 310 and 311, the syst3m det~ine~ whether th~ detected pulse data i~ with~n ~he a3~pli'cudQ and ris~ t:ime sc:reen~ng l~mits.
q~he ~creening limits are updated i.n hlock 336 of Figure 3C. In bloclc 310, i~ the amplihlde of the det~cted pulse 20 is within the amplitude screening li:mits, 1:he system contimles at bloeX 311, else the system disregar~s the pulse data and loops to bloc3c 305 o~ Figure 3A to wait ~or th~ next paak. In block 311, i~ the ris~ time oî the d~tscted p Lalse is within the ris~ time ss:reening l~mits, 25 ~en the 5y5tel~ sntinues at blo~: 312, else the syst~
loc~p~ to block 3 Q5 in Figure 3A to wait ~or the n~ct pealc.
In bloclcs 3 12 through 3 19, the system determiIIes whether ~e detected pulse ~atche~ any other pulse data storQd in tha Temporary Table at the particular target 3 0 pre~sure . I~ the detected pulse matchas a pulse in the Te~porary Table, then the syste~n updates the Oscillon~etric q~abla and con~inues at bloclc 302 o~ Figure 3A to detensina i;S th~re is enouqh data in the C~scillometric Table to be ~valuated. I~ no match is ~ound., then the sys~e~ adds the 35 detected puls~ data to thQ T~mporary Table and continues at block 305 in Figure 3A to wait for the next peak. In bloc:k 312, if the detected pulse is the f irst pulse at the target pressure to pass the screening limits, then there is no da ta in the Te~porary ~ai:;le to Dlat ;:h an~ the sys~em continues at bloc3c 317, else the systeIQ continues at bloc:3c 313. In bloc:Xs 313 throu7h 316, the system loops S comparing the detected pulse with each of the ~ntries in Temporary Tabl6~ a match is ~oulld, the syste: n continues at bloc~c 3 18 . In bloc~ 3 13, the sy~tem selects tha next pulse in the Temporary Table starting with ~he las~ pulse stored in ~he tableO In block 314, the sy~tem 10 determines whel:her the amplitude ol~ 1:he sele~d pulse matches the amplitude of the detected pulse. In a preferred emb~di~ent, the amplitude of ~.~ro puls~s mat::h whR~ the ab~s~lute value oi~ their di~ rences is les than or equal to a constant value plus 9% o~ the amplitude of 15 thQ detected pulse. In a pre~erred em~odiment, the constant is s~t to ac ::ount f or the inherent noise in the d~tection equipmQnt. Ona skilled in the art would . appreciate tha~ althc~u~h amplitudQs preferably D~atch when th~y are within 9%, other matching criterions are 20 acceptable. If an aJnplitude match is found, the syst~m continues at bloc~ 315, else the system continues at bloc3c 316. In block 315, the system determines whether the rise t~ o~ the s~lected pulse m;~tches th~ ris~ t~me o~ l:he d~tected puls~ a pr~f errad embodi.mel~t, the rise times 25 o:e pulses matsh when th~y arQ w:ithin approximately 8 millisecondsO Cin~ s~cill~d i~ thQ art would appreciat2 that oth~r rise! time ~natching criteria would produc~
ac:ceptabl~ results. I~ the ri5e times ~natch, then the system continues at block 318, else the system contir~ues 30 a~ ~alock 316. In bLock 316, if all the data in th~
~emporary Table hAs been chec3ced ~or a match, then the ~ystem corltinues at block 3 1.1, else thQ sys l:em l~ops to block 313 to ¢Rlect the nQxt pulse in th~ Te~p~rary Tabl~.
In bloc3c 317, no ~atch has been ~ound, and the syst~m 3 5 stores the detec:ted pulse data in 'che Temporary Tabl~ and continues at bloc3c 30S of Figure 3A to wait for the nex1:
peak. In bloc3c 318, the system plac~s the matched pulse 12 ~8~2~7 data in ~e Oscillometric Table. In a preferred em~odlment, ~he a~erage value of the amplitude and rise tLmes and associated cu~ DC pressure ~or the selected and detected pulses are stored in the Oscillometric Table.
The time interval betwee~ the two matching pulses is also stored to be u~ed to deter~ine ~he heart rate. In block 319, ~he system clears the Temporary Table for processing at ~he next target precsure level, and the system con~inues a~ bloc.~ 302 of Figure 3A to determine i~ there 0 i5 enough data in the Oscillometric Table for evalua~ionO
Referring to Figure 3C, ln block 330, the system evaluates th Oscillometric Table to determin~ the ~ystolic and diastolie blood press~xes. Three outco~es o~
this evaluation are p~ssible. First, the evaluation was i~complete because of insufficient data. Second, ~he evaluation found data inconsistencies aLthough there was suf~icient data. During evaluation, each amplitud~ ~nd h~ark- rate interYal in the Oscillometric Table is che~ked ~or internal consistency with the other entries. ~hese ch~cks are fully describ~d in th~ earlier cited patents.
Third, the evaluation was successful. In bloc~ 331, i~
the data i~ insu~ficient, then the~ system loops to block 303 in Fisur~ 3A to coll~ct addit:ional data at th~ next lower target pressure, else thQ system continues at bl~c~
2S 332. In bloc~ 332, if inconsiste:ncies with the data ar~
~ound, then th~ sy5tEm return5 wi~h an error c~de, els~
the syste~ continues at bloc~ 333. I~ bloc~ 333, ~he systQm detQrmines whether the scre~nins limits should be updated. In a pre~erred ~mb~diment, the screeninq lLmits should not be updatsd when too ~any pulses ~ail to pass the screening lLmits or when too many nonmatchin~ pulses were dQtected. This test ensures that the screenin~
limits are set as a function of the patient oscillometric data rather than a~ a ~unc~ion o~ artifacts~ In alternate embodiments, th~ scre~ninq limits are updated when majority o~ the pulses ~all outside the screening limits.
I~ the ~creening limit~ are to be updated, the syste~

~ ~ ;

continues at bloc~ 334, else ~he system retuxns. In block 334, the system calculates ~he aYerage highest amplitude ~or several blood pressure readings. In block 335, the system calculates the average lon~est rise ~ ~e ~or several blood pressure readings. In blo k 336, the system updates the screenin~ limits based on ~he average highest ituds ~nd the average longest rise time. In a prei~erred em~odiment, the maximum amplitude scr~ening li~it i5 set to 15C% of the average highest amplitude, and t~e mini~num amplitude SGreen~ng l~mit i5 s2t to lgg6 o~ the av~rage highest amplitude. In a preferred ~mbodiment, th~
m~ximum rise tlme screening limit i5 set to 20 milliseconds l~nger than ~he average lonyest rise tLme~
There is no ~inimum rise time sereening lLmit. In a 15 pre:l~erred e~odiment, the avera~es are calculated using 'che last ~ive blos~d pressure mea urements that resultec3. in an update of ~he screening limits. One skilled i~ the art would apprecia~e that the updating o~ these screening 1 ~ its can be varied and still produca accaptable results.
Th~ system then returns.
Figure 4 is a ~low dia~ram of the interrupt routine. The interrupt is tim~r-~rive~. The interrupt routine inputs di~itized data, detects when a peak occurs, ~nd checks ~or t~inq and pressure measurements. The interrupt routine sets ~lags ~hat are used by ~he main proc~ssing rou~in~ ~or detection of certain conditions.
B~cause o~ ~h~ potential dan~er to a patient~ ~he interrupt rou~ine firs~ determines whether the allowed ~L~e to have the cu~f pxessuriæ~d is exceeded. In bloc~
40~ the time is exce~ded, ~he routine conti~ues at bloc~ 402, else the rout me con~inues a~ bloc~ 403. In blocX 402, the rou~ine d~lates the cuff and re~urns ~o the calling routine, ra~her than the interrupted routine.
In block 403, the routine inputs the digitized AC
oscillometric pressure and the ~C cuff pressure and Lnitiates the next sa~pling by the ~/D converter. In bloc~ ~04, the routine deter~ines if a peak occurred. A

~, ': ; ' :

1~

p~ak occurs wh~n the data has passed from beLow a high trigger level to above ~he high trigger level, and ~hen passes ~rom above a low trigger level to below ~he low trigger level. One skillad in the art would appr~ciate tha~ other peak detection algorithms would produce acc ptable results. In bloc~ 405, if a peak was detected, then the rou~in~ sets the pe~k detected flag in bloc~ 406.
In blocX 407~ if ~he time at the target pressure has exceeded a pre~etarmined tim2, then ~he routine sets the target pressure timQ exceeded flag. The routine stores the raw digiti2ed data, and the ~ain processin~ routine calculates the amplitude and rise time~ In a prefe~red ~bodiment, the ris~ ti~e is the amo~nt of time it tak~s the pressure to rise between 25% and 88~ o~ the pulsa amplitude. The routine then returns to the interrupted routine.
Although the present invention has been d~cribed in t~r~s o~ pr~erred ~bodiments, i~ is n~t intended that the inv~ntion be limited to these ~mb~diments. M~di~ication Withill the spirit of the invention will be apparsnt to thoe;e skilled in the art.
Th~ scope of the present invention is defined by ~he claims that ~ollow.

Claims (22)

1. A blood pressure monitor, comprising:
a blood pressure cuff;
an air pressure pump pneumatically coupled to the blood pressure cuff;
a valve pneumatically coupled to the blood pressure cuff;
a pressure transducer pneumatically coupled to the blood pressure cuff, the pressure transducer generating a signal indicative of air pressure in the blood pressure cuff and a signal corresponding to the oscillometric pulses in the blood pressure cuff;
processor means connected to the pressure transducer and receiving the signal indicative of the air pressure in the blood pressure cuff and the signal corresponding to oscillometric pulses in the blood pressure cuff, the processor means connected to the air pressure pump and the valve, the processor means comprising means for energizing the air pump to inflate the blood pressure cuff, periodically energizing the valve to incrementally reduce the air pressure in the blood pressure cuff, detecting an occurrence of an oscillometric pulse, determining an amplitude and rise time of the oscillometric pulse, determining whether the amplitude and rise time are within amplitude and rise time screening limits, disregarding the pulse when the amplitude and rise time are not within the screening limits, determining whether the screened pulse matches another screened pulse detected at the same cuff air pressure level based on the amplitude and rise time of the pulses, saving pulse data when pulses match, and determining systolic and diastolic pressures based on the saved pulse data.

2. The blood pressure monitor of claim 1 wherein the amplitude and rise time screening limits are updated based on data specific to a patient.

3. The blood pressure monitor of claim 2 wherein the amplitude screening limit is based on the average highest amplitude of previous blood pressure readings for the patient.

4. The blood pressure monitor of claim 3 wherein the amplitude screening limit has a maximum limit that is set to approximately 150% of the average highest amplitude.

5. The blood pressure monitor of claim 3 wherein the amplitude screening limit has a minimum limit that is set to approximately 19% of the average highest amplitude.

6. The blood pressure monitor of claim 3 wherein the rise time screening limit is based on the average longest rise times of previous blood pressure readings for the patient.

7. The blood pressure monitor of claim 6 wherein the rise time screening limit has a maximum limit that is set to approximately 20 milliseconds longer that the average longest rise time.

8. The blood pressure monitor or claim 1 wherein the saved pulse data is the average data of the matched pulses.

9. The blood pressure monitor of claim 1 wherein the processor means determines whether the screened pulse with a matches another pulse by comparing the screened pulse with a pulse other than the immediately previous pulse screened.

10. The blood pressure monitor of claim 1 wherein pulses are considered to match when the absolute value of the difference in amplitude is less than or equal to a percentage of the amplitude of the last screened pulse.

11. The blood pressure monitor of claim 10 wherein the percentage is approximately 9%.

12. The blood pressure monitor of claim 1 wherein the rise time is the time between the occurrences of approximately 25% and 88% of pulse amplitude.

13. A blood pressure monitor, comprising:
a blood pressure cuff;
an air pressure pump pneumatically coupled to the blood pressure cuff;
a valve pneumatically couple to the blood pressure cuff;
a pressure transducer pneumatically coupled to the blood pressure cuff, the pressure transducer generating a signal indicative of air pressure in the blood pressure cuff and a signal corresponding to the oscillometric pulses in the blood pressure cuff;
processor means connected to the pressure transducer and receiving the signal indicative of the air pressure in the blood pressure cuff and the signal corresponding to oscillometric pulses in the blood pressure cuff, the processor means connected to the air pressure pump and the valve, the processor means comprising means for energizing the air pump to inflate the blood pressure cuff, periodically energizing the valve to incrementally reduce the air pressure in the blood pressure cuff, detecting an occurrence an oscillometric pulse, determining whether characteristics of the pulse are within screening limits, disregarding the pulse when not within the screening limits, saving pulse data when within the screening limits, determining systolic and diastolic pressures based on the saved pulse data, and updating the screening limits based on the characteristics of a plurality of previously detected pulses.

14. The blood pressure monitor of claim 13 wherein the characteristics of the pulse include the pulse rise time.

15. The blood pressure monitor of claim 14 wherein the pulse rise time is defined as the time between the occurrences of approximately 25% and 88% of pulse amplitude.

16. The blood pressure monitor of claim 13 wherein the screening limits are updated based on data specific to a patient.

17. The blood pressure monitor of claim 16 wherein the screening limits are updated based on the average characteristics of previous blood pressure readings for the patient.

18. A blood pressure monitor, comprising:
a blood pressure cuff;
an air pressure pump pneumatically coupled to the blood pressure cuff;
a valve pneumatically coupled to the blood pressure cuff;
a pressure transducer pneumatically coupled to the blood pressure cuff, the pressure transducer generating a signal indicative of air pressure in the blood pressure cuff and a signal corresponding to the oscillometric pulses in the blood pressure cuff;
processor means connected to the pressure transducer and receiving the signal indicative of the air pressure in the blood pressure cuff and the signal corresponding to oscillometric pulses in the blood pressure cuff, the processor means connected to the air pressure pump and the valve, the processor means comprising means for energizing the air pump to inflate the blood pressure cuff, periodically energizing the valve to incrementally reduce the air pressure in the blood pressure cuff, detecting an occurrence an oscillometric pulse, determining an amplitude and rise time of the detected pulse, determining whether the detected pulse matches another pulse detected at the same cuff air pressure level based on the amplitude and rise time, and when pulses match saving

19 pulse data, and determining systolic and diastolic pressures based on the saved pulse data.
19. The blood pressure monitor of claim 18 wherein determining whether the detected pulse matches another pulse compares the detected pulse with a pulse other than the immediately previous pulse.

20. A method of screening out artifact-induced oscillometric pulses in a blood pressure monitor, the method comprising the steps of:
detecting an occurrence of an oscillometric pulse;
determining a rise time of the oscillometric pulse;
comparing the rise time of the oscillometric pulse to a rise time screening limit; and disregarding the oscillometric pulse if the rise time is outside the rise time screening limit.

21. The method of claim 21 wherein the rise time screening limit is based on the average longest rise time of previous blood pressure reading for the patient.
23. The method of claim 22 wherein the rise time is the time between the occurrences of approximately 25% and 88%
of pulse amplitude.
24. The method of claim 23 wherein the rise time screening limit has a maximum limit that is set to approximately 20 milliseconds longer than the average longest rise time.
25. The method of claim 20 further including the steps of:

determining an amplitude of the oscillometric pulse;
comparing the amplitude of the oscillometric pulse to an amplitude screening limit; and disregarding the oscillometric pulse if the amplitude is outside the amplitude screening limit.
26. A method of detecting blood pressure induced oscillometric pulses in a blood pressure monitor for a patient, the method comprising the steps of:
detecting occurrence of oscillometric pulses:
determining characteristics of the oscillometric pulses;
comparing the characteristics of an oscillometric pulse to detected pulses other than the immediately previous oscillometric pulse detected; and when a comparison is found, using the comparing oscillometric pulses to determine the blood pressure.
27. The method of claim 26 wherein the characteristics include the amplitude and rise time of the oscillometric pulses.
28. The method of claim 26 wherein the characteristics include the amplitude of the oscillometric pulses and a comparison is found when the amplitudes are within approximately 9%.
29. The method of claim 26 wherein the characteristics include the rise time of the oscillometric pulses measure between the occurrences of approximately 25%
and 88% of pulse amplitude, and wherein a comparison is found when the rise times are within approximately 20 milliseconds.
30. A system for calculating blood pressure from a plurality of oscillometric pulses detected at a plurality of pressures, comprising:

first means for determining whether the amplitude and rise time of each oscillometric pulse are within amplitude and rise time screening limits;
second means for disregarding the oscillometric pulse when the amplitude and rise time of the oscillometric pulse are not within the screening limits;
third means for saving pulse data corresponding to the oscillometric pulse when the amplitude and rise time of the oscillometric pulse are within the screening limits; and fourth means for calculating blood pressure parameters based on the saved pulse data.
31. A system for calculating blood pressure from a plurality of oscillometric pulses detected at a plurality of pressures, comprising:
first means for determining whether a characteristic of each oscillometric pulse is within a screening limit;
second means for disregarding the oscillometric pulse when the characteristic of the pulse is not within the screening limit;
third means for saving pulse data corresponding to the oscillometric pulse when the characteristic of the pulse is within the screening limit;
fourth means for calculating blood pressure parameter based on the saved pulse data; and fifth means for updating the screening limit based on the characteristic of a plurality of previously detected pulses.
32. A system for calculating blood pressure from a plurality of oscillometric pulses detected at a plurality of pressures, comprising:
first means for determining whether the amplitude and rise time of each oscillometric pulse matches the amplitude and rise time of another oscillometric pulse detected at the same pressure;

22 second means for saving pulse data corresponding to the matched amplitude and rise time of the oscillometric pulses; and third means for calculating blood pressure parameters based on the saved pulse data.
33. A system for calculating blood pressure from a plurality of oscillometric pulses detected at a plurality of pressures, comprising:
first means for determining whether each oscillometric pulse has a characteristic similar to an oscillometric pulse detected at the same pressure other than the immediately previous pulse detected at the same pressure;
second means for saving pulse data corresponding to the oscillometric pulses with similar characteristics; and third means for calculating blood pressure parameters based on the saved pulse data.