6 ——————__ PATIENT MONITORING SYSTEMS YSTEM CONCEPTS ‘The objective of patient monitoring is to have a quantitative assessment of the important physiological variables of patients during critical periods of biological functions. For diagnostic eiresearch purposes, it is necessary to know their actual value or trend ‘of change. Patient aaaioring systems are used for measuring continuously or at regular intervals, automatically, the values of the patient's important physiological parameters. There art several categories of patients who may need continuous monitoring or intensive care Critically ill patients, recover ing from surgery, heart attack or serious illness, are often placed in special units, generally tnpwn as intensive care units, where their vital signs can be watched constantly by the use of ecrpnie instruments. The long term objective of patient monitoring, oe spelled out by Gardner (1972) is to decrease mortality and morbidity by (i) organizing and displaying infor- mation in'a form meaningful to improved patient care, (i) correlating multiple parameters for clear demonstration of clinical problems, (iii) processing the data to set alarms on the develop- ment of abnormal conditions, (iv) providing information, based on automated data, regarding therapy and (v) ensuring better care with fewer staff members. Koster (1978) states that during @ surgical operation, the patient is deprived of several Muinral eaction mechanisms which normally restore abnormalitics in his physical condition or alert other people. Indications or alarms that cannot be given by the patient himself can be presented by patient monitoring equipment. Besides this, in special cases, it is not uncommon forsurgical procedures to last for several hours. ‘During these lengthy operative procedures, it it dificult for the anesthesiologist and the surgeon to maintain intimate contact with the Patient’s vital signs and at the same time attend to anaesthesia, surgery, fluid therapy and Many other details that are required under such circumstances. Also, when a patient is con- Rected to life-support apparatus, ¢-8:» heart-lung machine or ventilator, correct functioning of Ce eee artes well, A patient monitoring system thus infor the surgeon and the anaesthesiologist of the patient's With patient monitoring systems, the Tis that surgery involves has been consi possible to detect the Complications before they prove dangerous as suitable measures can be taken well in ee The choice of proper parameters which have @ high information content is one of thely carried out ‘at the bedside, central station and bedside with Au asap these is dependent upon medical requirements, avail i i different man esi itors + ilable in a variety of configurations from Tana acaied oh cater different parameters but the common feature amongst all facility to Continuously monitor and provide non-fade display of ECG waveform and h instruments also include pressure, temperature and respiration rate mon Rania Ge Gia), Figure 6.1 (b) is a simplified block diagram of a typical bedside monitor, imi -d is the heart of the instrument wherein the master clock pro eee ot the instrument, The analog-to-digital circuits convert the two channels of parameter waveform information into the digital form and this is stored in two channels of RAMs. A D/A converter converts the stored information into analog form for isplay. The display speed is fixed at 25 mm/s and 50 mm/s, similar to the paper speed in conventional ECG recorders. i ‘The numerics board generates digit addresses in coincidence with each block address. Each digit in the parameter BCD code is addressed sequentially to enable transfer of the parameter information from the digital section to the numerics board. The character generator is pro- grammed to generate the numeric display values. This’ generator receives the parameter input data and converts this data into the correct format for display. A 60 kHz clock signal generates the CRT high voltage power supplies. The XYZ driver circuits amplify the incoming signals to a level suitable for driving the CRT. The alarm circuits receive sensing signals from the patient parameters under measurement and generate audio and visual alarms if a parameter is not within preset limits. The numeric value is displayed for both parameters selected for waveform display. The advent of microcomputers has marked the beginning of a fundamentally new direction in patient monitoring systems. Such systems are intended to replace the traditional monitoring devices with a single general purpose unit capable of recognising the nature of the signal Source and processing them appropriately. The hardware responsible for physiological signal analysis, information display and user interaction is actually a set of firmware modules implemented in terms of microcomputer program. The firmware gives the system its functional Personality and the usual switches, knobs, dials and meters can be replaced by a touch- sensitive character display (Deutsch et al. 1977). A typical example of a microprocessor-based patient monitoring instrument built around Z80 microprocessor is ‘Life-Trend 15° from Albury Instruments, UK. The system is designed to display electrocardiogram, digital heart rate with high and low alarms, digital pulse rate, ‘dynamic pressure or other waveforms from external pre-amplifiers. It also gives immediate - ‘Scanned with CamScannerPatient Monitoring Systems 137 lt Sas ( I c | I Eanarienle (>) Courtesy Ms H.P., USA). (Block diagram of a bedside monitor ~ (Based on Hewlett Packard Model 7834). formation on heart rate, temperature, and sys- urs, Figure 6.2 is a block diagram Rig. Tat) A opical bedside patient monitor Play of ECG, heart rate and blood pressure 4nd histori, i historical data on the patient for trend in and diastolic blood pressures for periods up to eight ho Seanned with CamScanne!CRT display tube FIG. 6.2 Block diagram of the bedside patient monitor (Courtesy Albury Instruments, UK). of this system. The system basically consists of three circuit blocks: Preamplifier section, Logie boards and Display part. The preamplifiers incorporate patient isolation circuits based on optical couplers. The ECG waveform has facilities for ‘lead off’ direction, ‘pacer’ detection and quick recovery circuit for overload signals. Various amplified signals are carried to 2 multis plexer and then to an analog to digital converter, included in the logic board. The central processing unit along with memory gives X and Y output for the CRT display. The character generator output is mixed with the Y output for numeric display on the CRT. The alarm set- tings, selection switches for different parameters and defibrillator synchronisation “system ‘communicate with the CPU. The alarm signals are also initiated under its control. The memory comprises 5K bytes ROM and 3.25 K bytes RAM with 256 samples of ECG. Eight seconds delayed ECG is available for recording purposes. Figure 6.3 is another example of a bedside monitor designed to carry out monitoring and display values for ECG and heart rate, arterial blood pressure (systolic/diastolic/mean); pul- monary artery pressure (systolic/diastolic/mean); mean central venous pressure; temperature and ree In addition, cardiac output, wedge pressure and fluids input/output measure- pase en ec psoogdire functions can be added to the basic system. The system is basi- ly patient hemodynamic oriented monitoring system. The system consists of (i) physio- logical signal amplifiers for various parameters, (ii) a four-channel non-fade di i fea 2 fade display for presentation of the monitored physiological waveforms, (ii 1 ; . for presentation of all derived vital signs and user request en Ee funsioning as central procesing element of the bedside, aed (y epeval parvo Keypad for user-bedside interaction. and, (v):a special purpose Kayo Seanned with CamSeannerPatient Monitoring Systems 139 FIG. 6.3 Bedside patient monitor based on micro- computer-system solo. (Courtesy Menon Medical Inc, USA). The LSI-11 microcomputer consists of 28 K words of RAM memory, a 64-channel A-D con- verter, a video generator and multiple serial interfaces. Analog waveforms and status lines are fampled continuously by the processor and algorithms invoked for the derivation of the sociated physiological parameter values. The derived values are stored in memory for later fetrieval in both graphical and tabular formats. The bedside monitor facilitates on-line communi- tation with a centrally located host processor through a standard serial ASCII link. The host processor may be computer-supported nurse station or a data base management facility, Central Monitors With central monitoring, the measured values are displayed and recorded at a central station. Umally, the signal conditioners are mounted at the bedside and the display and alarms, etc. arg located in a central station The central station monitoring equipment from Hewlett Packard known as Patient Informa: tion Centre (PIC) incorporates a multi-microprocessor architecture to display a flexible mixture of smooth waveforms, alphanumerics and graphics on a single cathode ray tube. This presents all the information at a glance and thus assists the hospital staff in the following ways. First, itgenerates audible and visual alarms if preset vital sign limits are exceeded. It is important ~ that the central station announces these emergencies without generating too many false alarms, tising due to patient movements, etc. Secondly, it displays the patient's vital sign data. By Witching this deta, the attending staff can detect problems before they reach ‘he alarm stage. Trend plots of vital signs provided by the PIC aid in guiding the patient's therapy. Thirdly, "provides recording ‘of the ECG and sometimes of other parameters, especially of the few 6 just before an alarm, which shows what kind of irregularity led to the alarm. Blancke ‘nd Neilsen (1980) explain the system concepts of the PIC. Primarily designed for coronary Seanned with CamScanne!140 Recording and Monitoring Instruments ients, it dis -rate information for eight pat; tients, it displays the ECG waveforms and heart-rate i ght pa The idiotay shows (Fig, 6.4) four seconds of real time ECG waveform and alarm informatia: A long trend (for either 9 or 24 h) or short trends (90 min) may be selected for display fog EOC OM. Cenos Orr Thor ECG INVALID) thrbeterbepcy ———— reece HR UG) uk x A cr heer eee FIG 6.4 Typical displays on the patient information centre. The dj Pe ces play depicts information on the Hl il Bid More hae ak ger heart rate, alarm limits, display of peat Bees FE | 7a ECG waveforms from four to eight i a patients (Courtesy Hewlett Packard, USA). observation and/or documentation. The information for the PIC is collected from the bedside, Each bedside cable contains as many as fifteen analog signals representing physiological para- meters which may include several blood pressures, ECG, heart rate, respiration, end-tidal CO, and temperature. Status information such as alarm signals is also carried by the same cable. The 80 or so incoming physiological values are then sampled and digitized at appropriate rates by a 10-bit analog-to-digital converter. ECG waveforms are sampled every two milliseconds to maintain the 0-100 Hz bandwidth. Slowly varying variables such as temperature are sampled every four seconds. The display part has two subsections—raster type display for waveforms and a conventional 300260 picture-element bit map for alphanumerics and graphics. To make the waveforms look smooth, 2 1200 line vertical raster has been used. The display subsection also manages a 16 K-word memory which is used as temporary storage for waveform and hard copy data. ECG waveform data for each patient is continually stored temporarily to provide a delayed (typically 8) waveform output for recording. The delay is used to capture a snapshot of any ECG abnormality leading to an alarm. If the recorder happens to be busy, the display processor stores it until a recorder becomes available, he trend memory can hold patient data for 24h. It contains 6 K words of CMOS RAM filter, explained in detail later in the chapter. There has been an increasing realisation meen with regard to assuring minimum breakdown of patient monitoring equipment. Goldberg (1980) explains a microcomputer based self-test Seanned with CamScanne!142 Recording and Monitoring Instruments ‘et eeeme tin uch 02 . v ‘a SUT. ww ie se FIG. 6.6 Diode pump circuit. If there are N pulses in a time interval ¢ such that each adds a charge q on the capacitor, they the total change in charge would be Q=Ng=NCV ‘Thus the average current over a period ¢ would be fee eNO a NOW) cop tire t The equation shows that the average current is directly proportional to the number of pulses per unit time. Thus, a current meter can be calibrated to give a direct reading of the average heart rate in beats/min. ‘When a positive voltage pulse of amplitude Vis applied at the input terminals of the circuit, capacitor C; would be charged to C,V through the diode D, which would conduct and offer negligible forward resistance. Therefore, the charging of the capacitor would be governed by the time constant R,C; which should be much smaller than the width of the input pulse. When the input pulse returns to zero, the cathode of D: is negative by V volts with respect to the anode, so that diode D3 is forward biased and starts conducting. Capacitor C, then dis- charges through the diode D:, the meter and the resistances R, and R. Capacitor Cy is used to average the current through the meter and hence it should be much larger than Cj. The circuit is arranged in such a way that capacitor C, is completely discharged before the next pulse appears at the input terminal. Another pulse of magnitude V would again contribute a charge q which is pumped through the meter when the input pulse returns to zero. If the current Iyy passes through a resistance R (resistance of the indicating meter), then the voltage across the capacitor is e=CVR This relation is true only if ¢ is made a small proportion of V. A linearit % 1 is Z ty of 0.1% can be achieved by eat ar 150 V and e= 1 V. This is not practicable most of the fa in solid state circuitry. Therefore, some form of modification is carried out to obt which has a linear relationship with frequency. eae The block diagram of a direct reading average heart rate me i ter is shown in Fig. 6.7, The ECG pulse received from the electrodes is amplified in a preamplifier to a level chet would operate the Schmidt trigger circuit. The Schmidt trigger converts each R wave into a rect: is then differentiated in an RC differentiator to yield multivibrator. The output of this multivibrator which ide and duration goes to the integrator (diode pump Proportional to the input frequency. Burbage (1973)ee nee aul cca ahe c two consecutive pulses mé ped related to the instantaneous pulse frequency. Pu aes relationship between output voltage and lis fer a im i ‘capacitor charging curve. The linearity was found to be within 6% ul es 0 ‘range. Roy and Wehnert (1974) used the same technique but ac ne inearity be frequency and output voltage by operating on a specific portion of a Sa charging. _ where the exponential charging curve can be best approximated by a hyperbola. +V Bistoble Imultivibrator Low level adjust FIG. 69 Circuit diagram of a linear scale beat-to-beat heart rate meter operating on the specific portion of the ‘capacitor charging cycle (Redrawn after Roy and Wehnert, 1974; by permission of Med. and Biol. Eng). pulse is directly proportional to the time interval between heart beats. Two identical RC Retworks are used to integrate the bistable pulse converting time to corresponding voltage. While one of the RC networks is being charged, the other has already been charged and is holding the charge for the meter to read. The ‘discharge switches’ momentarily short-circuit the integrating capacitors to earth whenever the bistable output goes from zero to a positive Voltage. The summing diodes couple whichever capacitor has the higher positive voltage to a Voltage follower and hence the meter, Use of a Voltage followet is essential to prevent the charge from leaking-off the capacitors during monitoring. , Parviainen et al. (1978) explained the construction of a cardiotachometer based on analog divider for achieving the function f = 1/7. Hartley (1976) used multiplying analog-to-digital Seanned with CamSeannerLo eee ) 146 Recording and Monitoring Instruments segment LED displays. The meter will indicate heart rate from 30-300 bpm with a calculated error of +0.2% at 200 bpm (Taylor and Mandelberg 1975). This circuit req storage registers, such as latches, between the R-R interval timer and the divider so that the count from the interval timer is then reset. A:method to overcome this problem and simplify the citcuitry is suggested by Nijhawan et al. (1981) and Yonezawa et al. (1983). Pratapa Reddy (1977) also used a programmable counter and bistable latches to convert the measured period into corresponding frequency. . ‘The use of an inexpensive calculator chip for obtaining the division required to convert the measured time interval to frequency in beats per minute has been suggested by Mason and Shoup (1979). It was shown that all the available calculator chips can fulfil the speed require. ments rieeded for heart rates up to 200 bpm. Ludwig (1977) used a Rockwell 5 function calcu lator to determine the time necessary to detect the occurrence of eight heart beats and to automatically calculate the rate in beats per minute. The instrument designed for heart rates from 20 to 200 bpm gives an accuracy of +1%. The use of ROM (read only memory) for storing a table for necessary division has been suggested by some authors (Voulgaris, 1984; Rehak, 1981). Dotsynsky et al. (1980) used an EPROM with a capacity of 1024 bytes for constructing a digital heart rate meter. The hyper- bolic transformation into beats per minute after measuring the R-R interval by the number of fixed frequency clock pulses is programmed in this memory so that the content of each address is the BCD value of the corresponding frequency in beats per minute. Assuming a freedom from error of R-wave identification, the devices error is zero in the range of 29 to 188 beats] min. The upper range with zero error can be extended by using a larger memory. Calculation of heart rate, from patients’ ECG, is based upon the reliable detection of QRS complex (Thakor, 1983). Most of the instruments are, however, quite sensitive to the muscle noise (artefact) generated by patient movement. This noise often causes a falsely high rate that may exceed the high rate alarm. A method to reduce false-alarm is by using a QRS matched filter, as suggested by Hanna (1980). This filter is a fifteen sample finite-impulse-response-filter whose impulse response shape approximates the shape of a normal QRS complex. The filter, therefore, would have maximum absolute output when similarly shaped waveforms are input. The output from other parts of the ECG waveform like T-wave will produce reduced output. Figure 6.11 is a block diagram of the scheme. The ECG is sampled every 2 m/s. Fast transi- Fibrillation detector ECG Slew rate 30 Hz 1.25 Hz ars 2-ms limit low-pass high - pass matched samples average filter filter Low limit Beat Heart rate Alarm detector —_— Taal r generator ‘comp. FIG. 6.11. Block diaram ofthe cardiotachometer based on matched , Rs a Hanna, 1980; by permission of Hewlett Packard, USA). eatin cane iof fntinitesimpalser we, Butterworth filter is employed. This preddestes fr ant rated “ae Bet with the signal is removed by a 1.25 Hz high-pass fier, The clamped end tenet oes aeform is finally passed through a QRS matched fier. The beat detector retossies ORS complexes ‘in the processed ECG waveform value that has occurred since the last heart beat to ineshold value. If this value exceeds a threshold : Value, a heart beat is counted. The beat interval averaged over several beats is used to calculate the heart rate for display, alarm limit comparison, trending and recorder annotation. The threshold in this arrangement gets auto- jn depending upon the val it f ‘matically adjusted depending up alue of QRS wave amplitude and the int {QRS complexes. Following each beat, an inhibitor af needed N 'y period of 200 ms is introduced duri which no heart beat is detected. This reduces the pias crite , eat the possibility of T-wave from getting counted. The inhibitory period is also kept varied as an inverse function of the high rate limit, with Jower high-rate limits giving longer inhibitory periods. Based on the power spectra estimation of QRS complex, Thakor et al. (1984) have suggested that a bandpass filter with a centre frequency of 17 Hz anda Q of five yields the best signal tonoise ratio. Such a simple filter should be useful in the design of heart rate meters, arrhy- thmia monitors and implantable pacemakers. ‘Measurement of Pulse Rate For pulse rate measurement, photoelectric transducer suitable for use on finger or ear lobe is used. The signal from the photocell is amplified and filtered (0.5 to 5 Hz passband) and time interval is measured between two successive pulses. The measuring range is 0-250 bpm. Monitoring peripheral pulse is more useful and dependable than heart rate derived from ECG in the case of heart block because it can immediately indicate the cessation of blood circulation in the limb terminals. Moreover, photoelectric pick-up transducer is very much easier to apply than the three ECG electrodes. The amplitude of the signal obtained by this method is also quite large as compared to the ECG signal and therefore it gives better signal- tonoise ratio. However, the technique is severely subject to motion artefacts. VLOOD PRESSURE MEASUREMENT (Blood pressure is the most often measured and most intensively studied parameter in medical and physiological practice! The determination of only its maximum and minimum levels during ach cardiac cycle supplemented by information about other physiological parameters is an invaluable diagnostic aid to assess the vascular condition and certain aspects of cardiac per- formance. Pressure measurements are a Vital indication in the successful treatment and manage- Ment of critically ill patients in an intensive cardiac care unit (Russel and Rackley, 1974) or the Patients undergoing cardiac catheterisation. The tremendous research and development for an ae blood pressure monitor has resulted in several methods but only very few have been Smmercialised due to certain practical difficulties. 4 saa The blood is pumped by the et hheart into the aorta which supplies it to the arterial circuit. Due Seanined with CamScanne!148 Recording and Monitoring Instvuments i illares, it loses most of its pressure and re is f the arterioles and precapillaries, i l wa low pressure via highly distensible veins. The right hart pumpsitt i ala Rela operates at a lower pressure. The heart supplies blood to both ¢ heen flow pulses of variable rate and volume. The maximum. p, multaneous intermittent flow j i » The maxims teached during cardiac ejection is‘called systolic pressure and the minimum pi occuring i ion i iastolic pressure. The mean arterial preg, tricular relaxation is termed as diastol e : : = eee ae is approximated by adding one-third of the pulse pressure (differen felneea systolic and diastolic values) to the diastolic pressure. All blood pressure measuremenis re made with reference to atmospheric pressure. . Typical hemodynamic pressure values are shown in Fig. 6.12. The nominal values in the to the load resi basic circulatory system are as follows: Arterial system 30-300 mmHg Venous system 5-15 mmHg 6-25 mmHg Pulmonary system FIG. 6.12 Typical hemodynamic pressure values present in the basic circulatory system (Courtesy Hewlett Packard, USA). The most frequently monitored pressures which have clinical usefulness in medium and long term patient monitoring are the arterial pressure and venous pressure. There are two basi¢ methods for making measurements of blood pressure—direct and indirect. (The indirect methods consist of simple equipment and cause Very little discomfort to the subject but they are intermittent and less informative, They are based on the adjustment of Seanned with CamSeannerae is i | damper consisting of ¥ damping. The other method is by using a parallel damper isting of a s - needle raldnscs parallel to the manometer and in series with a distensible p | ‘connected to a syringe. The parallel damper is able to flatten the resonance peak ang response curve is achieved almost upto the original peak. i Pressure waveform distortion due to transmission in fluid-filled catheters can be compen, menstein et al. (1976) can be by electronic means. The compensator described by Dai 7 n with the catheter in place by changing the natural frequency and damping ratio whey catheter parameters are altered by fibrin deposits, gas bubbles and softening of the cathy wall, The compensator is based on the fact that fluid-filled catheters can be characteriseq second order systems and the distortions introduced by catheters are typical of the output of systems (Melbin and Spohr, 1969). It is obvious that a compensator whose transfer function the exact inverse of such a second order system would facilitate recovery of the original ways, | form shape. , t Hok (1976) describes two methods of dynamic calibration of manometer systems Which are useful for detecting resonance frequencies and damping below 100 Hz. One method makes use of the step-function technique and the other is a sinusoidal pressure generator using magnetic fluids. It was found that resonance frequencies could be determined with an error of about 5%, whereas the error in damping factor is greater, of the order of +507%, The step function technique was found to be quicker and easier to apply under clinical environment. In an attempt to evaluate the pressure distortion due to clinical catheter-manometer systems, it becomes necessary to establish their linearity. A system is considered linear if its transfer function coefficients are independent of pressure and time in the applicable zone of pressure and frequency. Fromm and Delara (1974) showed that linearity is to be individually established for different catheters having differing cross-sectional structures. It was also shown that Dacron catheters whose cross-section contains an annulus of homogeneous Dacron matrix are seen to behave linearly while the linearity of other cross-sectional types of woven Dacron catheters remain in question. Patel et al. (1965) studied the harmonic analysis to determine the dynamic response requirements of catheter manometer systems. \ Alndirect Methods of Blood Pressure Measurement / The classical method of making indirect measurement of blood pressure is by the use of cuff over the limb containing the artery. This technique was introduced by Riva-Rocei for the determination of systolic and diastolic pressures. Initially, the pressure in the cuff is raised to @ level well above the systolic pressure so that the flow of blood is completely terminated — Pressure in the cuff is then released at a particular rate. When it reaches a level which is below the systolic pressure, a brief flow occurs. If the cuff pressure is allowed to fall further, just below the diastolic pressure value, the flow becomes normal and uninterrupted.)-/ ; #The problem here finally reduces to determining the exact instant at which the artery just opens and when it is fully opened|/The method given by Korotkoff and based on the sounds Produced by flow changes is the one normally used in the conventional sphygmomanometers. ‘The sounds first appear (Fig. 6.17) when the cuff pressure fas to just below the systolic press Bbey are produced by the brief turbulent flow terminated by a sharp Seanned with CamSeanner‘Patient Monitoring Systems 155 of Korotkoff sounds with variations in cuff pressure d 7 ‘A number of automated blood pressure measuring instruments have been designed. making Systolic pressure value Pressure, 1 t t ' Measuring duration” Blood tlow FIG. 6.17 Principle of blood pressure : \\\J pulsations measurement based on Korotkof sounds, ue of the Riva-Rocci method. They operate in a manner analogous to that employed by a uman operator, but differ in the method of detecting the pulsations of blood fiow at the systolic and diastolic levels.)/Frequency bands that best discriminate the Korotkoff sounds at systole and at diastole from the sounds immediately preceding these events must be defined for achieving a high degree of reliability in the automatic electronic blood pressure instruments. Golden et al. (1974) carried out a special analysis of seven Korotkoff sounds centred about the fystolic and diastolic ausculatory events and found that a maximum increase in amplitude at the systolic transition occurred in the 18-26Hz band. Similar results were obtained by Schulze (1968) who empirically found a bandpass of 15-30 Hz to provide optimal discrimination for changes in energy at the systolic Korotkoff sound transition. Similarly, a maximum decrease in spectral energy of diastolic Korotkoff sounds, at ausculatory cessation, was observed within 440-60 Hz passband. Wolthuis et al. (1974) constructed a Korotkoff sound processor for automatic identification of the systolic and diastolic sound transitions. Alexander et al. (1977) discuss in detail the criteria to be used in the choice of an occluding cuff for measurement of blood pressure. Automatic Blood Pressure Measuring Apparatus using Korotkoff’s Method The ae consists in putting a cuff around the upper part of the patient’s arm and applying @ ed waist the brachial artery. The compressed air required for inflating the cuff is prods by rs a ‘Ystem incorporated in the apparatus. Usually the inflating is done to eek ante ae Well beyond the systolic value at the rate of approximately 30 mm¥g/s, Tae a eae 'sthen decreased at a relatively slow pace at the rate of 3-5 mmHg/s. The ¢ such a way that the veins are not occluded. While air is allowed to leak from the cuff, t 2 Piezoelectric microphone. The corresponding electrical sign i ial tkoff sounds are picked up by @ spec ne ora cgnals are fed to a preamplifier. The ee Seanined with CamScanne!amplified signals are then passed on to a bandpass filter having a bandwidth of 25 to 1; With this passband, a good signal-to-noise ratio is achieved when recording Korotkoff from the brachial artery beneath the lower edge of the cuff. ‘The system is so designed that the appearance of the-first Korotkoff sound switches in systolic manometer and locks the reading on the indicating meter. In a similar way, the stolic value is fixed by the last Korotkoff sound. The cuff is completely deflated, automatically, after an interval of 2-5 s aftet the determination of the diastolic value. : Such instruments usually incorporate a sensitivity control which is required to be adjusteg. from patient to patient to compensate for extremely weak or strong sound intensities which are encountered. Correct setting of the instrument may be assured by monitoring the sound signals either by an electrostethoscope or on an indicating meter. Instruments operating on this principle are subject to serious errors, particularly in restless patients, unless steps are taken to ensure protection against artefacts. One method of doing this is to design the control system in such a way that when pressure is registered, the first sound must be followed by a second one within the preset interval. If this is not the case, the record. ed value is automatically cancelled and the measurement starts again with the subsequent sounds, The measuring accuracy of such type of instruments is not very high and the error is usually of the same order (5 mmHg/s) as is obtained in clinical sphygmomanometers. ‘A complete cycle of measurement consists of cuff pumping, controlled deflation, picking up and evaluation of the Korotkoff sounds, fixing of systolic and diastolic pressure and then complete deflation of the cuff. The cycle is initiated by a time delay and the operation is co trolled by a command pulse from a Schmidt trigger. Manual operation is, however, alway possible. ‘A number of shortcomings limit the application of Riva-Rocci method. Probably most serious among them is already mentioned, that the measurement is not continuous. Even for a particular single measurement a number of heart cycles intervene between the determination of systolic and diastolic pressures. Moreover, large errors are common since the pressure applied to the exterior vessel wall is not necessarily identical to that in the cuff, but is attenuated by the inter- vening tissue and an exact state of flow cannot be precisely determined. This problem is so severe that in most of the measurements made with instruments based on this principle, the diastolic pressure value is less reliable than the systolic. The diastolic pressure can be determined with greater accuracy and more reliability if the microphone output is amplified and fed to a chart recorder. The recorder can be calibrated in terms of pressure by feeding simulated signals corresponding to 60, 120, 180, 240, 300 mmHg. ‘Automatic built-in pump system of inflating and deflating the cuff must be provided with safety devices so that the patient does not experience any discomfort in case of system failure. Provision should be made for immediate switching off of the pump when the pressure in the cuff reaches the preset maximum over-systolic value and in no case the pressure in the cuff be allowed to exceed 300 mmHg. An additional arrangement must switch off the pump at any pressure after 20s from starting and deflate the cuff at a constant rate. These devices shall ensure that the pressure in the cuff will not reach too high a value and that no pressure is kept longer than approximately 20 s, Katona and Bolvary (1975) describe an indirect blood pressure measuring method which is an improvement over the conventional measurements made usi i i : P tions ing Riva-Rocci Korotkoft technique. HB this method, determination of both systolic and diastolic pressure values is by appearance of the Korotkoff sounds, an‘event which is more easily determined than their dis- Seanned with CamSeanner. Thus, the maximum cuff Pre: by reducing the measurement time ances physical oe aa sane the rate of change of eid ‘small in the region of measure- ents igh outside thus avoiding the 282d fo alata’ le this region, measurement time, The accuracy of the equines} was (iad (5 bo ane mmHg when compared with the intra-arterial blood pressu; Based on the same technique of detecting Korot one oats (Katona and Bolvary, 1973). ieee by Lewis eh al (1976), which forme i: Sounds on the inflation run is a toring system for intensive care patients, During a sure by means of an electropneumatic converter a signal from a microphone by noting when the sou: however, Bae ae during a time window which is opened following each QRS complex of the bes Ey on eer aes enc Program, false readings due to noise are conside- 1 pas a 4 correlation coefficient of 0.942 and 0.837 fe ic and diastolic values res; A waetees os ie eee Pectively, when the measurements were compared with readings } Sibley et al. (1983) made recordings of brachial artery pressure waves beneath a standard blood pressure cuff and found that Korotkoff sounds are propagated at slower velocities than | are other sounds. This phenomenon can be used to identify Korotkoff sounds in a noisy environment and to define systolic and diastolic pressure in quantitative terms, rather than by subjective interpretation of sound quality or disappearance. Phase Shift Method Another method of automatic indirect measurement of blood pressure is due to Dobbeleer (1965). The measuring procedure is similar to that used by Riva- Rocci of occluding the blood flow by a pressurised cuff and reading the pressure when the pulse appears at the systolic level and also when ‘it disappears at the diastolic level. But it differs from the system described above in the method of detecting the instant at which the cuff pressure is equal to the systolic and diastolic value. The arrangement employs a cuff consisting of three bags or compartments inflated as usual about 30mm higher than the expected systolic value. One of these bags is called the systolic bag and the other two are called diastolic bags. The systolic value is determined by detecting first blood flow pulse which passes underneath the diastolic bags by means of the small third bag. The diastolic value is determined by detecting the disappearance of the phase shift of the signals from both diastolic bags, the phase shift being caused to flow pulsations due to inertia of blood. Heated thermistors are used to detect air flow impulses produced by the blood flow underneath the cuff. The Rheographic Method A fully automatic apparatus for measuring systolic and diastolic blood Pressures using ordinary Riva-Rocci cuff and the principle of rheographic detection of an arterial pulse is the Philips blood pressure apparatus. Basically the method is similar to the Dobbeleer’s method except that here the change in impedance at two points under the occluding’ cuff forms the basis of detection of the diastolic pressure instead of the phase shift produced. in the blood flow pulse (Parrot, 1969). " A set of three Tag (Fig. 6.18) which are attached to the ‘cuff are placed in Saas With the skin. A good contact is essential to reduce the skin electrode contact impe ee Electrode B which acts as a common electrode is positioned slightly distal from the a 5 es Of the cuff. Electrodes A and C are placed at a certain distance from the electrode B, o Seanined with CamScanne!158 Recording and Monitoring Instruments distally and the other proximally. connected to the electrodes A and C. When we mi Ca eee ansancinge electrodes before pressurising the cuffs, it shows modulation in cy daciodelatoaiadll pulsations in the artery. Therefore, arterial pulses can be detecte fication of this modulation. Distal lamp. a = = J) Systolic pressure indicator Pump iS FIG. 6.18 Rheographic method of indirect blood pressure measurement. When the cuff is inflated above the systolic value, no pulse is detected by the electrode 4, The pulse appears when the cuff pressure is just below the systolic level. The appearance of this fist distal arterial pulse results in an electrical signal which operates a valve to fix a manometer pointer on the systolic value. As long as the pressure in the cuff is between systolic and diastolic values, difference signal exists between the electrodes A and C. This is because the blood flow is impeded underneath the occluding cuff and the pulse appearing at the elec trode A is time delayed from the pulse appearing at C. When the cuff pressure reaches diastolic pressure, the arterial blood flow is no longer impeded and the difference signal disappears, A command signal is then initiated and the diastolic pressure is indicated on the manometer. In the rheographic method of measuring blood pressure, the cuff need not be. precisely Positioned as in the case with Korotkoff microphone which is to be fixed exactly above an artery. Also the readings are not affected by ambient sounds. Similar to the rheographic method of measuring arterial blood pressure non-invasively, is the photoelectric plethysmograph which detects cardiac volume pulse and transduces it to an Glectrical signal. The rate of pulsatile blood volume change in a peripheral site, such as the earlobe or forehead, is used for indicating the magnitude of the arterial systolic blood pressure ater calibration against a standard blood pressure measuring means (Warner and Sankar, 1977). The correlation between rate of peripheral pulsatile blood-volume change and arterial systolic blood pressure is more close if these quantities are measured during a portion of each cardiac cycle, i.e. period of diastole or systole and for just one instant during this period. Seanned with CamScannerPatient Monitorir szolis pressure, the occluded artery sna re ; in a Doppler shift i Yiqion results ina Doppler shift in the returnin es e if ig ult i i a Tecognised to register the tentative systole Sta ed ae < ecome fixed as true s from recording a false systole reading. As a mined for wi wats from DEINE accepted as true artery ‘lina Poe Seed Se ape ving more frequently than every 250 milliseconds are rejected te Gant eee i This sets the upper limit (240 bpm) on the patient's heart rate Se ea 1 gion, The lower limit of response is 24 bpm. At diastole, cuff presu sein ash is, acral wall Pressure. AS a result, wall snapping ceasteand iiwedtee, cae iy no longer receives Doppler shifted returns, The reading is re saat ies allowed to deflate rapidly to atmospheric pressure. The sad oae ica ae Ferree’ line gs are held fixed until Mpoosive cuff is placed on the arm (Fig. 6.20) in the usu ; ; vamiveer on the arm over the brachial artery. The eufT is cneated a above sacar id artery Hoe. This reduces the possibility of artefact eiyet check, the audio returns are exa FIG. 6.20 Measurement of blood pressure using ultrasonic Doppler- shift principle (Courtesy Roche “em pRESsURE Medical Electronics Division, USA). CUFF then deflated at a specified rate. A low energy ultrasonic beam (less than 50 mW/em!) at & fietheney of 2 MHz is transmitted into the arm. The portion of the ‘ultrasound that is reflected bythe arterial wall shifts in frequency when the wall of the artery moves. Above oa a Bees cioced duc to the pressure of the occluding cuflyané mont signals ae not eetived, As cuff pressure falls to the point where it is just overcome by brachial ee Sure, the artery wall snaps open. This opening wall frowement,corespondingt the oecrrenes Stthe fist Korotkoff sound, produces a Doppler shift. which is inerpreted BY 105 at eee nad aisplayed accordingly. With cach sbSd005 ie ag Tequeney shift is produced, until at the diastolic presser the artery is no ae ae Mpid motion suddenly disappears and the Doppler shift BerOeT eine anal ett notes the sudden diminution in the amplitude of the Doppler shi help ort is displayed as diastolic pressure. Sp . i i ent ecial electronic circuits used in he a ium Aiscriminate against extraneous motion artefacts. ‘A coupling medit is Seanned with CamScanne!162 Recording and Monitoring Instruments ents? ski ficient transmission of ultrasonic between the transducer and the patients’ skin for ¢ J lt “ La Roche, manufacturers of ultrasonic based blood pressure instruments ree uble and can be removed easily from the patients’ skin ann M{S Hoffman- mend Gelisonde. This is water sol the transducer. : 7 Unlike the Korotkoff method, the instruments based on ultrasonic Doppler shift princip, often provide reliable blood pressure measurements in severe hypotensive states, at unfa able sites such as the popliteal artery, in neonates where no other indirect method of measure, ment is feasible, in patients too obese for successful ausculation, under unfavourable condition such as high ambient noise, and in many species of laboratory research animals. Hochberg and Salomon (1971) using pooled data from seven teaching hospitals, compared readings on ultrasonic based blood pressure measuring instruments (Arteriosonde) against intra arterial values in 58 patients and against Korotkoff-method determinations in 244 patien Blood pressures were measured on patients in the operating room, recovery room, intensive care unit, obstetrical service and cardiopulmonary laboratory. A correlation coefficient of about 0.95 was obtained in systolic and diastolic measurements. MEASUREMENT OF TEMPERATURE The transducer normally used for temperature measurement in a patient monitoring system is a thermistor. Changes in resistance of thermistor with temperature are measured in abridge circuit and indicated on a calibrated meter. The measuring range is 30-42°C. In case a digital display is required, the temperature monitoring circuit can take the form of Fig. 6.21. Initially, the switches S; and S; are in position as shown. In this state, the voltage appearing across ger (0°C) is applied to the non-inverting input of Ay. The output of 4; is applied to both the inverting and non-inverting inputs of 4z, so the output of 4; is balanced, However, during this period, the capacitor C at the inverting input of 4, is charged to the reference signal amplitude (Vger °C). At time Tz, the position of switches is reversed and the input to 4, is available from thermistor. Also, the inverting input from 4, to Ay gets dis- connected. A voltage less positive than (Vger °C) appears at non-inverting input of 4, The differential between inverting and non-inverting inputs causes the output of 4; to go negative. During this period, the capacitor C, discharges through Rj. It continues till the voltage across , equals the voltage generated by Rr (thermistor). When both inputs are equal, the output of 42 becomes positive. In actual circuit, FET’s are used as switches S; and S». High frequency clock pulses are counted during the period 7; to Tz and the duration T; to Tp is proportional to the measured temperature. In a patient monitoring system, two channel temperature measure~ ments are usually carried out. Similar to ECG monitoring, the output circuits are isolated through opto-couplers. Provision for inoperate conditions are also included in such type of monitoring equipment. MEASUREMENT OF RESPIRATION RATE a transducers commonly used to measure respiration rate include thermistor placed in front a the ae microswitch or displacement transducer Put across the chest, impedance elect- rodes and signal from CO; measurement. The respiratory signal from anyone of these trans~et al. (1969) conducted experiments on man 5 to. of sinusoidal alterna ing current which can be passed eign ta utes Ti to avoid stimulation of sensory receptors, nerves and muscle, currents gher i frequency 1 Sisiz must be used for the measurement of physiological events by impedance, Prone jower than 5 KHz are particularly hazardous since ventricular fibrillation may be produced with ‘aoxantial current flow. The use of higher frequencies not only provides the protection jn the avoidance of tissue stimulation, but also provides the safe use of cerreate ofa itis ich ebuld be lethal if the frequencies were lower. a "The electrical impedance changes associated with physiological activi i ‘ensively. Some of the physiological quantities which have Ce a ae nae ciltimpedance method include respiration, blood flow, stroke volume, autonome nergoue stem activity, muscle contraction, eye movement, endocrine activity and activity of brain cells (Gaker et al. 1973). Baker (1971) hhas reviewed applications of impedance technique for both fuanitative and qualitative measurement of physiological events in animals and man. CO; Method of Respiration Rate Measurement Respiration rate can also be derived by monitor- jing continuously the CO2 contained in the subject’s alveolar air. Measurement of CO, in expired air is otherwise useful in several ways; for example for original setting up the respirator and in ‘making adjustments to it afterwards, supervising patients suffering from respiratory paralysis, and other cases where there is respiratory involvement. The measurement is based on the absorption property of infrared rays by certain gases. Suitable filters are required to determine the concentration of specific gases (like COz, CO, and 1,0) constituting’ the expired air. Rare gases and diatomic gases do not absorb infrared rays. When infrared rays are passed through expired air containing a certain amount of COs, some of the radiations are absorbed by it. There is a proportional loss of heat energy associated with the rays. The detector changes the loss in heating effect of the rays into an electrical signal. This signal is used to obtain the average respiration rate. Figure 623 shows the arrangement for detection of CO: in the expired air. Two beams of equal intensity of infrared radiations from the hot-wire spirals fall on one half of each of the ‘condenser microphone assembly. The detector has two identical portions separated by a thin, flexible metal diaphragm. The detector is filled with a sample of pure CO. Because of absorption ‘of CO; in the analysis cell, the beam falling on the test side of the detector is weaker than that falling on the reference side. The gas in the reference side would, therefore, be heated more than that ori the analysis side. As a result, the diaphragm is pushed slightly to the analysis side ofthe detector. The diaphragm forms one plate of a capacitor. The infrared beams are chopped at 25 Hz and the alternating signal which appears across the detector is amplified, shaped and Suitably integrated to give respiration rate. APNOEA DETECTORS Apnoea is the, cessation of breathing which may precede the arrest of the heart and circulation Seanned with CamSeannerPatient Monitoring Systems 167 ircuits. The sensor may be a strain the amplifier is adjusted to zero volts Motion channer Posi detector the input signal from the sensor pad to the logi a ance embedded in the mattress. The output oh, Motion /Resp. discrimination cirevit detector From Apnea period sensor selector circuit Alarm circuit Respiration channel ‘Schmitt trigger FIG.624 Block diagram of apnoea monitor (Courtesy B-D Electrodyne, USA). with offset adjustment provided in the amplifier. The amplified signal goes to motion and respiration channels connected in parallel. The motion channel discriminates between motion and respiration as a function of frequency. In case of motion signals, high level signals above fixed threshold are detected from the sensor. In the respiration channel, a low pass filter is incorporated. Low frequency signals below 1.5 Hz (respiration) cause the output of the Schmidt trigger circuit to pulse at the respiration rate. Higher frequency signals, above 1.5 Hz (motion), cause the output of the trigger to go positive. The absence of signal (apnoea) causes the output of the Schmidt trigger to go negative: The outputs of the motion and the respiration signals are combined in NAND gate compa- Tator which compares the polarities of the motion and respiration channel signals to indicate respiration. The presence of respiration is indicated by a flashing lamp. The output of the dis- ctimination detector also goes to an apnoea period selector circuit, a low frequency alarm oscillator and driver, a tone oscillator and audio amplifier connected to a speaker. Audible alarm is given at a frequency of 800-1000 Hz, which is pulsed at2Hz. An alternative method is based on electromagnetic induction and is described by Yoshikawa and Saito (1972), It consists in passing a high frequency alternating current through a trans- mitting coil and creating an alternating magnetic field. The transmitting coil is placed at some distance from the infant, The receiving coil is applied on to the abdomen of the infant. The alternating magnetic field induces an emf in the receiving coil. The movement of the abdominal Wall with infant’s respiration results in inducing an amplitude-modulated signal in the receiving