RU2125399C1 - Device for analysis of psychophysiological information - Google Patents

Abstract

FIELD: medical engineering. SUBSTANCE: device has unit of sensors. Unit outputs are connected to respective inputs of multichannel analog-to-digital converter. Device also includes reproduction unit, outputs of which are connected to display unit inputs, storage assembly connected to reproduction unit and unit of functional converters through busbars. Former of animated picture signals is introduced into device. Corresponding outputs of multichannel analog-to-digital converter are connected to inputs of similar functional converters of unit. Outputs of every functional converter are connected to similar inputs of animated picture signal former, outputs of which are connected to reproduction unit inputs. EFFECT: enhanced reliability. 2 cl, 4 dwg

Description

 The proposed technical solution relates to the field of analysis of the psychophysiological reactions of a person to verbal tests and can find application both for medical purposes and judicial practice.

 The devices of this class include, in particular, polygraphs, stress-based voice detectors, and strain gauge platforms for assessing the psychophysiological state of a person. The purpose of all such devices is to establish the honesty or dishonesty of the subject in answering a specific question asked him. The problem with all devices of this type is opening the latent structure of the test subject. In other words, revealing, through a specially organized experiment, the manifestations of the subject's “reflex of truthfulness” on the verbal impact (information communicated to him in the form of a series of questions).

 This technical solution is aimed at improving one of the most common in law enforcement technical means of honesty-polygraph control.

 A polygraph (lie detector) is usually called a device for analyzing the psychophysiological parameters of the human body, based on the experimentally confirmed, stably reproducible connection between the subject’s attempt to lie and the “small” changes (emotional stress) that are triggered as a result of this attempt.

It is customary to interpret the questions posed to the subject as part of a psychophysiological experiment as tests, and the reflex reactions to these influences of the body of the test subject as a manifestation of what is considered to be a characteristic “truth reflex”
Its essence is that in the continuation of the interval from several units to several tens of seconds after the test question, the body experiences reflex changes in the dynamic characteristics of the pulse, respiration, galvanic resistance of the skin, blood pressure, as well as a number of other parameters directly measurable. At the same time, there are fundamental differences in the nature of reactions to relevant and irrelevant issues. The latter is, in fact, the methodological basis on which the principle of any polygraph is based.

 The effect of verbal tests on the psychophysiological state of a person has been studied in practice in sufficient detail. Since the body’s reactions have a stable course, they can be used as a binary (“yes” or “no”) answer to the corresponding test question, since this answer is not controlled by consciousness, the test result carries complete information about the honesty of the test subject and, therefore, represents interest for use in the practice of investigation and in judicial procedures. Since the polygraph test procedure in some cases can influence the decision-making regarding the fate of a person, the reliability aspect is especially important with regard to the analysis of specific test results.

 A polygraph test (Poligraph Examination) as a method that is used in the course of investigative actions and judicial procedures, is most widespread in the United States, where more than two thousand units of used equipment of this type are registered. The concern of the American society about the reliability of polygraph testing was manifested with the adoption by the US Congress of a special law on protection against polygraphs - the Employee Polygraph Protection Act (EPPA) in 1988. This act makes it possible to protect an employee from arbitrary interpretation, as well as from violation of the confidentiality of test results.

 As for the source of these results, it is impossible to guarantee 100% correct identification of the reactions of a particular subject to all tests in principle. Therefore, it is logical to strive for the maximum value of the posterior probability of making the right decision. For this, it is necessary, firstly, to standardize and unify all polygraph test procedures, in order to minimize the arbitrariness of the personnel servicing the device, and, secondly, to expand as much as possible the analysis base of the recorded reactions of the subject to test influences.

 However, the most complex and difficult tasks are to analyze the reactions of the test subject to test questions. In fact, if you do not delve into the psychophysiological details, the task of analyzing the polygraph data itself does not go beyond the standard procedures used in the statistical processing of multidimensional data samples of any other field experiment. However, one should reckon with the fact that in this case the data source is a person. Therefore, these data are obviously unique and non-averagable (both for the ensemble of sources and for the sample). But this situation calls into question not only and not so much this or that method of statistical processing of test data, but the entire statistical paradigm of decision-making on the psychophysiological reactions of a person to verbal tests. For this technical solution, this position is of fundamental importance, so here you will have to go into some details.

 The minimum of arbitrary actions on the part of the staff is achieved by the formalization and transparency of all training and asking questions. The Questionnaire is prepared individually and represents a series of control, relevant and irrelevant (neutral) questions. For each of the test questions, possible answers are provided only in the form of "yes" or "no." The subject should be familiarized with the list of questions before the start of the test. The list of issues may not include topics affecting freedom of conscience, politics and the sphere of personal life. Inappropriate and degrading personal dignity questions cannot be asked. During the test, any psychological pressure is excluded as a method, questions should be asked monotonously and unarticulated. There is the possibility of termination at any time at the request of the test subject to continue the procedure. Written guarantees of the psychological safety of the test for the test subject on a polygraph, signed by a psychiatrist, are mandatory. In general, the experience gained in operating polygraphs in world practice makes it possible to significantly standardize testing and unify its procedures, minimizing the subjectivity brought by the operator.

 The results of the patent search, discussed below, as well as materials of scientific publications on the problem of polygraph tests, make it possible to judge the undivided dominance of the statistical approach to making decisions on the honesty or dishonesty of the test subject. Common to each of the technical solutions discussed below is the existence of a contradiction between the specialization of reaction detection tools and the unification of their processing methods. For example, skin-galvanic reactions (RAG), respiration, pulse and blood pressure are recorded by fundamentally different types of sensitive elements (sensors), but the electrical signals received from them after analog-to-digital conversion are processed using a single decision rule for all types of reactions. Reactions are recorded in the form of graphs, the deviation of which at some points from certain threshold values is considered in terms of confirmation or non-confirmation of the a priori hypothesis. Each reaction is analyzed separately. The analysis consists in comparing the behavior of the graph in certain areas with a certain reference type of behavior. The difference between the homogeneous curves gives a result that is interpreted as a binary ("yes" or "no") response to the stimulus. Thus, we can say that the essence of all the identified technical solutions of polygraphs is reduced to solving the problem of detecting in a non-stationary random process a signal whose shape is known. And since a problem of this type is similarly solved in various applications, it seems logical to use the standard approach of the theory of statistical solutions to this problem.

 A simple and important rule of the theory of statistical decisions is to compare the posterior probabilities of various hypotheses with each other and take the one with the highest posterior probability. Since polygraph tests involve a choice between two hypotheses, the decision rule has the form of a two-alternative simple hypothesis. We briefly describe the rule formalism.

The symbols h ₁ and h ₂ denote the actual values of those latent variables (true values of the answers) that should be identified during the psychophysiological experiment, and the symbols H ₁ and H ₂ denote the answers that the subject gives during the test session. Otherwise, these answers can be called hypotheses that need verification.

После получения измерительной информации e₁, e₂ от чувствительных элементов полиграфа единственное изменение знаний о состоянии субъекта, которое возникает в связи с этим, состоит в переходе от априорных вероятностей p₁ и p₂ к вероятностям апостериорным p(h₁/e₁) и p(h₂/e₂).The theory of statistical solutions assumes that the probabilities g ₁ = p (h ₁ ) and g ₂ = p (h ₂ ) of the real states of the subject h ₁ and h _{2 are} known a priori, and these states themselves form a complete group of events, so

After receiving the measurement information e ₁ , e ₂ from the sensitive elements of the polygraph, the only change in knowledge about the state of the subject that arises in this regard is the transition from a priori probabilities p ₁ and p ₂ to posterior probabilities p (h ₁ / e ₁ ) and p (h ₂ / e ₂ ).

These numbers are the only real information that the polygraph operator can use to make a decision. Having received the indicated numbers, he compares them with each other and, if p (h ₁ / e ₁ )> p (h ₂ / e ₂ ), then he accepts the hypothesis H ₁ , and if, on the contrary, p (h ₁ / e ₁ ) ≤ p (h ₂ / e ₂ ), then the alternative hypothesis is H ₂ . The meaning of the described decision rule is that a more probable hypothesis is accepted, taking into account the data of a natural experiment.

This decisive rule can also be written in a more general form, using the Bayes theorem:
p (h _i / e _i ) = p (e _i / h _i ) g / i / p (e _i ).

Along with the latter, in the practice of psychophysiological research, other (transformed) forms of writing the expression for the decision rule are used, but the essence is not in the form, but in the fact that the measurements e _i , no matter how large their number, in the end only translate some probabilities into others.

 Let us consider more specifically some of the most important features of the analogues of the proposed technical solution.

 The well-known "System for automated processing of polygraph data" (US, patent, N 5327899, class A 61 B 51/04, 12.7.94), in which the data of the polygraph itself is the input, and the assessment of the probability of an unfair response is the output. The process of processing polygraph signals in a system begins with their analog-to-digital conversion and transformation into some dimensionless form. Further, the signal in a dimensionless form undergoes a standardizing transformation in order to identify subtle features inherent in the responses to each of the questions. The subtle features of the subject's reactions to all relevant questions due to such a transformation are ultimately compared with responses to control questions. Of all the standardized results of reactions to relevant questions, those that correspond to a percentile of 80% are selected. As a result, through the logistic regression model (logit), the data that correspond to the specified percentile is converted into an estimate of the probability of an unfair response. To achieve the task of standardization and transformation into a dimensionless form, this system contains blocks that serve the purpose of suppression in the data of cyclic and aperiodic trends. Data on the skin-galvanic reaction (RAG) and cardiological information are processed to exempt from slowly changing trends, pulse information is obtained by filtering the cyclic component, a derivative is taken from the blood pressure, upper breathing is calculated from the baseline.

 The result of the transformation is the release of the recorded reactions from time to time. In other words, time-dependent quantities, which are random functions in nature, are transformed into a system of dimensionless quantities independent of time. Since the dependence on time disappears, the result of these transformations can be treated as a sample of a random variable and make a decision based on an analysis of the behavior of its distribution function (some upper percentile of this function).

 The disadvantage of this approach is the idea of freeing itself from the trend for the sake of convenience, working with the apparatus of statistics, and not with the apparatus of the theory of random functions, developed by far worse example. The thing is that trend characteristics are, in turn, random functions, and an attempt to exclude them (detrend) without an accurate knowledge of these characteristics produces, most often, "smoothing" of the original signal. As a result, the risk of an error of the 2nd kind (signal skipping) increases. And, although from the point of view of legal practice, this is still better than the risk of a mistake of the first kind (false alarm), which is fraught with an unfair accusation of the subject in dishonesty, the approach itself is associated with an uncontrolled mistake. Even an experienced specialist will not be able to confidently decide after a trend has been removed: was a useful signal or was not removed along with the trend? Note that the very idea of constructing percentiles from the control and relevant experimental data is based on the principle of increasing a posteriori probability (reliability of the conclusion about dishonesty) described above. But, since some editing of the source signals interferes with the formation of the hypothesis of dishonesty, the improvement of the probability estimate with an increase in the number of relevant questions asked may be slower than we would like, or not occur at all.

 It is known "Device for monitoring physiological information" (RU, patent 2010326, class G 06 F 15/46, 10.20.92), which fixes attention on the details that are available in any polygraph (sensors, ADC, processor), but As a methodological basis, statistical analysis is also implied.

 The well-known "Method and system for monitoring physiological parameters measured by sensors on the subject (US patent 5047930, class A 61 F 5/04, 09/10/91)," Method and system for analyzing records of long-period parameters of polygraph test results "(US , patent, 5299118, CL G 06 F 15/42, 03/29/94), "Device for preventing the threat of sudden cardiac arrest" (US patent 5306293, CL A 61 N 1/36, 04/26/94).

 Each of these technical solutions carries out data processing in the form of a multidimensional process. Ultimately, all of these solutions operate on data as a sample, that is, in approximately the same sense as it was analyzed above for US Pat. No. 5,327,899. But data is not always necessarily a sample. In this case, the functional dependence of the data on time and on the parameters that determine the personality is critically important. However, it turns out that it is this dependence that is eliminated as a result of the centering operation of the random process. There is, generally speaking, a paradox: according to its purpose, the device is designed to identify individual personality reactions to certain verbal stimuli, but an indispensable attribute of any of the known approaches implemented in the above technical solutions is the elimination of an individual data form. At the same time, we note that not only centering takes place, the indicated action introduces some systematic subjective edition into the data. Thus, the centering takes place according to a randomly calculated, and not at all the reference level of the analyzed random process.

 A common drawback of the approach based on the statistical processing of information that carries individual characteristics is the risk of losing this personality when moving from time-dependent functions to time-independent values. This risk cannot be reduced to zero by improving the statistical method, since the idea of statistics and the idea of an individual approach are opposite in meaning.

 Thus, a common feature of the identified analogues of the technical solution is to maximize the posterior probability of the decision. In fact, the differences between analogues are reduced only to one form or another of the engineering embodiment of this approach. The result is inevitably probabilistic. It reflects the level of reliability of the conclusion that the deviation of the registered reaction from the reference level is not an accidental fluctuation, but a significant psychophysiological event.

 Another, more particular, but equally inherent feature of all the identified analogues is that all functions - taking, processing, presenting and analyzing measurements - are concentrated in one compact device. This feature requires the gathering in one place and at the same time of the subject, the auditor, experts and technical personnel. This procedure for conducting tests on a polygraph is enshrined in law and even has a certain justification in terms of psychology. However, it should be noted that the presence of the actual processor and technical personnel at this point or in the remote is unprincipled. It is enough that in the same place at the same time there was a test subject, an auditor, data acquisition equipment and registration tools.

 The closest technical solution to this is a device for analyzing psychophysiological information, containing a sensor unit whose outputs are connected to the corresponding inputs of a multi-channel analog-to-digital converter, a playback unit, the outputs of which are connected to the inputs of the display unit, a memory unit connected via buses to the playback unit and a block of functional converters (US patent 5029118, class A 61 B 5/04, 1991).

 With this in mind, the known prototype device has the following disadvantages.

 An increased level of error of the 2nd kind, which entails the risk of missing informationally valuable reactions when the subject answers the relevant questions.

 Leveling individual differences of personalities, creating inconvenience for studying the characteristics of psychotypes.

 Insufficient reliability (probabilistic nature) of the conclusion about dishonesty of the test subject when answering a relevant question, requiring repeated confirmation by varying the form of the question.

 These shortcomings put a limit to the development of the polygraphic method of studying personality characteristics, limit the possibility of reproducibility of an experiment on a polygraph, and reduce the effectiveness of forensic procedures based on the use of a polygraph.

 The technical result of the proposal is to expand the range of the studied signals by analyzing parameters that are hidden and unaccounted for during statistical processing (functional "sensitivity" of the detector), increasing the reliability of the analysis results by reconstructing the dynamic process of the subject's response to the stimulus in the form of a phase trajectory.

 In addition, the device has the ability to ensure the convenience of using the equipment due to the remote diversity of the subject of control, a set of sensors and output equipment for visual monitoring of the reactions of the subject and the equipment for signal processing.

 The technical result is achieved by the fact that in the device for the analysis of psychophysiological information containing a sensor unit, the outputs of which are connected to the corresponding inputs of a multi-channel analog-to-digital converter, a block of functional converters, a playback unit, the outputs of which are connected to a display unit, a memory unit connected via buses to block playback unit functional converters, introduced a signal generator of the animated image, the corresponding outputs gokanalnogo analog-to-digital converter connected to the inputs of similar functional blocks converters, the outputs of each of which are connected to the input of the same name animated image signals, the outputs of which are connected to the inputs of the rendering unit corresponding to the memory unit bus connected to the buses of the same name shaper animated image signals.

 In addition, groups of transceivers are introduced into the device, the connection between the corresponding outputs of the multichannel analog-to-digital converter and the inputs of the functional block converters of the same name and the connection between the outputs of the playback block and the display unit are made in the form of the same transceivers of the first and second groups, respectively.

 The essence of the present invention is reduced at the ideological level to the rejection of the stability postulate of the analyzed process in favor of the dynamic stability postulate. Specifically, this translates into a transition from the analysis of graphs (images) representing certain dependences of psychophysical parameters on time to the analysis of phase portraits (animations), which are dynamic in nature, casually connected sequence of scenes.

 The phase portrait is a conditional image of the behavior not of individual parameters, but of the dynamic system "subject" as a whole.

 The phase portrait does not include the time axis in explicit form. Time enters the phase portrait through animation, dynamic portrait change. In the limit determined by the relaxation time of the processes caused by stress from the question asked, the phase portrait is a loop in N dimensional space, which, having described several turns, returns to the region of space that it left under the influence of the stimulus (another test). The exclusion of time during the transition to the phase portrait does not occur through an artificial method of suppressing slowly changing trends and / or filtering cyclical trends, as described above for analogues, but through the natural expression of synchronously occurring processes through each other.

 The simplest analogue of such a representation is the transition from the deployment of dependency images on the oscilloscope screen using a sawtooth-shaped sweep generator to the deployment of one signal by another. Possible options for this presentation of information, the so-called cognitive presentation, are described in the article by Rakcheeva T. A. "Cognitive representation of the rhythmic structure of the ECG" in the journal "Software Products and Systems", N 2,1992, p. 38 - 47. Various types of presentation measured ECG parameters in the form of a "phase portrait" in which each ECG report is depicted as a point characterized by three parameters (x, y, z), where x (t), y (t) are the coordinates of the point on the plane, and z ( t) is a discrete parameter encoding the ECG values obom, for example, a different color or symbol.

 The main feature of the phase portrait is the possibility of detailing without loss of generality. The movement of the depicting point along the phase portrait will be adequate to the ideas of the subject under study, since the human mental activity is, in a generalized sense, one of the types of matter motion. The type of movement, the form of which is a persistently manifesting psychic phenomenon, will, by definition, be stable. By virtue of the latter circumstance, the parameters of this motion must be measurable and reproducible in a full-scale experiment, which here means a polygraph test. Note that the probability cannot be measured physically, its estimate is calculated as a function of the data obtained in the experiment, it is an indicator that is indirect in meaning.

 Heterogeneous changes can be distinguished not only as a result of calculations, but also visually, by means of graphical display of the phase trajectory (in the form of an animated image). At the same time, the observer captures heterogeneous images that correspond to reactions to the relevant and irrelevant types of questions. In other words, it is not difficult for him to attribute the differences in the dynamic type of behavior of the image point in these two alternative cases. After the expert accumulates certain experience in analyzing the nature of the observed changes, we can talk about the unconditional recognition of two textures of different nature (two- or three-dimensional images).

In FIG. 1 shows one possible embodiment of the device, FIG. 2 shows a variant of the device with a remote arrangement of parts of the equipment. In FIG. 3 illustrates an embodiment of a possible shaper of an animated image signal. The device comprises a sensor unit 1, including sensors designed to measure various physiological parameters, such as respiratory rate, blood pressure, galvanic skin reactions, etc. The sensors 2 ₁ -2 _{n are} connected to the multiplexer 3, the connection of certain sensors can be adjusted if necessary by setting the operating mode of the multiplexer on the control bus 4 ₁ . Information signals from sensors 2 _i -2 _n , by means of a multiplexer, are fed to the corresponding channels of a multi-channel analog-to-digital converter 5, in which each type of measured parameters is converted into digital form for transmitting signals for direct processing and conversion into a form convenient for analyzing the reactions of the subject during testing. In the simplest case, a multi-channel analog-to-digital converter can be made in the form of analog-to-digital converters combined in a block.

 For the first stage of processing the measured parameters of each group of physiological sensors, different functional dependencies are used, and therefore the system contains a block of functional converters 6, each of which is connected to the corresponding channel of the multi-channel ADC 5 (or to the analogous ADC).

 After preliminary processing, the information signals are supplied to the inputs of the same name of the shaper 7 signals of the animated image. This block is designed to convert parameters to the corresponding phase space point on the screen of the display unit.

 The position of the specified point forms a portrait of the animated image, which ultimately represents the visualized multidimensional reaction of the subject to the test.

 To convert the signals received in the shaper 7 into a visual image, a standard playback unit is used on the screen of the display unit 8. The display unit 9 is intended for visual control of the testing process.

 The device also includes a memory unit 10. Communications between units 6,7 and 8 of the device and units by unit 10 are carried out using standard control buses 4, data buses 11 and address buses A.

In FIG. 2 shows an embodiment of the device when the test subject, the sensor unit 1 with ADC 5 and the display unit 9 are in one place, and the processing equipment (blocks 6,7,8 and 10) in another, connected by the transceivers 13 ₁ and 13 ₂ first and the second group with the ADC and the display unit, respectively. In FIG. Figure 3 shows an example of a possible implementation of an animated image signal conditioner.

 The operation of the device on the indicated principle is carried out as follows.

The examiner presents the subject with a sequence of questions. Block 1 of the sensors records the reactions of the test, and block 5 of the ADC implements their presentation in digital form. The signal converted to digital form and transmitted over the communication channel must be converted to a phase trajectory point corresponding to the current time t, each question is processed separately. In general, the process is repeated n times, where n is the number of discrete samples of the signal from the sensor 2 _i (test subject's reaction) after presenting him with the next question of the auditor.

 The parameter signal, for example, of a skin-galvanic reaction (RAG), is digitally supplied to the functional converter 6 of the same name, in which, using, for example, the well-known Weber-Fechner law, the RGR parameters measured in time are converted into a function of the intensity of the irritating factor, which is monotonous and non-linearly increasing logistic function with increasing argument. Another parameter is processed simultaneously and similarly in another Converter unit 6.

 However, in order for the measured parameters to be converted into an animated image, in which on the screen of the display unit one parametric signal was a sweep signal for another parametric signal, and the third signal could, for example, change the color gamut of the image, it was necessary to normalize, as mentioned above, signals so that they are close, for example, in frequency and amplitude, and for this, an imaging signal generator 7 is provided.

 To determine the deployment parameters of a vector in a phase portrait, the device provides a calibration mode, a mode for asking control questions, which can be used to determine with a certain degree of certainty a certain level of the measured parameters when answering relevant and irrelevant questions and the module for changing it for each individual.

 These parameters can be measured on separate outputs 12 from each channel from the ADC block; they are transmitted via buses 12 to the operator console (not shown in the drawing) for use in calculating the specified operating range of the corresponding channels of the animated image signal generator 7. These values can be empirically calculated and entered into the device as driving signals. The response time (reaction) of the subject to the question can also be determined, it has a more or less stable value. This characteristic is used to set the period of the displayed signal.

 Thus, the shaper 7 serves to normalize the measured and converted parameters, taking into account the individual data of the subject, and pairing the parameters of the signals received to reproduce the phase portrait of the subject from different sensors.

A possible embodiment is shown in FIG. 3. Shaper 7 contains multipliers 15 ₁ -15 _n according to the number of sensors (measured parameters), one input of the multiplier is connected to the output of the corresponding functional converter 6, the other input is input 14 ₁ -14 _{n of} setting the magnitude of the factor, which can be calculated as It was indicated above, based on the condition of commensurability of the signals during the sweep and the range of work with each subject. On the control bus 4 set the start mode and the scan period of each signal. The signal from each functional converter is scaled in the shaper 7, as a result of which a “polygram” of psychophysiological information is formed in each processing channel (for each measured parameter). Scalable data can also be sent via bus 11 and address bus to memory node 10 for storage.

 Thus, the “polygram” of various psychophysical parameters is supplied to the playback unit from the animated image shaper.

 In a multidimensional image of a phase portrait, the polygrams form a spatial image tinted with different polygrams, such as color, brightness, sound, etc. receptions. Means for implementing the playback unit are standard computer graphics with multiple driving signals. An example of a possible implementation of the playback unit in the event of a change in the brightness and grayscale of an image is the video processing subsystem, which is a co-processor of the grayscale image processing described by Yu. E. Golyas. and other systems of input and image processing in a PC. Designing hardware. M .: Engineering, 1993, pp. 111-115), the view of which is shown in FIG. 4.

 The coprocessor of this system is built on a functionally modular type and is designed to accelerate the execution of standard procedures, for example, grayscale images. The processed data is an array of codes that can be received via data buses and can be stored in specialized video memory.

 The coprocessor has a set of functionally-oriented blocks that allow regular repetitive operations of transfer and arithmetic-logical data processing.

 The block contains interfaces 16 and 17, the microprogram control block 18, the histogram logic block 19, the block arithmetic-logical pipeline unit 20. Interface 16 is designed to organize the exchange of data between the base PC (in our case, the channels from the output of the animated image forming unit or data bus 11) and the playback unit 8 of the coprocessor. The interface 17 provides communication between the video input module (FRAME) and the histogram logic unit 19, as well as for generating control signals of the display modes.

In addition, through the interface 17, the coprocessor is synchronized from the clock sequence of the input module (f _T ).

 Block 18 is a control device that determines the sequence of microcommands, for example, the signals received via interface 16 integrates them as the addresses of the microprograms stored in the control ROMs of the coprocessor units (not shown in the drawing).

 Control ROMs have a common address field Y, formed by block 18. The sequence of microcommands in the microprocessor is influenced by the status signals received in block 18.

 Block 19 consists of two main devices: histogram memory and histogram logic. The histogram memory is a high-speed operational memory designed to provide hardware support for the coprocessor to perform arithmetic-logical processing of words and data up to 24 bits long. Data exchange is carried out through the following channels: data bus 11 histogram memory of block 19; between video memory and histogram memory; and histogram memory and block 20. Logic is used to calculate the histogram of the array of brightness codes Q (MXM). The calculation results (histogram) are in RAM.

 Block 20 is a high-speed conveyor-type ALU with a pyramidal parallel architecture. Block 20 is designed to accelerate the operation of convolution of an array of Q (MxM) brightness codes with a low-pass filter matrix and a 3x3 high-pass filter. In addition, block 20 allows you to perform a number of additional procedures, such as image overlay, image subtraction, direct image comparison and some other data operations.

 The data transformed in the former 7 or from the video memory (not shown in the drawing) is transmitted via interfaces to the coprocessor (Fig. 4), in which the incoming signals are converted into code signals, for example, of different brightness, as described above. Further, the obtained values are standardly displayed on the screen of the display unit 9.

 If the visually fixed image that has formed in the display unit as a response to verbal influence has not undergone any (other than large-scale) changes, then the subject’s reaction is a psychophysiological answer to an irrelevant question. On the contrary, if the image has become topologically different, then there is a reaction to the relevant question. If, suppose, the initial figure had the form of an ellipse, as in the case of harmonic rhythmograms described in the article by Rakcheeva T. A., then with an irrelevant question, it will remain an ellipse, unless it changes the size of its semiaxes; when the question is relevant, the reaction can take the form of a figure-eight or some other figure - one that cannot be recognized as an ellipse. The result can be obtained by the following algorithm: there is a topological change in the phase portrait - there is a reaction in the sense of a positive reflex response to the relevant question being asked; there are only large-scale changes in the phase portrait - there is a reaction in the sense of a negative reflex response to the relevant relevant question; if changes in the topology of the displayed phase portrait occur in the form of a reaction to an irrelevant question - there is a spontaneous process in the psyche of the subject, indicating the impossibility of further continuation of the session on the polygraph.

 In the case of remote testing, all signals forming a graphic image of the test subject’s reaction to test questions are transmitted through the transceiver equipment to the display unit installed at the site of the polygraph test session. This ensures the removal and display of data in one place, and their mathematical processing in another. Such an organization allows the use of time sharing and parallel processing.

 As noted above, the duty cycle of the device is determined by the question-response interval. The order of questions asked by the auditor to the subject is determined by certain rules, which, within the framework of the description of the principle of operation of the device, does not make sense to describe. It is enough to indicate only the following. The questions that the examiner asks the subject in a predetermined order are divided into three groups: control, irrelevant, and relevant. Control questions allow us to trace the reaction of the subject in a deliberately false answer, because the meaning of the control questions is such that it forces the subject (for moral reasons, for example) to resort to lies. Irrelevant questions make it possible to trace the test subject’s reaction in a knowingly true answer. The meaning of irrelevant questions is such that it obviously does not cause emotional reactions in the subject. Relevant questions relate to the problem of the investigation and are related to possibly hidden information by the subject. The subject's response to a relevant question should answer the question about the latent structure of the object. If the subject’s reaction to the relevant question is similar to that which occurred when answering the control question, then we can conclude that the answer is false; if the reaction is similar to that which was caused by an irrelevant question, then the answer is true. The test subject's answers to the questions of the auditor are recorded together with those reflex responses that they triggered and recorded by the sensors in the form of so-called polygrams - a graphic display of the psychophysiological responses of the subject to the effects of verbal tests (the logging tools are not indicated on the drawing). Block 1 of the sensors is a standard medical equipment that allows with high accuracy to record such, in particular, the parameters of the subject's body, such as pulse, ohmic resistance of the skin, heart rate, blood pressure, spirometric characteristics.

 Valuable information for solving the test problem on a device (polygraph) is carried not by the values of the parameters measured by the sensors themselves, but by their changes. The range of changes of any parameter varies from person to person, but in principle is subject to the well-known Weber-Fechner law.

 Suppose, for example, this particular sensor 21 measures the so-called skin-galvanic reaction (RAG). Any question causes the subject this reaction in the form of an unsteady process that develops within a few seconds after the auditor presents the question to the subject.

 The reaction to irrelevant and relevant questions proceed along different paths. In traditional technical solutions, the difference between certain comparable points of these trajectories (they have special names, which do not make sense to indicate in the description of the principle of operation) are a measure of the proximity of an expert’s conclusion to one or another alternative option. The expert who analyzes the RAG polygram shows that the trajectory of this reaction, recorded on the loop diagram, goes high or low, concludes whether the answer to the relevant question is false or true.

 At the same time, the test subject’s reaction to control and irrelevant questions gives the expert a “training” sample of data, as they set the range of individual reactions of this subject in the sense of the Weber-Fechner law. Essentially, through these questions, the subject's response scale is calibrated. When analyzing reactions to relevant questions, an expert already knows where one end of the range (true) and the other (false) are located.

 The problem of traditional technical solutions, as already noted, is that individual differences “blur” the boundaries of this range and deprive the test results of certainty, which, in turn, forces experts to resort to the formation of a sample of the same reactions and its analysis by a statistical method.

Let the subject be asked an irrelevant question. Then the signals from the sensor 2 _i RAG and all other types of sensors will collectively describe a stable psychological process. This means that the topological structure of the phase portrait of this process will be unchanged for the reaction to all irrelevant questions. In other words, the image that is formed on the screen or other graphic means as a result of the deployment of one polygram (for example, RAG) with the help of another (for example, breathing) will be simply connected. The appearance of the image (two-dimensional texture) can undergo large-scale changes (shrink or stretch along the coordinate axes) when moving from one irrelevant question to another, but cannot change in a topological sense. Meaning: a circle can turn into an ellipse, but cannot become a figure-eight.

 If a two-dimensional image (in this example, RHR in the breathing function) is deployed using another parameter, for example, blood pressure, the result of this transformation will have the form of a three-dimensional texture. All arguments retain their previous meaning: when moving from one irrelevant question to another, a ball can turn into an ellipsoid, but cannot become a cubic (or conical) figure.

 The multi-channel control allows duplicating reactions to questions, but when analyzing the visualized representation of these reactions, the phase portrait is limited to three-dimensional Cartesian space. If you need to increase the number of simultaneously controlled reactions to 5-7, you can resort to either constructing “parallel” three-dimensional textures from different sets of controlled parameters, or, in addition to purely geometric images of the reaction intensity on the display means, use visual means such as a color scale, height sound tone, frequency of flashes, the speed of the image, and other techniques for constructing animated images. For the described purposes, for example, software tools for multimedia presentation of multidimensional information, which are equipped with modern personal computers, can be used.

 Let the subject be asked a relevant question. Then, if the answer is true, then the texture of the image (phase portrait) will not undergo changes in comparison with a series of control questions. On the contrary, if the answer contains a false statement, then the stressful changes caused by it in the psyche of the subject will be recorded by the sensor equipment, and the changes revealed by it will demonstrate the loss of stability of the course of the mental process. The emergence and development of instability over time will be displayed by the topological (qualitative) deformation of the phase portrait. In other words, a two- or three-dimensional texture that is completely different in nature will arise, which differs from that which was fixed after asking irrelevant questions, not only by large-scale changes, but also by a change in the topological nature of the image.

The loss of stability caused by stress resulting from an untruthful answer to a relevant or control question is temporary. The process ends as a result of coming to an equilibrium state, from which a new question is deduced, which the examiner asks the subject at time intervals, allowing him to trace the development of the psychophysiological reactions of the subject in the necessary details. Thus, a series of different types of questions causes a series of responses that are recorded by the sensor equipment and, being converted into discrete samples in the ADC, arrive at the block of functional converters. Next, in the shaper 7, the signal is modulated by discrete samples that carry information about the psychophysiological reactions. In this block, signals are generated by which specific values of the animated image are set (dimensions, relative scales, color, etc.)
That is, taking into account the type of parameter and the ranges of truthfulness and falsity in it, the signals are scaled and each signal (polygram) is identified with the parameters for changing the animated image.

 The playback unit 8 receives the control signals from the shaper 7 and turns them into screen forms. In addition, this block involves applying labels and inscriptions to the image that accompany the testing process, in particular questions and answers, expert comments and other information. This block can be a standard element of a personal computer, including the necessary software.

 Thus, as follows from the description of the operation of the proposed device, its technical solution not only achieves the goal of increasing the resolution of the analysis, but also simplifies the problem solved by the operator. The operator (expert), who is forced to observe up to 5-7 loop plots in the usual approach, in this case receives for review and analysis a holistic image (the so-called "gestalt"), the dynamic properties of which are no longer weakly distinguishable against the background of an unsteady process, as in the first case, and the essence is the product of a change in the state of the subject, easily distinguishable at a qualitative level.

The location of the equipment in different places creates convenience in terms of accumulating a single archive of test results, as well as the ability to promptly request test data from the same subject at another time and / or in another place. For example, in order to obtain additional confirmation for the purposes of authentication or double-checking the newly appeared information about the subject. Thus, according to the principle of operation, this device is a multimedia polygraph using teleconferencing capabilities (Fig. 2)
In addition to the functions of the polygraph itself, this device can, in principle, be used for some diagnostic tasks for medical purposes.

Claims (2)

 1. A device for analyzing psychophysiological information, comprising a sensor unit, the outputs of which are connected to the corresponding inputs of a multi-channel analog-to-digital converter, a functional converter unit, a playback unit, the outputs of which are connected to the inputs of the display unit, a memory unit connected via buses to the playback unit and unit functional converters, characterized in that the device implements the signal generator of the animated image, the corresponding outputs gokanalnogo analog-to-digital converter connected to the inputs of similar functional blocks converters, the outputs of each of which are connected to the input of the same name animated image signals, which outputs are connected to inputs of the rendering unit corresponding to the memory unit bus connected to the buses of the same name shaper animated image signals.  2. The device according to claim 1, characterized in that the connection between the corresponding outputs of the multi-channel analog-to-digital converter and the inputs of the same functional converters of the unit and the connection between the outputs of the playback unit and the display unit are made in the form of the same transceivers of the first and second groups, respectively.

Images