CN104771164A - Method utilizing event-related potentials equipment to assist in screening mild cognitive impairment - Google Patents

Abstract

The invention relates to a method utilizing event-related potentials equipment to assist in screening mild cognitive impairment. The method adopts the event-related potentials technology to assist in screening mild cognitive impairment (MCI) and judging clinical intervention effect, adopts the learning-recognizing experimental mode to allow a testee to learn and recognize three emotional faces (positive, neutral and negative) during operation, and also utilizes event-related electrocortical potentials equipment to record the electroencephalogram of the testee. Collected data are calculated according to the unique algorithm, so that four indexes, such as the response accuracy, the N170 ingredient, the early-stage old/new effect and the late-stage old/new effect of the testee, become effective indexes for recognizing MCI, and the method is used for early-stage diagnosis and curative effect determination of MCI, is noninvasive, can be easily accepted by patients and is suitable for clinical popularization.

Description

Events related potentiometer is used to assist the method for examination mild cognitive impairment

Technical field

The present invention relates to a kind of method using event related potential (ERPs) instrument to assist examination mild cognitive impairment.

Background technology

Alzheimer disease (Alzheimer ' s disease, AD) be lasting progressive hypophrenia, and lack effective treatment means, etiology is complicated, and EARLY RECOGNITION is comparatively difficult.Mild cognitive impairment (mild cognitive impairment, MCI) is considered to the commitment of AD, is a kind of Cognitive function damage state between AD and usual aging.Normal person in dull-witted evolutionary process, tested meeting significantly shows intelligence and declines, and most outstanding behaviours is hypomnesis.In the mild cognitive impairment phase, tested meeting shows as and declines to de novo thing memory ability.Patient is often for missing some important thing and occasion and worried, and housewife also can show the taste of cooking and change, and this with the addition of how many flavouring agents on earth because they have forgotten themselves.Along with increasing the weight of of intelligence damage, this memory ability also gradually to remote memory development, until forget All Around The World completely, can comprise children and the companion of oneself.Correlational study shows, and the incidence rate of mild cognitive impairment in old people is 5.3%, about has the patterns in patients with mild cognitive impairment of 10% ~ 15% to be converted into AD every year, and Elderly people crowd is converted into AD every year and is only 1% ~ 2%.Scholar Huang C, Wahlund LO, Almkvist O etc. (2003) carry out the follow-up investigation of 2 years by a definite date to 82 routine patterns in patients with mild cognitive impairments, wherein 28 examples (34%) example is converted into AD, 54 examples (66%) example keeps steady statue, according to this mild cognitive impairment is divided into Advancement Type (Progressive MCI, PMCI) and stable type (Stable MCI, SMCI).The most urgent task how effectively to identify PMCI patient at present, and carry out early intervention to it, delays it and develop to AD.

Research in the past shows, emotional information may preferentially be processed, and when to when being tried the face and Emotional Picture that wake degree equally up, emotion face can preferentially be processed again.That is, individual when the face of observation band emotion, can minimum resource be used, farthest be processed.But this emotion advantageous effect may in the tested middle infringement of MCI, and this infringement is the Early manifestation of MCI before dystropy.And this infringement is arrived by event related potential (Event-related potentials, ERPs) technology for detection.Based on this, we have invented the EARLY RECOGNITION technology of a kind of MCI.

ERPs is the brain Electrical change of generation when carrying out Cognitive Processing to objective event, it has comparatively high time resolution, can the Electrophysiology process that is associated with Cognitive Processing of real time record, and may when behavior problem is also unconspicuous, the abnormal brain electrical activity of visible part cortex neural.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of technology identifying MCI, and namely use event related potential (ERPs) instrument to assist the method for examination mild cognitive impairment, it early can find dementia, for early treatment dementia provides useful help.

For solving the problems of the technologies described above technical scheme of the present invention be:

A kind of method using events related potentiometer to assist examination mild cognitive impairment of the present invention, comprises the steps:

A. according to the mild cognitive impairment diagnostic criteria of Petersen (1999), judge that whether patient is the patient of mild cognitive impairment;

B. programming software E-prime is used to present emotion face picture on computers, allow and can carry out the memory of emotion face suspected of the subjects of mild cognitive impairment, patterns in patients with mild cognitive impairment is allowed to learn three class emotion face (positivity faces respectively, neutral face, negativity face) each 50, allow the tested type of emotion to each face classify when first time study, and the photo of classification error is corrected, browsed revision face photo every 30 minutes later, learn four times altogether;

C. after last has learnt 1 hour, carry out learning effect test, namely test is re-recognized, the face learnt and the face do not learnt are mixed in the ratio of 1:1, present to tested at random, testedly please distinguish one by one, while picture distinguishes task, adopt its brain electricity of events related potentiometer record tested carrying out;

D. in next day, in the same time with the time in step c, the face remembered and new face (do not see above and met) are presented at random in the mixing of 1:1 ratio, allow and testedly again judge that whether this face is browsed.Tested carry out differentiation task while, adopt events related potentiometer record brain electricity;

E. scan 4.3 pairs of eeg datas are adopted to carry out pretreatment, the event related potential composition extracted mainly contains N170, three indexs such as early stage new and old effect (300-500ms) and the new and old effect in late period (500-800ms), and adopt above-mentioned four indexs of Spss statistical package analysis to change in 24 hours;

F. the tested degree of accuracy of re-recognizing is lower, event related potential composition (N170 composition, early stage new and old effect, the new and old effect in late period) to have learnt after 24 hours compared with when just completing, if this tested ratio is less, show its be mild cognitive impairment may be larger.

In described a step, judge that whether patient is the standard following points of mild cognitive impairment:

A) be often worry of forgetting things; B) tested appearance and schooling, age disproportionate memory impairment; C) tested general cognitive function is normal; D) tested activity of daily living is normal; E) dull-witted performance is not had.

In described step c, d, when tested brain electricity collected by use case related potential instrument, major parameter arranges as follows:

Electrode cap is by international 10/20 system electrode installation method electrodes, adopt 64 crosslinking electrode caps, record PO7, PO8, Fz, Cz, the electroencephalogram of Pz electrode, reference electrode is one-sided mastoid process after being placed in left ear, the vertical eye electricity (VEOG) of upper and lower 2 the electrode records of left eye, two 2, outside electrode recording level eyes electricity (HEOG), earth point is positioned at the mid point of Fpz point and Fz point line, during continuous record, cake resistancet leads to for 0.05-100Hz, and sample rate is 500Hz, and the resistance between each electrode and scalp is all less than 5K Ω.

In described step e, carrying out eeg data, in preprocessing process, mainly containing following main points:

Intercept each stimulation and present front 100ms to the brain electricity presenting rear 800ms, with Return Law automatic calibration eye electricity artefact, and reject wave amplitude and exceed ± the brain of 75 μ V electricity, cake resistancet leads to for 0.05-40Hz;

N170 analyzes PO8 and PO7 two electrodes, and average wave amplitude span is between 160 to 240ms.New and old effect is chosen 3 electrodes altogether and is carried out analyzing (Fz, Cz, Pz).Early stage new and old effect analysis time window is 300 ~ 500ms, late period new and old effect analysis 500 ~ 800ms.

In described step f, the tested degree of accuracy at learning phase is as follows with ratio calculation method when just completing after event related potential composition (N170 composition, early stage new and old effect, the new and old effect in late period) has learnt 24 hours:

Tested degree of accuracy (representing with X1): the rate of hitting deducts false alert rate, and (rate of hitting refers to the tested probability correctly re-recognizing the picture learnt; False alert rate refers to that the tested face not learning says into the probability of study);

X1＝p[hits]-p[false positives]

Rate is hit in described p [hits] expression, and described p [false positives] represents false alert rate.

N170 wave amplitude and incubation period (representing with X2): the change of tested electroencephalogram when re-recognizing after having learnt 24 hours of this index reaction, the difference of new face and old face is not considered when N170 calculates incubation period, " old face " wave amplitude learnt is used to deduct " new face " that do not learn, using the end value of its difference as each counter electrode when N170 wave amplitude calculates;

$X2=\frac{[P07(F_A/B_A)+P08(F_A/B_A)]/2+[P07(F_L/B_L)+P08(F_L/B_L)]/2}{2}$

(annotation, in each formula, each alphabetical implication is as follows: wave amplitude-A; Incubation period-L; First time test-B; Second time test-F; Electrode PO7 point-PO7; Electrode PO8 point-PO8; F _arepresent second time test wave amplitude, all the other roughly the same);

Early stage new and old effect (representing with X3): this index dominant response is tested when recognizing the old picture learnt and the new picture do not learnt, in the early stage new and old effect that 300ms-500ms causes, in the original value timing of this index, first calculate the average wave amplitude of each electrode at 300ms-500ms, then by the difference of old face and the new face end value as this electrode:

$\begin{array}{c}X3=[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{PZ}(F_N/B_N)]+[\mathrm{Fz}(F_P/B_P)+\mathrm{CZ}(F_P/B_P)+\\ \frac{\mathrm{PZ}(F_P/B_P)]}{[\mathrm{Fz}(F_c/B_c)+\mathrm{CZ}(F_C/B_C)+\mathrm{PZ}(F_C/B_C)]*2}\end{array}$

(annotation, in each formula, each alphabetical implication is as follows: positivity face-P; Negativity face-N; Neutral face-C; Incubation period-L; First time test-B; Second time test-F; Electrode Fz point-Fz; Electrode Cz point-Cz; Electrode Pz point-Pz; F _nrepresent the average wave amplitude of 300ms-500ms during the test of negativity face second time; Fz (F _n/ B _n) represent at brain district Fz point, the ratio of negativity face average wave amplitude when negativity face was tested with first time during second time test, all the other are similar);

New and old effect in late period (representing with X4): this index dominant response is tested when recognizing the old picture learnt and the new picture do not learnt, in new and old effect in the late period that 500ms-800ms causes.In the original value timing of this index, first calculate the average wave amplitude of each electrode at 500ms-800ms, then by the difference of old face and the new face end value as this electrode;

$X4=\frac{[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{Pz}(F_N/B_N)]+[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{PZ}(F_N/B_N)]}{[\mathrm{Fz}(F_C/B_C)+\mathrm{CZ}(F_C/B_{A.C})+\mathrm{PZ}(F_C/B_C)]*2}$

(annotation, in each formula, each alphabetical implication is as follows: positivity face-P; Negativity face-N; Neutral face-C; First time test-B; Second time test-F; Electrode Fz point-Fz; Electrode Cz point-Cz; Electrode Pz point-Pz; F _nrepresent the average wave amplitude of 500ms-800ms during the test of negativity face second time; Fz (F _n/ B _n) represent at brain district Fz point, the ratio of negativity face average wave amplitude when negativity face was tested with first time during second time test, all the other are similar).

The present invention compared with prior art significant advantage is:

The inventive method has hurtless measure, and expense cost is relatively low, and sensitivity is relatively high, when psychological test and behavior expression also do not occur abnormal, can identify patterns in patients with mild cognitive impairment in time, for dull-witted high-risk group early finds, early intervenes and provides valuable chance.

Accompanying drawing explanation

Fig. 1 is total evaluation flow chart, and this figure reflects in evaluation scheme tested so several stages of experience.

Fig. 2 is learning phase program flow diagram, and this figure reflects the tested concrete steps in learning phase needs experience;

Fig. 3 is test phase flow chart, and this figure is reflected in the concrete steps of test phase individual need experience;

Fig. 4 is ROC curve chart, and this figure reflects the ROC curve chart of four indexs adopting SPSS16.0 to draw.

Detailed description of the invention

The invention will be further elaborated by the following examples.

See Fig. 1 ~ Fig. 4.

The present embodiment is implemented under premised on technical solution of the present invention, gives detailed embodiment and content, but the scope that comprises of the present invention is not limited to following embodiment.

Implementation process mainly comprises six stages (as shown in Figure 1).The first, screening has that the main suit's of amnesia is tested, and the second, carry out necessary psychometry, the 3rd step, carries out the learning and memory of emotion face; 4th step, carries out face and re-recognizes test, and record electroencephalogram after study completes 1 hour; 5th, again carry out emotion face in the next day same time and re-recognize test, and record electroencephalogram; 6th, carry out data analysis and process.

1. screening is tested

Object of study: the tested and 24 routine normal healthy controls groups of 24 routine mild cognitive impairments are chosen in this research altogether.Mild cognitive impairment is tested is mainly derived from the patient gone to a doctor to The First Hospital of Hebei Medical University year February in May, 2012 to 2014, and normal healthy controls group is tested mainly comes from community's volunteer recruiting.

Enter group standard: mild cognitive impairment enters (2004) that group standard meets Petersen and enters group standard about mild cognitive impairment, mainly contain following some: is often a) worry of forgetting things; B) with schooling, age disproportionate memory impairment; C) general cognitive function is kept; D) to take care of oneself daily life; E) dull-witted performance is not had.

Inapplicable colony: a) organic history of being ill; B) psychiatric history; C) severe physical history of disease; D) severe psychiatric diseases history.

2. psychological test

The scoring of clinical dementia rating scale equals 0.5 point; Hachinski ischemic score scale score≤4 point; Mini-mental test (MMSE) scoring is between 25 ~ 27 points.

3. emotion face learning and memory

Stimulus material: altogether employ 300 face photos in the present embodiment, wherein has 1/3rd (100) for positivity face, and 1/3rd is neutral face, and 1/3rd is negativity face.All face photos are selected from Chinese emotion material database (CAFPS) (Wang Yan, Luo Yuejia, 2005), the picture men and women half and half of selection.All pictures unify size, background, color.Photo is presented on the flat-screen CRT monitor of 17 cun, and background is Lycoperdon polymorphum Vitt.All pictures are presented on screen center.

At learning phase, three kinds of Emotional Pictures each 50 (totally 150 are presented to tested, concrete allocation proportion is as shown in table 1, the joyful degree of picture, degree of waking up are as shown in table 2), allow and tested button reaction is carried out to the type of emotion face in picture, in order to increase tested familiarity, every pictures need repeat to see 4 times, and all pictures present at random.Before learning phase starts, tested being apprised of need examine and remember seen face.Learning phase program is (as shown in Figure 2): point of fixation 500ms, there is a picture 3000ms afterwards, it is (glad that the tested classification to this emotion judges, neutral, unhappy), and with right hand forefinger, button feedback is carried out to three keys, there is the blank screen of a 1000ms afterwards, specifically as shown in Figure 2.

4. emotion face re-recognizes test

At test phase, the face learnt and the face do not learnt (comprising three kinds of emotions) are mixed and presents in the ratio of 1:1, allow the tested familiarity to face carry out button feedback.Specific procedure is (as shown in Figure 3): after each unit starts, first "+" point of fixation 500ms is presented, there is a face (terminating after the longest 2000ms of presenting or tested button) afterwards, testedly need carry out button feedback (whether learning) to the familiarity of this face, after tested feedback, occur the blank screen of a 2000ms.

Test unified instruction language, has the exercise of 12 unit before formally re-recognizing experiment and starting.

Table 1 emotion face test material distribution table

Table 2 emotion face material parameter

Adopt the tested test phase electroencephalogram of events related potentiometer record that Neuroscan company produces, Amplifier type is SynAmps2, by international 10/20 system electrode installation method electrodes, adopt 64 crosslinking electrode caps, record PO7, PO8, Fz, Cz, Pz electrode electroencephalogram, reference electrode is one-sided mastoid process after being placed in left ear, the vertical eye electricity (VEOG) of upper and lower 2 the electrode records of left eye, two 2, outside electrode recording level eyes electricity (HEOG), earth point is positioned at the mid point of Fpz and Fz line.During continuous record, cake resistancet leads to for 0.05-100Hz, and sample rate is 500Hz.Resistance between each electrode and scalp is all less than 5K Ω.Use Scan4.5 software analysis EEG data.Programming software E-Prime is adopted to present stimulation and behavioral data collection.

5. date processing

Index one, tested degree of accuracy of re-recognizing.Re-recognizing the stage, experimenter is to the order of accuarcy of face photo new/old decision in degree of accuracy reaction, is the important indicator of reacting tested cognitive function.

X1＝p[hits]-p[false positives]

Hit rate p [hits]: it refers to the tested probability correctly re-recognizing the picture learnt;

False alert rate p [false positives]: refer to that the tested face not learning says into the probability of study.

ERPs analyzes time-histories for stimulating the rear 800ms of appearance, and baseline occurs front 100ms for stimulating.The average wave amplitude of the composition (N170) that ERPs data analysis is early stage and the new and old effect in late period (the new and old effect in frontal lobe district, the new and old effect in top district).N170 analyzes PO8 and PO7 two electrodes, and average wave amplitude span is between 160 to 240ms.New and old effect is chosen 3 electrodes altogether and is carried out analyzing (Fz, Cz, Pz).Early stage new and old effect analysis time window is 300 ~ 500ms, late period new and old effect analysis 500 ~ 800ms.

Before computational prediction index, first pretreatment is carried out to EEG data.Intercept each stimulation and present front 100ms to the brain electricity presenting rear 800ms, with the average wave amplitude of-100-0ms, baseline is corrected.With Return Law automatic calibration eye electricity artefact, and reject wave amplitude and exceed ± the examination of 75 μ V time, cake resistancet leads to for 0.05-40Hz.

Index two, based on N170 wave amplitude and preclinical index.The change of tested electroencephalogram when re-recognizing after having learnt 24 hours of this index reaction.Do not consider the difference of new face and old face when N170 calculates incubation period, when N170 wave amplitude calculates, use " old face " wave amplitude learnt to deduct " new face " that do not learn, using the end value of its difference as each counter electrode.

$X2=\frac{[P07(F_A/B_A)+P08(F_A/B_A)]/2+[P07(F_L/B_L)+P08(F_L/B_L)]/2}{2}$

(annotation, in each formula, each alphabetical implication is as follows: wave amplitude-A; Incubation period-L; First time test-B; Second time test-F; Electrode PO7 point-PO7; Electrode PO8 point-PO8; F _arepresent second time test wave amplitude, all the other roughly the same)

Index three, based on the index of early stage new and old effect.This index dominant response is tested when recognizing the old picture learnt and the new picture do not learnt, in the early stage new and old effect that 300ms-500ms causes.In the original value timing of this index, first calculate the average wave amplitude of each electrode at 300ms-500ms, then by the difference of old face and the new face end value as this electrode:

$X3=\frac{[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{PZ}(F_N/B_N)]+[\mathrm{Fz}(F_P/B_P)+\mathrm{CZ}(F_P/B_P)+\mathrm{PZ}(F_P/B_P)]}{[\mathrm{Fz}(F_C/B_C)+\mathrm{CZ}(F_C/B_C)+\mathrm{PZ}(F_C/B_C)]*2}$

(annotation, in each formula, each alphabetical implication is as follows: positivity face-P; Negativity face-N; Neutral face-C; Incubation period-L; First time test-B; Second time test-F; Electrode Fz point-Fz; Electrode Cz point-Cz; Electrode Pz point-Pz; F _nrepresent the average wave amplitude of 300ms-500ms during the test of negativity face second time; Fz (F _n/ B _n) represent at brain district Fz point, the ratio of negativity face average wave amplitude when negativity face was tested with first time during second time test, all the other are similar.)

Index four, based on the index of new and old effect in late period, this index dominant response is tested when recognizing the old picture learnt and the new picture do not learnt, in new and old effect in the late period that 500ms-800ms causes.In the original value timing of this index, first calculate the average wave amplitude of each electrode at 500ms-800ms, then by the difference of old face and the new face end value as this electrode:

$X4=\frac{[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{Pz}(F_N/B_N)]+[\mathrm{Fz}(F_N/B_N)+\mathrm{PZ}(F_N/B_N)]}{[\mathrm{Fz}(F_C/B_C)+\mathrm{CZ}(F_C/B_{A.C})+\mathrm{PZ}(F_C/B_C)]*2}$

(annotation, in each formula, each alphabetical implication is as follows: positivity face-P; Negativity face-N; Neutral face-C; First time test-B; Second time test-F; Electrode Fz point-Fz; Electrode Cz point-Cz; Electrode Pz point-Pz; F _nrepresent the average wave amplitude of 500ms-800ms during the test of negativity face second time; Fz (F _n/ B _n) represent at brain district Fz point, the ratio of negativity face average wave amplitude when negativity face was tested with first time during second time test, all the other are similar.)

6 results:

As shown in table 3, two groups tested at the age, and schooling ADL mark does not exist the difference with statistical significance, and at MMSE mark, on MoCa mark, normal healthy controls group is greater than mild cognitive impairment group, and difference has statistical significance.

Table 3 basic document

As shown in table 4, the X1 of normal healthy controls group, X2, X3, X4 tetra-indexs are all significantly greater than mild cognitive impairment group, and difference has statistical significance.

The table 4 liang tested each indicator difference of group compares

Spss 16.0 is adopted to draw ROC curve, concrete as Fig. 4, shown in.The design parameter of each index is as shown in table 5.The AUC of X1 is 0.913, and sensitivity is 0.870, and specificity is 0.883; The AUC of X2 is 0.821, and sensitivity is 0.838, and specificity is 0.842; The AUC of X3 is 0.915, and sensitivity is 0.800, and specificity is 0.837; The AUC of X4 is 0.872, and sensitivity is 0.839, and specificity is 0.874; The AUC of X1+X2+X3+X4 is 0.943, and sensitivity is 0.911, and specificity is 0.892.

Table 5 ROC tracing analysis

Above-mentioned example, only for technical conceive of the present invention and feature are described, its object is to person skilled in the art can be understood content of the present invention and be implemented accordingly, can not limit the scope of the invention with this.

Claims (5)

1. use events related potentiometer to assist a method for examination mild cognitive impairment, it is characterized in that, comprise the steps:

A. according to the mild cognitive impairment diagnostic criteria of Petersen (1999), judge that whether patient is the patient of mild cognitive impairment;

B. programming software E-prime is used to present emotion face picture on computers, allow and can carry out the memory of emotion face suspected of the subjects of mild cognitive impairment, patterns in patients with mild cognitive impairment is allowed to learn three class emotion face (positivity faces respectively, neutral face, negativity face) each 50, allow the tested type of emotion to each face classify when first time study, and the photo of classification error is corrected, browsed revision face photo every 30 minutes later, learn four times altogether;

C. after last has learnt 1 hour, carry out learning effect test, namely test is re-recognized, the face learnt and the face do not learnt are mixed in the ratio of 1:1, present to tested at random, testedly please distinguish one by one, while picture distinguishes task, adopt its brain electricity of events related potentiometer record tested carrying out;

D. in next day, in the same time with the time in step c, the face remembered and new face (do not see above and met) are presented at random in the mixing of 1:1 ratio, allow and testedly again judge that whether this face is browsed.Tested carry out differentiation task while, adopt events related potentiometer record brain electricity;

E. scan 4.3 pairs of eeg datas are adopted to carry out pretreatment, the event related potential composition extracted mainly contains N170, three indexs such as early stage new and old effect (300-500ms) and the new and old effect in late period (500-800ms), and adopt above-mentioned four indexs of Spss statistical package analysis to change in 24 hours;

F. the tested degree of accuracy of re-recognizing is lower, event related potential composition (N170 composition, early stage new and old effect, the new and old effect in late period) to have learnt after 24 hours compared with when just completing, if this tested ratio is less, show its be mild cognitive impairment may be larger.

2. a kind of method using events related potentiometer to assist examination mild cognitive impairment as claimed in claim 1, is characterized in that, in described a step, judges that whether patient is the standard following points of mild cognitive impairment:

A) be often worry of forgetting things; B) tested appearance and schooling, age disproportionate memory impairment; C) tested general cognitive function is normal; D) tested activity of daily living is normal; E) dull-witted performance is not had.

3. a kind of method using events related potentiometer to assist examination mild cognitive impairment as claimed in claim 1, is characterized in that, in described step c, d, when tested brain electricity collected by use case related potential instrument, major parameter arranges as follows:

Electrode cap is by international 10/20 system electrode installation method electrodes, adopt 64 crosslinking electrode caps, record PO7, PO8, Fz, Cz, the electroencephalogram of Pz electrode, reference electrode is one-sided mastoid process after being placed in left ear, the vertical eye electricity (VEOG) of upper and lower 2 the electrode records of left eye, two 2, outside electrode recording level eyes electricity (HEOG), earth point is positioned at the mid point of Fpz point and Fz point line, during continuous record, cake resistancet leads to for 0.05-100Hz, and sample rate is 500Hz, and the resistance between each electrode and scalp is all less than 5K Ω.

4. a kind of method using events related potentiometer to assist examination mild cognitive impairment as claimed in claim 1, is characterized in that, in described step e, is carrying out, in preprocessing process, mainly containing following main points to eeg data:

Intercept each stimulation and present front 100ms to the brain electricity presenting rear 800ms, with Return Law automatic calibration eye electricity artefact, and reject wave amplitude and exceed ± the brain of 75 μ V electricity, cake resistancet leads to for 0.05-40Hz;

N170 analyzes PO8 and PO7 two electrodes, and average wave amplitude span is between 160 to 240ms, and new and old effect is chosen 3 electrodes altogether and analyzed.Early stage new and old effect analysis time window is 300 ~ 500ms, late period new and old effect analysis 500 ~ 800ms.

5. a kind of method using events related potentiometer to assist examination mild cognitive impairment as claimed in claim 1, it is characterized in that, in described step f, the tested degree of accuracy at learning phase is as follows with ratio calculation method when just completing after event related potential composition (N170 composition, early stage new and old effect, the new and old effect in late period) has learnt 24 hours:

Tested degree of accuracy, represents with X1: the rate of hitting deducts false alert rate, described in hit rate and refer to the tested probability correctly re-recognizing the picture learnt; False alert rate refers to that the tested face not learning says into the probability of study;

X1＝p[hits]-p[false positives]

Rate is hit in described p [hits] expression, and described p [false positives] represents false alert rate.

N170 wave amplitude and incubation period, represent with X2: the change of tested electroencephalogram when re-recognizing after having learnt 24 hours of this index reaction, the difference of new face and old face is not considered when N170 calculates incubation period, " old face " wave amplitude learnt is used to deduct " new face " that do not learn, using the end value of its difference as each counter electrode when N170 wave amplitude calculates;

$\begin{array}{c}X2=\\ \frac{[\mathrm{PO}7(F_A/B_A)+\mathrm{PO}8(F_A/B_A)]/2+[\mathrm{PO}7(F_L/B_L)+\mathrm{PO}8(F_L/B_L)]/2}{2}\end{array}$

In each formula, each alphabetical implication is as follows: wave amplitude-A; Incubation period-L; First time test-B; Second time test-F; Electrode PO7 point-PO7; Electrode PO8 point-PO8; F _arepresent that second time tests wave amplitude, all the other roughly the same;

Early stage new and old effect (representing with X3): this index dominant response is tested when recognizing the old picture learnt and the new picture do not learnt, in the early stage new and old effect that 300ms-500ms causes, in the original value timing of this index, first calculate the average wave amplitude of each electrode at 300ms-500ms, then by the difference of old face and the new face end value as this electrode:

$\begin{array}{c}X3=\\ [\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{PZ}(F_N/B_N)]+[F_z(F_P/B_P)+\mathrm{CZ}(F_P/B_P)+\\ \frac{\mathrm{PZ}(F_P/B_P)]}{[\mathrm{Fz}(F_C/B_C)+\mathrm{CZ}(F_C/B_C)+\mathrm{PZ}(F_C/B_C)]*2}\end{array}$

(annotation, in each formula, each alphabetical implication is as follows: positivity face-P; Negativity face-N; Neutral face-C; Incubation period-L; First time test-B; Second time test-F; Electrode Fz point-Fz; Electrode Cz point-Cz; Electrode Pz point-Pz; F _nrepresent the average wave amplitude of 300ms-500ms during the test of negativity face second time; Fz (F _n/ B _n) represent at brain district Fz point, the ratio of negativity face average wave amplitude when negativity face was tested with first time during second time test, all the other are similar);

New and old effect in late period (representing with X4): this index dominant response is tested when recognizing the old picture learnt and the new picture do not learnt, in new and old effect in the late period that 500ms-800ms causes.In the original value timing of this index, first calculate the average wave amplitude of each electrode at 500ms-800ms, then by the difference of old face and the new face end value as this electrode;

$\begin{array}{c}X4=\\ \frac{[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{Pz}(F_N/B_N)]+[\mathrm{Fz}(F_N/B_N)+\mathrm{CZ}(F_N/B_N)+\mathrm{PZ}(F_N/B_N)]}{[\mathrm{Fz}(F_C/B_C)+\mathrm{CZ}(F_C/B_{A.C})+\mathrm{PZ}(F_C/B_C)]*2}\end{array}$

(annotation, in each formula, each alphabetical implication is as follows: positivity face-P; Negativity face-N; Neutral face-C; First time test-B; Second time test-F; Electrode Fz point-Fz; Electrode Cz point-Cz; Electrode Pz point-Pz; F _nrepresent the average wave amplitude of 500ms-800ms during the test of negativity face second time; Fz (F _n/ B _n) represent at brain district Fz point, the ratio of negativity face average wave amplitude when negativity face was tested with first time during second time test, all the other are similar).