INTERNATIONAL OIML R 89 RECOMMENDATION Edition 1990 (E) Electroencephalographs - Metrological characteristics - Methods and equipment for verification Electroencéphalographes - Caractéristiques métrologiques - Méthodes et moyens de vérification OIML R 89 Edition 1990 (E) ORGANISATION INTERNATIONALE DE MÉTROLOGIE LÉGALE INTERNATIONAL ORGANIZATION OF LEGAL METROLOGY
Foreword The International Organization of Legal Metrology (OIML) is a worldwide, intergovernmental organization whose primary aim is to harmonize the regulations and metrological controls applied by the national metrological services, or related organizations, of its Member States. The two main categories of OIML publications are: International Recommendations (OIML R), which are model regulations that establish the metrological characteristics required of certain measuring instruments and which specify methods and equipment for checking their conformity; the OIML Member States shall implement these Recommendations to the greatest possible extent; International Documents (OIML D), which are informative in nature and intended to improve the work of the metrological services. OIML Draft Recommendations and Documents are developed by technical committees or subcommittees which are formed by the Member States. Certain international and regional institutions also participate on a consultation basis. Cooperative agreements are established between OIML and certain institutions, such as ISO and IEC, with the objective of avoiding contradictory requirements; consequently, manufacturers and users of measuring instruments, test laboratories, etc. may apply simultaneously OIML publications and those of other institutions. International Recommendations and International Documents are published in French (F) and English (E) and are subject to periodic revision. This publication reference OIML R 89 (E), edition 1990 which is under the responsibility of TC 18/SC 4 Bio-electrical instruments, was sanctionned by the International Conference of Legal Metrology in 1988. OIML publications may be obtained from the Organization s headquarters: Bureau International de Métrologie Légale 11, rue Turgot - 75009 Paris - France Telephone: 33 (0)1 48 78 12 82 and 42 85 27 11 Fax: 33 (0)1 42 82 17 27 E-mail: biml@oiml.org Internet: www.oiml.org 2
ELECTROENCEPHALOGRAPHS METROLOGICAL CHARACTERISTICS METHODS and EQUIPMENT for VERIFICATION 1. Scope 1.1. This Recommendation concerns analogue electroencephalographs, which are defined as diagnostic medical instruments having at least eight channels. Such instruments are intended to accurately measure and record bioelectrical potentials and their time sequences as detected by electrodes attached to the surface of the human scalp. 1.2. This Recommendation does not apply to information-processing electroencephalographs, nor instruments that digitize the signals, nor other special-purpose instruments. 1.3. This Recommendation is intended for the use of metrological services and other responsible governmental agencies. It specifies the metrological characteristics to be determined, methods of and equipment for testing, and initial and subsequent verifications of electroencephalographs that are regulated by laws. Metrological characteristics include errors in measuring voltage signals and time intervals as well as a number of other characteristics that affect measurement accuracy. Procedures are given for determining the relative error of measurement of sixteen instrument characteristics, and formulae are provided for calculating the associated permitted limits of these characteristics. These testing and control procedures provide a basis for ensuring that electroencephalographs of adequate accuracy are available for clinical measurements. The characteristics of the instrument covered are those that have been deemed to be important in order to achieve the necessary measurement accuracy in clinical practice; therefore, any changes in the requirements prescribed should be based on clinical experience and the needs of electroencephalography. 1.4. This Recommendation does not address pattern approval, terminology, technical requirements, electrical safety and the associated test methods. Some of these matters are being considered by the International Electrotechnical Commission (IEC) in the (draft) Publication 601-2-xx "Medical electrical equipment Part 2: Particular requirements for the safety of electroencephalographs" under the auspices of SC 62D on "Electromedical equipment". A subject that is also under consideration by SC 62D is " Particular requirements for the performance of electroencephalographs". The requirements of this Recommendation are also compatible with "Recommended specifications for electroencephalographs" developed by the International Federation of Societies for Electroencephalography and Clinical Neurophysiology (IFSECN). Those specifications can be found in "Recommendations for the Practice of Clinical Neurophysiology", IFSECN, Elsevier, Amsterdam- New York-Oxford (1983). 3
2. Characteristics to be verified The characteristics shown in Table 1 should be measured during verification of electroencephalographs. Table 1 Metrological characteristics to be verified. Metrological characteristics Point Verified during Initial Subsequent verification Relative voltage-measurement error 5.3.1 + + Relative sensitivity-setting error 5.3.2 + Relative time-interval measurement error 5.3.3 + + Relative error of the recording speed 5.3.4 + Recording hysterisis 5.3.5 + Relative errors of internal calibrator and time marker 5.3.6 + + Overshoot 5.3.7 + Time constant 5.3.8 + Amplitude-frequency response curve 5.3.9 + + Input impedance 5.3.10 + Common-mode rejection ratio 5.3.11 + + Baseline width 5.3.12 + Baseline drift 5.3.13 + + Intrinsic noise level referred to the input 5.3.14 + + Interchannel crosstalk coefficient 5.3.15 + Relative error of the interelectrode-resistance measurement 5.3.16 + Verification of current in the patient circuit is also required at initial verification (see point 5.4). 4
3. Measuring instruments used for verification The measuring instruments listed in Table 2 are recommended for the verification of electroencephalographs. Table 2 Measuring instruments used for verification Measuring instrument Symbol Main characteristics Sine-wave signal generator (*) G1 Frequency range 0.01 Hz 100 Hz Max.frequency error ± 1 % Voltage range 10 mv 10 V RMS Max. voltage error ± 1 % Double-ended output Square-wave signal generator (*) G2 Frequency range 0.01 Hz 1 000 Hz Max. frequency error ± 1 % Voltage range 1 mv 10 V Max. voltage error ± 0.5 % Double-ended output Voltage divider (**) D1 Division factor 10 000 (R1 = 100 kω; R2 = 10 Ω; max.error ± 1 %) Max.error of division ± 2 % Simulated skin-electrode Z1 R3 = 4.7 kω ± 1 % impedance (***) DC voltage source U1 Voltage 1.5 V ± 5 % Length-measuring device Measurement length 1 mm 100 mm Max.error ± 0.1 mm on measuring length from 0 mm to 10 mm, and ± 1 % on measuring length from 10 mm to 100 mm. Magnifying glass Magnification 5 Resistors R4-R12 R4 = 200 Ω ; R5 = 50 Ω ; R6 = 100 kω ; R7 = 620 kω, R8 = 5 kω, R9 = 1 kω ; R10 = 5 kω; R11 = 20 kω ; R12 = 50 kω Max. error ± 1 % Capacitors C1, C2 C1 = 0.5 µf; C2 = 4.7 nf Max.error ± 5 % Impedance Z2 R7 in parallel with C2 (*) If the voltage or frequency of generator G1 or G2 do not meet the specifications of Table 2, a voltmeter and a frequency meter with the accuracy specified for the generator output voltage and frequency should be provided. (**) (***) See Figure 1. Impedance Z1 (called in the text "patient skin-electrode impedance" is intended to simulate the electrical characteristics of the electrode-skin interface. 5
4. Verification conditions and preparation for verification 4.1. The recommended conditions for verification are as follows: ambient temperature: 15 C to 25 C, atmospheric pressure: 96 kpa to 104 kpa, relative humidity of air: 50 % to 80 % (no condensation), mains voltage fluctuations: ± 2 % of nominal voltage, mains frequency: (60 Hz or 50 Hz) ± 2 %. The ranges of atmospheric conditions may be extended for regions with extreme temperatures (or other climatic conditions) or high altitude. 4.2. Deviations of the power-supply voltage should not exceed the values indicated in the manufacturer's manual. 4.3. Before the electroencephalographs are verified, they and the measuring instruments required for verification should be assembled according to the requirements of the national metrology service or other officially authorized organizations. Measuring instruments used during verification should be checked by authorized personnel. 5. Verification 5.1. External examination External examination includes a check that the manufacturer's manual, which should provide the necessary information on the electroencephalograph, is available and includes the following: values of commonly accepted characteristics, their tolerance limits and procedures for their determination, diagrams and construction details necessary for carrying out verification procedures, operating and maintenance instructions, instructions for special medical applications. External examination includes checks for: absence of corrosion and mechanical damage, and freedom from any trace of deterioration of the lead cables. 5.2. Testing The electroencephalograph shall be tested after the warm-up period specified by the manufacturer. The testing shall include checks for the presence and deflection of the recording trace, the pen pressure, either the ink feed or the adjustment of the heating system for thermal pens, the movement of the record at different speeds and the presence of calibration signals and sensitivity controls (step and smooth). 5.3. Determination of metrological characteristics To determine the metrological characteristics each measurement shall be repeated at least three times and each of the measured values shall lie within specified limits. Unless otherwise stated, the electrode selector shall be set so as to apply an input test signal to each channel simultaneously. The recorded output shall be measured in a way that excludes the trace line width. 6
The limiting values of error include both the errors of the EUT (*) and of the reference instruments. If the settings for filters or sensitivities indicated by the manufacturer differ from those specified in this Recommendation the procedure shall be modified to take these settings into account (see footnote, Table 3). The associated diagrams in this Recommendation are given for a 16-channel instrument as an example. The diagrams for other instruments having at least 8 channels are similar. 5.3.1. Determination of relative voltage-measurement error Definition: The relative voltage-measurement error is the difference between the voltage recorded by an electroencephalograph and the voltage applied to its input (whose reference value is taken as a conventional true value), divided by the input voltage. Method of measurement: The relative voltage-measurement error shall be determined directly by measuring the amplitude of the square-wave signal recorded, referring it to the input using the sensitivitysetting, and comparing it with the amplitude of the input voltage as determined by the reference voltmeter (conventional true value). Measurement circuit: The measurement set-up is shown in Figure 1. The appendix presents the table of correspondence between the system of numerical definitions of electrodes and the International "10-20" System. Measurement procedure: The switches of the low-frequency (time constant) and high-frequency filters are set to cover the necessary EEG band width (e.g. 1 s and 70 Hz respectively). The electrode selector (s) is (are) set as shown in Figure 1. The recording speed is set at 30 mm/s. Switch S1 is set at position 2 (Z1 is switched on) and switch S2 at position 3. Generator G2 is set at a frequency of 10 Hz. The sensitivity of all channels is set as specified in Table 3. A square-wave signal from generator G2 and voltage divider D1 with a peak-to-peak amplitude shown in Table 3 is applied to the electroencephalograph's input through the simulated skin-electrode impedance Z1. For each voltage of the input signal and corresponding sensitivity the dimension of the recorded signal is measured. The measurement is repeated in the presence of a constant voltage that simulates the maximum value of electrode potential polarization equal to ± 300 mv, switch S2 being set at positions 1 and 2, in turn. Table 3 Sensitivity-setting and input signal voltage for determining the relative voltage measurement error (**) 5 25 50 100 250 500 Input voltage, µv (peak-to-peak) 10 50 100 200 500 1 000 20 10 200 400 1 000 2 000 Sensitivity-setting, µv/mm 1 5 10 20 50 100 (*) (**) EUT: Equipment under test. If a listed sensitivity is not available on the instrument, the next lower available value may be used with the input voltage scaled down proportionately. See also note (*) page 9. 7
8
Calculation: The relative voltage measurement error, in percent, shall be calculated by the following formula: U m Uin δu = 100 (1) Uin where: U m = h m S n * is the recorded voltage peak-to-peak amplitude, in µv, h m is the peak-to-peak amplitude of the recorded output signal, in mm, S n * is the nominal value of the sensitivity setting, in µv/mm, (*) U in is the peak-to-peak amplitude of the input voltage, in µv. Requirement: For the signals recorded with and without a constant D.C. polarization voltage of ± 300 mv, the error given by formula (1) shall not exceed the value: 10 (1 + U 1 / U in ) (2) where: U 1 is the lowest value of the voltage measurement range, e.g. 5 µv. The commonly accepted polarity shall be checked by applying a negative input signal to input 1 (G1) which shall cause the pens to deflect upwards. 5.3.2. Determination of relative sensitivity-setting error Definition: The relative sensitivity-setting error is the difference between the nominal and measured values of sensitivity, divided by the nominal value. Method of measurement: The relative sensitivity-setting error shall be determined directly by measuring the peak-to-peak amplitude of the recorded sine-wave signal and of the stimulus, calculating the value of sensitivity and comparing it with the nominal value. Measurement circuit: The measurement set-up is shown in Figure 2. Measurement procedure: The filter switches are set as specified in point 5.3.1. The sensitivity is set at 1 µv/mm and the recording speed at 30 mm/s. A sine-wave signal from generator G1 and voltage divider D1 with a peak-to-peak amplitude of 20 µv and a frequency of 10 Hz is applied to the electroencephalograph's input. The amplitude of the recorded signal is measured. The measurements are repeated for sensitivity equal to 5, 10, 20, 50 and 100 µv/mm and for input signals with peak-to-peak amplitude of 100, 200, 400, 1 000 and 2 000 µv respectively. The footnote of Table 3 applies. Calculation: The relative sensitivity-setting error, in percent, shall be calculated by the following formula: * * Sn Sm δ s = 100 (3) * S where: S m * = U in / h m is the measured sensitivity value, in µv/mm, h m is the recorded signal peak-to-peak amplitude, in mm, U m is the input voltage peak-to-peak amplitude, in uv, S n * is the nominal sensitivity value, in µv/mm. n (*) In this Recommendation, the term "sensitivity" is defined by the ratio of the input to the output amplitude, is expressed in µv/mm and is represented by the symbol S *. This definition of sensitivity, commonly used in the field of electroencephalography, is opposite to the definition of sensitivity S in all the other fields (S * = 1/S). 9
10
Requirement: The relative sensitivity-setting error as determined by formula (3) shall not exceed ± 5 %. 5.3.3. Determination of relative time-interval measurement error Definition: The relative time-interval measurement error is the difference between the measured and conventional true values of the input signal period, divided by the latter. Method of measurement: The relative time-interval measurement error shall be determined directly by measuring the period (linear length) of the recorded square-wave signal, dividing it by the nominal recording speed and comparing it with the reciprocal of the input frequency. Measurement circuit: The measurement set-up is shown in Figure 1. Measurement procedure: The switches S1 and S2 are set at positions 2 and 3 respectively. The filter switches are set as specified in point 5.3.1. The sensitivity is set at 10 µv/mm. A square-wave signal from generator G2 and voltage divider D1 with a peak-to-peak amplitude of 200 µv is applied to the electroencephalograph's input. The values of frequency of generator G2 and the recording speed values are taken as specified in Table 4, for different intervals of measurement. In each case no fewer than 3 cycles of the input signal are recorded. It is sufficient to measure top and bottom channels only. Table 4 Generator G2 frequencies to be adjusted and time intervals to be measured Time intervals to be measured, s 5 1 0.5 0.5 0.3 0.2 0.2 0.1 0.05 Generator G2 frequency, Hz 0.6 3 6 6 10 15 15 30 60 Recording speed, mm/s 15 30 60 Calculation: The relative time-interval measurement error, in percent, shall be calculated by the following formula: Tm Tin δt = 100 (4) Tin where: T m = L m /V n is the recorded time interval, in s, L m is the length of 3 cycles, in mm, V n is the nominal recording speed, in mm/s, T in is the time interval corresponding to 3 cycles of the input signal, in s. Requirement: The relative time-interval measurement error determined by formula (4) shall not exceed the value: 5 (1 + T 1 /T in ) (5) where: T 1 is the lower limit of the time-interval measurement range, e.g. 0.05 s. 11
5.3.4. Determination of the relative error of the recording speed Definition: The relative error of the recording speed is the difference between the measured value of the recording speed and the nominal value, divided by the nominal value. Method of measurement: The relative error of the recording speed shall be determined directly by measuring the recorded sine-wave signal period, calculating the value of the recording speed from the generator frequency, and comparing it with the nominal value. Measurement circuit: The measurement set-up is shown in Figure 2. Measurement procedure: The filter switches are set as specified in point 5.3.1. The sensitivity is set at 10 µv/mm, and the recording speed is set at the value to be determined. A sine-wave (or square wave) signal from generator G1 (or G2) and voltage divider D1 with a peak-to-peak amplitude of 180 µv is applied to the electro-encephalograph's input. Measurements are made at speeds of 15, 30 and 60 mm/s. The generator frequency is chosen so that the length of the period of the recorded signal is at least 2 mm. At least 10 cycles are recorded at each speed. Calculation: The relative error of the recording speed, in percent, shall be calculated by the following formula: where: V m = L m /nt e is the measured value of the recording speed, in mm/s, L m is the length of the section recorded for n cycles (n 10), in mm, T e is the period of the input signal, in s, V n is the nominal value of the recording speed, in mm/s. δ v Vm Vn = 100 (6) V Requirement: The relative error of the recording speed as determined by formula (6) shall not exceed ± 5 %. 5.3.5. Determination of recording hysteresis Definition: The recording hysteresis is the distance between the base lines of a trace obtained, respectively, after a positive and a negative input signal is returned to zero (see Figure 3). n Figure 3 Determination of recording hysteresis 12
Method of measurement: The recording hysteresis shall be determined directly by measuring the distance between the baselines, respectively after a positive and a negative input signal is returned to zero. Measurement circuit: The measurement set-up is shown in Figure 4. Measurement procedure: For determination of recording hysteresis a differentiating circuit with the time constant equal to 50 ms (e.g. R6 = 100 kω, C1 = 0.5 µf), is connected between voltage divider D1 and the input. The filter switches are set as specified in point 5.3.1. The sensitivity is set at 50 µv/mm and the recording speed is set 30 mm/s. A differentiated signal of 500 µv and 1 Hz is applied to the electroencephalograph's input, and the recording hysteresis is measured. Requirement: The recording hysteresis hi shall not exceed 0.5 mm. 5.3.6. Determination of relative errors of internal calibrator and time marker Definition: The relative error of the internal calibrator or of the time marker is the difference between the measured and nominal values of voltage output of the internal calibrator or of the time interval of the time marker, divided by their respective nominal values. Method of measurement: The relative errors of the internal calibrator and of the time marker shall be determined by comparing the nominal values of the voltage and time interval of the internal signals of the instrument with the voltage and time interval of a signal applied to the input, the recorded value of the signal being made equal to the recorded internal value. Measurement circuit: The measurement set-up is shown in Figure 1. Measurement procedure: Switches S1 and S2 are set at positions 1 and 3 respectively and the internal calibrator is selected. The filter switches are set as specified in point 5.3.1 with the recording speed of 15 mm/s. Signals are recorded from the internal calibrator and the time marker. Then the electrodes are selected. A square-wave signal from generator G2 and voltage divider D1 with a frequency of 1 Hz and an amplitude equal in value to the internal calibrator voltage, is applied to the input of the device to be verified. Generator G2 voltage and frequency is adjusted so that the dimensions of the recorded signals induced respectively by the internal and external stimuli coincide. The tests are performed at the values of the internal-calibrator voltages and the sensitivities given in Table 5. Table 5 The values of sensitivity to be selected and internal-calibrator voltages to be measured Internal calibrator voltage µv Sensitivity µv/m 1 000 500 200 100 50 20 10 5 2 100 50 50 10 5 5 1 1 1 Calculation: The relative internal-calibrator error, in percent, shall be calculated by the following formula: δ c U cm U cn = 100 (7) U cn 13
14
where: U cm is the measured peak-to-peak amplitude of the internal-calibrator voltage, in µv, U cn is the nominal value of peak-to-peak amplitude of the internal-calibrator voltage, in µv. The relative time-marker error, in percent, shall be calculated by the following formula: where: T m is the recorded value of the time-marker interval, in s, T n is the nominal value of the time-marker interval, in s. δ tm Tm Tn = 100 (8) T Requirement: Neither the relative internal-calibrator error nor the relative time-marker error as determined respectively by formulae (7) and (8) shall exceed ± 2 %. If there is no time marker on the EUT, the EUT shall be considered to have met the relative time-marker error. 5.3.7. Determination of overshoot Definition: The overshoot is the difference between the maximum peak-to-peak amplitude of a recorded square-wave signal and its minimum value, divided by twice the minimum value (see Figure 5). n Figure 5 Determination of overshoot. Method of measurement: The overshoot shall be determined directly by measuring the peak-to-peak amplitude of the recorded square-wave signal. Measurement circuit: The measurement set-up is shown in Figure 1. Measurement procedure: Switches S1 and S2 are set at positions 1 and 3 respectively. The filter switches are set as specified in point 5.3.1. The sensitivity is set at 10 µv/mm and the recording speed is 30 mm/s. A square-wave signal from generator G2 and voltage divider D1 with a peak-to-peak amplitude of 180 µv and a frequency of 10 Hz is applied to the electroencephalograph's input. At least 3 cycles are recorded, and the maximum and minimum peak-to-peak amplitude of each cycle is measured. Calculation: The overshoot, in percent, shall be calculated by the following formula: δ o h max h min = 100 (9) 2h where: h max and h min are the measured values of the linear dimensions of the maximum and minimum peak-to-peak amplitudes respectively of each period recorded, in mm. Requirement: The overshoot as determined by formula (9) shall not exceed 10 %. min 15
5.3.8. Determination of time constant Definition: The electroencephalograph time constant is defined as the time interval required for a recorded square-wave signal amplitude to decay to 1/e (37 %) of an initial value (see Figure 6). Figure 6 Determination of time constant. Method of measurement: The time constant shall be determined directly by measuring the linear dimensions of the decay of the recorded square-wave signal following the overshoot. Measurement circuit: The measurement set-up is shown in Figure 1. Measurement procedure: Switches S1 and S3 are set at positions 1 and 3 respectively. The highfrequency filter switch is set as specified in point 5.3.1 and the time-constant and low-frequency switches are set at a value to be determined. The sensitivity is set at 20 µv/mm and the recording speed is set at 30 mm/s. A square-wave signal from generator G2 and voltage divider D1 with a peak-to-peak amplitude of 180 µ.v and a frequency of 0.1 Hz is applied to the electroencephalograph's input. The length of the segment recorded between the moment of maximum overshoot and the moment when the amplitude of the recorded signal decreases to 37 % of the initial value is measured. The measurements are repeated for all positions of the time-constant switch. Calculation: The time constant, in s, for each of the recorded signals shall be calculated by the following formula: T c = L m /V n (10) where: L m is the length of the segment in which the signal amplitude decreases to 37 % of the initial value, in mm, V n is the nominal value of the recording speed, in mm/s. Requirement: The time constant as determined by formula (10) shall be within ± 10 % of the values specified by the manufacturer. 5.3.9. Determination of amplitude-frequency response curve Definition: The amplitude-frequency response curve is the variation with frequency of the recorded signal amplitude, the input signal amplitude being constant. Method of measurement: The amplitude-frequency response curve shall be determined by comparing the peak-to-peak amplitude of the sine-wave signal recorded at different frequencies with the peak-to-peak amplitude of the signal recorded at 10 Hz. Measurement circuit: The measurement set-up is shown in Figure 2. 16
17
18
Measurement of the pass band: The filter switches are set as specified in point 5.3.1 and the recording speed is set at 15 mm/s for frequencies less than 10 Hz and 30 mm/s for all other frequencies. A sine-wave signal from generator G1 and voltage divider D1 with peak-to-peak amplitude of 180 µv (held constant) is applied to the input, successively at frequencies 0.16, 0.5, 1.5, 5, 10, 15, 30, 60, 75 Hz. In each case at least 5 cycles are recorded and the peak-to-peak amplitudes of the recorded signals are measured. Requirement: The peak-to-peak amplitude of the signal recorded at frequencies between 1 and 60 Hz relative to the peak-to-peak amplitude of a signal recorded at 10 Hz shall be between 90 % and 110 %. Measurement of the cut-off frequencies of the filters: The procedures are the same as those specified above except that the filter switches are set at values to be determined. The generator G1 frequency is set, in turn, at 0.9 F c and 1.1 F c, where F c is the cutt-off frequency value to be determined. The linear dimension of the peak-to-peak amplitude of the recorded signal is measured. The measurements are made for all values of cut-off frequencies of the high-pass and low-pass filters. The entire amplitude-frequency characteristic for all filters should be provided in the manufacturer's manual. Requirement: The linear dimension of the recorded signals shall meet the following conditions: A0.9F c 0.7 A10 A1.1F c A0.9F c 0.7 A10 A1.1F c where: A0.9F c A10 Al.lF c for low frequencies for high frequencies is the linear dimension of the signal peak-to-peak amplitude recorded at a frequency of 0.9 F c, in mm, is the linear dimension of the signal peak-to-peak amplitude recorded at a frequency of 10 Hz, in mm, is the linear dimension of the signal peak-to-peak amplitude recorded at a frequency of 1.1 F c, in mm. 5.3.10. Determination of input impedance Definition: The input impedance is an impedance measured between any two inputs of the device. Method of measurement: The input impedance shall be determined by comparing the linear dimensions of peak-to-peak amplitudes of the sine-wave signals recorded with and without supplementary impedance connected in series. Measurement circuit: The measurement set-up is shown in Figure 7. Measurement procedure: The filter switches are set as specified in point 5.3.1. The sensitivity is set at 10 µv/mm and the recording speed is set at 30 mm/s. Switches S1 and S2 are set at 1 and 3 respectively. A sine-wave signal from generator G1 and voltage divider D1 with a peak-to-peak amplitude of 180 µv and a frequency of 10 Hz is applied to the input with the recorded length of 25 mm. The recorded signal peakto-peak amplitude is measured. Then the measurement is repeated with S1 at position 2. The measurements are repeated in presence of a DC voltage of ± 300 mv, with S2 being set at 1 and 2, in turn. Calculation: The input impedance, in MΩ, shall be calculated by the following formula: Z H 2 in = Z (11) 2 H1 H 2 19
where: H 1 is the linear dimension of the signal recorded with S1 at position 1, in mm, H 2 is the linear dimension of the signal recorded with S1 at position 2, in mm, Z 2 is the value of the impedance Z2 connected in series, in MΩ. Requirement: The input impedance as determined by formula (11) shall not be less than 1 MΩ when the same pair of input terminals is selected by up to 8 channels. 5.3.11. Determination of common-mode rejection ratio Definition: The common-mode rejection ratio is the ratio of the peak-to-peak amplitude of an in-phase signal applied at the input of the EUT to the peak-to-peak amplitude of an out-of-phase signal that results in the same peak-to-peak amplitude of the recorded signal. Method of measurement: The common-mode rejection ratio shall be determined indirectly by measuring the linear dimension of the peak-to-peak amplitude of the signal recorded by the EUT, when a sine-wave signal with the frequency 50 Hz or 60 Hz of a given amplitude is applied in common mode (between input and ground or earth). Measurement circuit: The measurement set-up is shown in Figure 8. Measurement procedure: The filter switches are set as specified in point 5.3.1 and switch S is set at 1. The sensitivity is set at 5 µv/mm and the recording speed is set at 15 mm/s. The peak-to-peak amplitude of the voltage of generator G1 is set at 400 mv with a frequency of 50 Hz or 60 Hz. The peak-to-peak amplitude of the recorded signal is measured. Then switch S is set to position 2 (unbalanced) and the measurements are repeated with a sensitivity of 100 µv/mm. Calculation: The balanced and unbalanced common-mode rejection ratio shall be calculated by the following formula: K U 3 = 10 (12) * h Sn where: h is the linear dimension of the recorded signal peak-to-peak amplitude, in mm, S n * is the nominal value of the sensitivity setting, in µv/mm, U is the peak-to-peak amplitude of the input voltage, in mv (in this case equal to 400 mv). Requirement: The common-mode rejection ratio as determined by the formula (12) shall not be less than 10 4 for the balanced condition and 200 for the unbalanced condition for each channel. 5.3.12. Determination of baseline width Definition: The base line width is the width of the line drawn on the paper, the input of the EUT being grounded. Method of measurement: The baseline width shall be determined directly by measuring it perpendicular to the trace (Figure 9). Figure 9 Determination of baseline width. 20
Measurement procedure: All input terminals are connected to ground (earth) and the filters are set to provide the narrowest band width. The sensitivity is set at 100 µv/mm and the recording speed is 15 mm/s. The width of the recorded line is measured. Requirement: The baseline width shall not exceed 0.5 mm. 5.3.13. Determination of baseline drift Definition: The baseline drift is the deviation of the baseline during a given time interval, the input of the EUT being grounded (earthed) (see Figure 10). Figure 10 Determination of the baseline drift. Method of measurement: The baseline drift shall be determined directly by measuring the baseline deviation during the time interval of 10 min. Measurement procedure: The sensitivity and recording speed are set as in point 5.3.12 and the filter switches are set as in point 5.3.1. The deviation of the baseline in 10 min is measured. Requirement: The baseline drift in 10 min shall not exceed 1.0 mm. 5.3.14. Determination of intrinsic noise level referred to the input Definition: The intrinsic noise level referred to the input is the maximum peak-to-peak amplitude of the signal recorded when the electroencephalograph's input is connected to the skin-electrode impedance Z1; it is evaluated for a given time interval and referred to the electroencephalograph's input (see Figure 11). Method of measurement: The intrinsic noise level referred to the input shall be determined directly by measuring the maximum peak-to-peak amplitude of the recorded signal for a time interval of 60 s and referred to the input through the sensitivity-setting. Measurement circuit: The measurement set-up is shown in Figure 12. Measurement procedure: The inputs are connected to simulated skin-electrode impedances Z1. The filter switches are set as specified in point 5.3.1. The sensitivity is set at 1 µv/mm and the recording speed is set at 15 mm/s. The recording lasts for 60 s. The linear dimensions of the maximum peak-to-peak amplitude of the recorded signal are measured for the entire record and also for 3 (15 mm long) segments i.e. for a time interval of 1 s. In addition the deviation of the medium value of the baseline (noise component less than 0.5 Hz) on the segment of 90 mm (i.e. for 6 s) is measured. 21
Note: In order to avoid external noise pickup, it is recommended that the resistors and switches of the measurement set-up (shown in Figure 12) should be placed in a shielded metal box with its case connected to ground (earth) (pin No. 24 of the terminal box). The resistors (or box containing the resistors and switches) should be connected to the terminal box by unshielded, untwisted leads that are as short as possible but no longer than 1.5 m. Calculation: The noise level referred to the input, in µv, shall be calculated by the following formula: * U N = h n S n (13) where: h n is the linear dimension of the maximum peak-to-peak amplitude of the noise measured on a recording, in mm, S n * is the nominal value of the sensitivity-setting, in µv/mm. Requirement: The intrinsic noise level referred to the input shall satisfy the following conditions: no more than 1 deflection greater than 4 µv in 60 s, no more than 1 deflection greater than 2 µv in 1 s, and the remaining deflections shall not exceed 1.5 µv in 60 s. Figure 11 Determination of the intrinsic noise level referred to the input. 5.3.15. Determination of interchannel crosstalk coefficient Definition: The interchannel crosstalk coefficient is the ratio of the peak-to-peak amplitude of the voltage induced on the channel under test to the peak-to-peak amplitude of the voltage applied at all the other channels. Method of measurement: The interchannel crosstalk coefficient shall be determined by comparing the peak-to-peak amplitude of signals recorded on the channel under test to the peak-to-peak amplitude of signals recorded on all other channels. Measurement circuit: The measurement set-up is shown in Figure 13. Measurement procedure: The filter switches are set as specified in point 5.3.1. The channel under test is connected through simulated skin-electrode impedance Z1 to ground and the sensitivity of this channel is set at 1µV/mm. The sensitivity of other channels is set at 200 µv/mm. A sine-wave signal from generator G1 and 22
23
24
voltage divider D1 with a peak-to-peak amplitude of 1 mv and frequencies manually swept from 10 Hz to 40 Hz are applied to these channels. Linear dimensions of the signal peak-to-peak amplitude recorded at the channel under test are measured. The measurements are repeated successively for each channel. Calculation: The interchannel crosstalk coefficient, in percent, shall be calculated by the following formula: where: i = 1,2,..n is the number of the channel under test, h i Si Wi = 100 (14) U n is the number of channels, h i is the peak-to-peak amplitude of the signal recorded on the channel influenced by the noise, in mm, U in is the peak-to-peak amplitude of the voltage applied to the inputs of the remaining channels, in µv, S i * is the nominal value of the sensitivity of the channel under test, in µv/mm. in Requirement: The interchannel crosstalk coefficient at frequencies from 10 Hz to 40 Hz shall not exceed 2 %. 5.3.16. Determination of relative error of the interelectrode-resistance measurement Definition: The relative error of the interelectrode-resistance measurement is the difference between the indicated resistance value and the conventional true value, divided by the latter. Method of measurement: The relative error of the interelectrode-resistance measurement shall be determined by comparing the data of the incorporated resistance measuring device with those of the resistor connected to the input of the EUT. Measurement procedure: The device under test is set in the mode of measurement of the interelectrode resistance. Resistances of 1, 5, 20 and 50 kω (R9-R12 in Table 2) are applied to the input of the device. Calculation: The relative error of the interelectrode resistance measurement, in percent, shall be calculated by the following formula: Rm R o δ R = 100 (15) R o where: R o is the resistance connected to the input of the device to be verified, in kω, R m is the indicated resistance value, in kω. Requirement: The relative error of the interelectrode-resistance measurement as determined by formula (15) shall not exceed ± 10 %. 5.4. Determination of current in the patient circuit Definition: The current in the patient circuit is defined as the current passing through a patient while connected to an electroencephalograph. Method of measurement: The current in the patient circuit shall be determined indirectly by measuring the voltage recorded by the electroencephalograph with a resistance connected to the input of the EUT and calculating the value of the current. Measurement circuit: The measurement set-up is shown in Figure 14. * 25
26
where: Measurement procedure: The sensitivity is set at 100 µv/mm and the recording speed is set at 15 mm/s. The filter switches are set as specified in point 5.3.1. The base line is recorded, then the linear dimension of the step signal recorded peak-to-peak is measured with S open. Calculation: The current in the patient circuit, in µa, shall be calculated by the following formula: I h r * 3 = (16) S n 10 h is the linear dimension of the step voltage recorded, in mm, S n * is the sensitivity, in µv/mm, r is the value of the resistance R8 connected to the input, in kω. Requirement: The current in the patient circuit as determined by formula (16) shall not exceed 0.1 µa. 6. Verification report and certificate 6.1. The verification report shall contain the following information: (a) the designation and serial number of the instrument verified, (b) the country of origin and manufacturer, (c) the equipment used for verification, (d) the date of verification, (e) the references to the applicable regulations and the procedures used, (f) the measurement results for the various characteristics, (g) the organizations and persons responsible for verification. 6.2. Electroencephalographs verified in accordance with this Recommendation may be granted a certificate of verification in accordance with the established laws and regulations of a given country. The certification shall specify the expiry date of the validity of the verification. 27
APPENDIX Table of correspondence between the system of numerical definitions of electrodes and the international 10-20 System 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Fp 1 Fp 2 F 3 F 4 C 3 C 4 P 3 P 4 O 1 O 2 F 7 F 8 T 3 T 4 15 16 17 18 19 20 21 22 23 24 T 5 T 6 A 1 A 2 F Z C z P z N Figure 15 Electrode lead terminal box. 28
Contents Foreword... 2 1. Scope... 3 2. Characteristics to be verified... 4 3. Measuring instruments used for verification... 5 4. Verification conditions and preparation for verification... 6 5. Verification... 6 5.1. External examination... 6 5.2. Testing... 6 5.3. Determination of metrological characteristics... 6 5.3.1. Determination of relative voltage-measurement error... 7 5.3.2. Determination of relative sensitivity-setting error... 9 5.3.3. Determination of relative time-interval measurement error... 11 5.3.4. Determination of the relative error of the recording speed...12 5.3.5. Determination of recording hysteresis... 12 5.3.6. Determination of relative errors of internal calivrator and time marker... 13 5.3.7. Determination of overshoot... 15 5.3.8. Determination of time constant... 16 5.3.9. Determination of amplitude-frequency response curve... 16 5.3.10. Determination of input impedance... 19 5.3.11. Determination of common-mode rejection ratio... 20 5.3.12. Determination of baseline width... 20 5.3.13. Determination of baseline drift... 21 5.3.14. Determination of intrinsic noise level referred to the input...21 5.3.15. Determination of interchannel crosstalk coefficient... 22 5.3.16 Determination of relative error of the interelectrode-resistance measurement... 25 5.4. Determination of current in the patient circuit... 25 6. Verification report and certificate... 27 Appendix: Table of correspondence between the system of numerical definitions of electrodes and the international 10-20 System... 28 29