Classifications

• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
  • A61B5/316—Modalities, i.e. specific diagnostic methods
  • A61B5/398—Electrooculography [EOG], e.g. detecting nystagmus; Electroretinography [ERG]
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
  • A61B5/7235—Details of waveform analysis
  • A61B5/7239—Details of waveform analysis using differentiation including higher order derivatives

Definitions

• the present invention relates to a method as defined in the introductory part of claim 1, as well as to an apparatus for the realization of said method for carrying out an electrooculograpinc (EOG) examination, said method being used for detecting certain types of damages in the retina.
• EOG electrooculograpinc
• the invention relates to a method for speeding up and facilitating to the EOG examination by automatizing it More generally the invention is directed to a method for determining a reference value for a potential leap from a sample signal comprising potential leaps and spurious noise.
• the invention further is related to an apparatus as defined in the introductory part of claim 7, for the realization of the method according to the invention.
• Electro-oculography is an electro-physiological examination centrally related to opthalmology, said examination being based on the measurement of slow fluctuations in the electrical potential differences between the retina (- pole) and the cornea (+ pole). The ratio between said cornea-retinal potentials is called the EOR-ratio.
• EOG is used for the examination of certain degeneration diseases in the retina. It may further be used for ascertaining whether a medication used e. g. for curing rheumatism is damaging the retina.
• EOG examination is considered to be essential in patient examination, but it is also considered to be a good examination method in experimental studies and especially when carrying out retina toxic examinations for medicines.
• the potential cannot be directly measured at the eye. This could be done by arranging an electrode on the cornea, but in this case the eye must be anesthetized, which would lead to erroneous results due to the abnormal function of the eye.
• This problem is avoided by measuring the EOG ratio indirectly using electrodes arranged in the vicinity of the patient' s eyes, whereby mutually different biopotentials, which can be measured, are formed by the movement of the eyes.
• the biopotential ratio for an eye can then be defined by calculation from the mutual ratio of said bio potentials.
• a varying amount of electrodes are arranged on the skin, and further there are numerous different positioning alternatives for them.
• the movement of the eyes is accomplished using e.g. blinking light sources.
• the patient looks at two lignt sources arranged at opposite sides of his eyes.
• the lights blink in opposite phases at a frequency of e. g. 1 Hz, so that the patient's eyes are in a constant movement back and forth.
• a test arrangement is known where a fixed light or some other distinct target is moved in front of the patient' s eyes.
• the potential always varies when the patient moves his eyes.
• the wave frequency of the potential amplitudes thus initiated is identical with the blinking frequency of the light source.
• the size of the potential amplitude is proportional to the EOG ratio, since the potential is zero when the eye looks straight ahead and correspondingly differs from zero when the eye looks sideways or upwards.
• the potential depends on the position of the eye and on the cornea-retinal potential it is impossible to define the real cornea-retinal potential. This fact, however, is unsignificant, since the potential ratio achieved as a result of the examination is more important. It is, however, important that during the examination the eye moves between the same suitably defined points.
• the examination itself is usually started in darkness.
• the potential difference signal is recorded for about 10 seconds once in every minute during a fifteen minute period.
• the signal is not measured continuously, because the continuous moving of the eyes is very exhausting, which would distort the results. Also, the changes in this type of cornea-retinal potential are very slow.
• the cornea-retinal potential decreases slowly when the eye adapts to the darkness. This should happen in about 8 to 9 minutes from the beginning of the examination. After this dark fifteen-minute period the light is changed to a very bright one Then another fifteen-minute period is recorded. The cornea-retinal potential should rise as the eye adapts to the new lighting This takes again about 8 to 9 minutes.
• EOG ratio is defined taking the highest and lowest cornea-retinal potential values and calculating their ratio. If this ratio is below a defined limit the case usually needs closer investigations.
• the object of the present invention is to overcome the disadvantages in the prior art and provide a quite new solution for performing an automatic EOG examination exactly and quickly, in a reliable and even very simple manner.
• Another object of the invention is to provide a functionally simple and easy-to-use method and apparatus for the performance of an EOG examination.
• Yet another object of the invention is to provide an apparatus and method which give the results of an EOG examination in real-time.
• Yet another object of the invention is to provide such an apparatus for the automatic performing of an EOG examination, which apparatus may be realized using easily accessible standard components and a micro processor.
• the invention is based on the idea that the signal obtained from the electrodes is subjected to a computer aided median filtering, i.e. a moving average value is defined from the signal's subsets, the signal favorably being digitized prior to the median filtering.
• the filtered signal is derivated and, using the peak of the derivative signal and a second point defined in accordance with the teaching of the present invention, the real level of the potential leap is defined by integration.
• the system's ability to differentiate between "good” and "poor” signals will be essentially improved.
• the digitalized signal is not continuous but represents the peak value for samples taken per time unit.
• integral is used to express the corresponding relation between two adjacent samples, which behaves like an integral and can be expressed as:
• the method according to the invention is mainly characterized by the features disclosed in the cnaracterizing parts of claims 1 and 6.
• the apparatus according to the invention is characterized by the features disclosed in the characterizing part of claim 7. Other characterizing features are found in the dependent claims.
• an EOG examination of the eyes is performed automatically so that the examiner does not at any stage after the initiation of the examination cycle need to interfere with the examination procedure, but the system according to the invention will independently perform the action cycles of the examination, perform the removal of error signals, analyze the obtained measured results as well as store and print them in a desired form.
• the form of the peak is utilized. After the median filtering the peak should have straight edges. Finding the position of the edges and the peak end is started at the maximum and the samples are searched one after the other, moving forwards away from the maximum until the average of all the samples between the present position and the maximum is half the maximum. Using this position point and the maximum point the end point of the peak is obtained for the calculation of the area. Finally the amplitude size is simply obtained by integrating over the peak area of the curve, whereby the real amplitude neight is obtained.
• the calculation is preferably performed using a computer and it can preferably be made faster by combining the median filter and derivation stages.
• the method according to the invention will give several remarkable advantages. Not only will the analysis of the signals obtained from the electrodes be faster with respect to prior art, but the obtained results will also be more exact and reliable.
• the apparatus needed for performing the examination and the method according to the invention is easy to realize and favorable regarding the costs, since all the components needed are of a type which can be commercially obtained.
• Figs 1a and 1b schematically show the electrical circuit formed by the eyes and they show the potential level when the eye looks straight ahead and to the side,
• Fig 2 is a schematic disclosure of the main idea of the invention
• Fig 3 is a schematic disclosure of an apparatus according to one preferred embodiment
• Fig 4 shows a measured signal
• Fig 5 shows the same signal as derivated but without median filtering
• Fig 6 shows measured results obtained using different filter lengths 1,
• Fig 7 shows sizes of potential leaps calculated from measured signal values
• Fig 8 shows EOG values obtained in one examination.
• Figure la discloses the electrical circuit formed by a human eye, and its potential when said eye looks straight ahead.
• the nose base and the temples are used as measuring points for the biopotential, i.e. as attachment points of the measuring electrodes 12, but other locations may be used too.
• Said electrodes 12 are connected to an amplifier 14, which can be a conventional EKG-apparatus or the like.
• the potential of the circuit is essentially zero when the eye is looking as indicated in the figure.
• a too excessive angle between light sources will cause the patient' s eyes to be rapidly wearied and then the signal will loose its sharp form.
• a too narrow angle will give too low potential leap values from which it is impossible to define the EOG ratio.
• a suitable angle would be about 15 from the center line to both sides of eyes, as indicated above.
• Figure 2 schematically shows the basic idea of the present invention in one of its simplest embodiments.
• the signal obtained from the electrodes 12 is amplified in an EKG-amplifler 14, filtered in a median filtering means 22, which is functionally connected to said amplifier, the correct signal transition is obtained by a calculation means 24, which is functionally connected to said median filter, and said signal is favorably stored and displayed in storage and display means 26.
• Said median filtering means 22 and calculation means 24 may also favorably be combined, as indicated by a line of dots and dashes, whereupon they will simultaneously handle the same signal.
• the operation of the blinking lights 36, 38 and the general lights 34 are controlled by incentive control means 30 All the sequences performed by these means, except for the amplifying phase performed by the amplifier 14, can preferably be performed programmatically using a microcomputer.
• FIG. 3 schematically shows a favourable apparatus embodying the invention, said apparatus being generally indicated by the reference 10.
• Said apparatus comprises electrodes 12, which are functionally connected to EKG-amplifiers 14. Usually one such amplifier is needed for each eye.
• Said EKG-amplifiers 14 are functionally connected to a selector device 16 which preferably is connected to an A/D converter device 18.
• Said apparatus 10 further comprises a computer 20, which is functionally connected to said converter 18 and comprises said median filtering means 22, said transition calculator means 24 as well as said storage and display means 26.
• Said computer 20 further comprises said incentive control means 28, which means are functionally connected to the actual incentive controller 30.
• Said controller 30 controls the function of said general light 34 which is functionally connected to said blinking incentive lights 36, 38 and a relay 32.
• Said amplifiers 14 must be completely isolated, i.e. none of the three electrodes may have a fixed potential. This is very important with respect to the patient security.
• said amplifier 14 comprises three electrodes 12. Two of said electrodes are differential electrodes, between which the potential difference is measured. The third electrode is an active zero electrode which feeds the measured signal back to the patient, which feature will decrease the measured noise level. This especially attenuates the strong 50 Hz signal.
• a test arrangement set up according to Figure 3 comprised two amplifiers, both being Kone 521 EKG amplifiers. In order to reduce the number of electrodes needed the active zero electrodes of each amplifier were interconnected. This is possible since said electrodes are completely isolated. Said test arrangement further comprised a notch filter tuned for 50 Hz.
• the amplification factor for the amplifiers used is 2000. This is quite suitable because a typical potential difference in an EOG examination is about 1 mV. Thus the amplifier's output voltage is between -2V and +2V.
• the interface card for the computer included in the test apparatus comprises several functions when an EOG value is measured. It stores all measured results, controls said light sources and acts as voltage supply for said EKG amplifiers.
• the used computer interface card had the following properties:
• the next task comprises the definition of the size of the potential leaps from the EKG signal.
• This is not any especially difficult task, since humans have a rather developed pattern recognition ability and thus a slight noise in the measured signal does not significantly disturb the recognition process.
• a computer cannot perform such a pattern recognition, and thus mathematical rules based on statistical methods must be deduced for this purpose.
• Figure 4b discloses the signal after derivation, which signal in Figure 4a is disclosed without a median filtering. The derivation is performed in order to define the points where the patient moves his eyes. As evident from Figure 4b a random noise will produce a derivative which is undulating to quite a high extent and which has very little regularity, except for a rather clear initial signal. From this curve it is difficult even for the human pattern recognition ability to find similarities with the actual signal curve. Thus, it is important to reduce the disturbing amount of noise.
• the real peaks come in a regular order, i e. a maximum must follow a minimum and vice versa. Further, it is clear that even if the exact eye movement interval is unknown the number of maxima and minima during the light blinking is in the range from n-1 to n+1. In other words it is known where the peaks should be located, i.e. they should be at the same location where the light source flashes. It must of course be understood that the peaks cannot be exactly at that location since due to the human reaction time the patients eyes react slightly behind the light source. This problem is removed by assuming a window around the location of the peak, the length l of said window being identical to the illumination time for one light source. Further, it is preferable to locate the center of the window slightly behind the theoretical maximum since it is much more probable that the patient will move his eyes slightly after the light source than before it.
• the first window comprises a maximum or a minimum.
• the signal phase can be investigated by modulating it with a "window signal". This is accomplished by first assuming the coming signal to be a maximum, and setting the windows accordingly. Thereafter the signal sample peak values are multiplied with the value 1 in the maximum windows and with the value -1 in the minimum windows. If the average of the values thus calculated is positive the assumption of the first coming maximum was correct, and if the average becomes negative the first window was set at the location of a minimum.
• the next step comprises the exact position of the edges of the triangular area of the peak.
• the rest noise in the signal makes the definition of the signal's edges using the zero points impossible.
• the last edge of the peak will usually be located farther from the zero point than the first edge.
• the form of the peak is used as a help.
• the peak should have straight edges.
• the edges of such a theoretical ideal peak is indicated in phantom line between said maximum X p and zero points ⁇ e . From said figure it is evident that if the average of a local maximum X p and a point X e is half of the maximum value said point X e is a zero point for said peak.
• finding the edges and zero points starts from a maximum value and the samples are analyzed one after the other proceeding away from said maximum, until the average of the investigated samples is half the maximum value. A straight line through this point and said maximum point defines the zero point ⁇ e for the peak, which point is needed in the calculations.
• the area of the triangle peak can easily be calculated. Since the peak represents the derivative for the measured signal the potential leap height can be obtained by integrating over the peak area. This can preferably be done by summing the signal values between said points X e and then the sum obtained corresponds to the height of one leap. This equation can be expressed as: ,
• s and e are the edges of the peak.
• Figure 7 shows as an example the sizes of potential leaps calculated from values for one minute as measured during an actual test series. It is probable that the first value is not reliable. If the measuring series comprise one or more clearly erroneous results like that above, one cannot simply take an average value of the measured series and define it to represent the whole one minute measuring series.
• the average value of the whole set of points is calculated. Thereafter the most remote point is removed. After this the average value is calculated for the remaining set of points and again the remotest point is removed. This procedure is repeated until there are n points left.
• the average of this set of points represents the whole set of points.
• the EOG ratio is finally calculated, printed and stored at some preferred means.
• the method according to the invention brings about an essential improvement.
• a median filtering where the signals firstly are suitably digitized in order to exploit all the benefits of the median filtering.
• Median filtering is namely a very effective way to reduce such random deficiencies which depend on the environment and even on the examined patient's brain functions, which deficiencies in the prior art have constituted a real problem.
• the median filtering or calculating the moving average value of the signals is performed by defining a new value for each new point utilizing the average of its neighboring points.
• This filtering method reduces the noise very efficiently, since the average value for the random noise is zero over the infinite interval.
• the interval used in the practical solutions is not infinite, but the filter still attenuates the noise very well
• the noise amplitude will rougnly be attenuated by a factor l (i.e. the length of the filter).
• the next step in the signal processing is to find the maxima and minima of the derivative signal, and the first problem then is to find local maxima.
• the task would not be difficult if the signal were completely clean but in the signal in practice there usually are small false maxima, due to the fact that the patient's eye sometimes "gets lost", i.e. it does not actually look directly at the light source 36, 38.
• the correct maxima should be found, and this is performed in the manner described above.
• the maxima are found according to the above the edges of the peaks are defined and finally the amplitude size is formed using integration, as above, has been described in greater detail. As the signal in question is a digitized one this can normally be easily performed by summing all signal values between the beginning and end points of a peak value.
• the EOG ratio must still be calculated on the basis of a set of values consisting of several peak values. In this set there probably still will exist also such values which for some reason do not represent typical values but rather should be considered as errors. For this reason those values which are considered to best correspond to the desired properties are separated for the calculation. According to a simple solution this elimination is performed using always smaller subsets so that the average of the remaining set is calculated and the farthest value is removed therefrom until suitably about the half of the original values are left. From this subset of values the EOG ratio is now calculated in a manner known per se.
• the EOG measuring program is written in Microsoft C/C++ 7.00 and requires a computer with at least an 80386 CPU and a VGA graphics adapter. It runs under MS-DOS 3.10 or equivalent.
• the source code is divided into six different files:
• This file is the header file all modules use. It contains all global function definitions and
• Code that controls other modules Determines the execution time of different phases of the program (measuring, calculation, etc.). This is the main program itself.
• tickcount blinklength
• savevalue roundcount, samplenr, gen_status & RECCHNR, sample
• drawcurve gen_status & RECCHNR, sample
• tickcount blinklength
• deinst_clockd()
• deinst_adcard () ;
• systimer _dos_getvect (TIMER_ INTNO) ;
• ptbuffer[0] (inc * ) malloc (sizeof (inc) * (RIGHTEDGE - LEFTEDGE)).
• _setvideomode (_DEFAULTMODE) ;
• stepsize (0] (inc *) malloc (rounds * sizeof (inc)) ;
• stepsize[1] (inc *) malloc (rounds • sizeof (inc)) ;
• cmpcable (inc *) malloc (smpls • sizeof (inc) ) ;
• samplesize sps - MEDFSIZE
• phasediff 3 * blinklength / 16
• tempfile fopen ( "test . log” , "w”);
• trg[i] src(i + MEDFSIZE] - src[i];
• tmptable[i] tmptable[i] - tmptable(i + MEDFSIZE] ; ⁇ /* Change the sign if necessary */
• osum sum - 2L * (long)tmptable (++pos);
• d2 (d2 ⁇ 0) ? -d2 : d2;
• d1 (d1 ⁇ 0) ? -d1 . d1;
• d2 (d2 ⁇ 0) ? -d2 : d2;
• mint[i] peak_ area (findm ⁇ n(b + blinklength / 2)), ⁇
• mint[i] peak_area(findmin(b) ) ;
• maxt[i] peak_area (findmax(b + blinklength / 2));

Landscapes

• Health & Medical Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Heart & Thoracic Surgery (AREA)
• Medical Informatics (AREA)
• Biophysics (AREA)
• Pathology (AREA)
• Engineering & Computer Science (AREA)
• Biomedical Technology (AREA)
• Ophthalmology & Optometry (AREA)
• Physics & Mathematics (AREA)
• Molecular Biology (AREA)
• Surgery (AREA)
• Animal Behavior & Ethology (AREA)
• General Health & Medical Sciences (AREA)
• Public Health (AREA)
• Veterinary Medicine (AREA)
• Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
• Eye Examination Apparatus (AREA)

Priority Applications (2)

Applications Claiming Priority (2)

Publications (1)

Family

ID=8537943

Family Applications (1)

Country Status (4)

Cited By (2)

Families Citing this family (6)

Citations (3)

Family Cites Families (1)

• 1993
  • 1993-05-14 FI FI932222A patent/FI96378C/fi active
• 1994
  • 1994-05-13 WO PCT/FI1994/000191 patent/WO1994026165A1/en active Application Filing
  • 1994-05-13 AU AU66517/94A patent/AU6651794A/en not_active Abandoned
  • 1994-05-13 US US08/586,642 patent/US5823190A/en not_active Expired - Fee Related

Patent Citations (3)

Also Published As

Similar Documents

Legal Events