RU2717214C1 - Method of assessing the reliability of reported information and device for its implementation - Google Patents

Abstract

FIELD: evaluation of information reliability.

SUBSTANCE: group of inventions relates to the field of criminalistics, examination, including forensic examination, and other fields of knowledge associated with assessment of reliability / accuracy of data and information and can be used in law, especially for obtaining evidentiary information in legal proceedings. Disclosed is a method of assessing reported information, where effect is carried out on a test block of neutral, control and verification stimuli, fixing parameters of motor responses to stimulation, study of dynamics of change of parameters, which are grouped on the basis of factor analysis into group of reactions. Method also involves detecting a trend of changes in each parameter and / or group of parameters and evaluating the reported information by comparing responses to test and control stimuli. Motor reactions are recorded by scanning. Detection of trends is carried out according to four types of response, selected based on dominant trends: by maximum and minimum parameters of responses to blocks of stimuli of control nature, ascending and descending reaction parameters on blocks of stimuli from neutral to check, wherein the stimulus action is performed by sets of blocks, after the exposure to the neutral block, the neutral, control and verification units are subjected to a chaotic action by a different number of units. Authenticity is estimated automatically in response to response types. If the reaction parameters prevail on the control stimuli units over the response parameters to the test stimuli, the reliability of the reported information is stated. If the reaction parameters prevail on the test stimuli units over the response parameters to the control stimuli, the unreliable information is stated to be false. Disclosed is a device for realizing a method, characterized in that it comprises an active unit based on a modular structure, comprising a microprocessor and a contactless sensor, connected to a microprocessor programmed to process signals from the non-contact sensor(s), unit for automated evaluation of transmitted information, to which active unit is connected.

EFFECT: group of inventions provides more objective evaluation of reported information with simplification due to exclusion of routine labor.

10 cl, 8 dwg, 6 tbl

Description

The group of inventions relates to the field of forensics, forensics, including forensic science, and other branches of knowledge related to the assessment of the reliability / accuracy of data and information and can be applied in the field of law, especially for obtaining evidence in legal proceedings.

The problem of assessing the reported data (information) is one of the most relevant for many areas of human activity (politics, economics, legal practice, etc.). Despite this, there are currently no reliable methods for evaluating the reported data with the establishment of these properties. Currently, to solve the problem of determining the reliability / inaccuracy (error / error) of information, the instrumental method based on the use of contact multifunctional biotechnological measuring and computing complexes [MBIVK (polygraphs]], which has low physiology due to significant methodological limitations, is mainly used. The main problem of MBIVK (polygraphs) is the low reliability of the results, since the conclusions cannot be confirmed by large arrays of measurements of comparative studies when revealing dominant trends indicating the reliability (error-free) or inaccuracy (fallacy) of the conclusions.

The current practice in assessing the reported information is focused on the concept of reliability. Reliability within the framework of special issues of situational information analysis (SIA) of data is associated with the identification of relationships, which implies the establishment of the likelihood of a conclusion based on accepted criteria. At the same time, “reliability” should be associated with the conceptual apparatus of the relevant GOSTs governing the work with information and data. Thus, the reliability of SIA is understood as error-freeness, respectively, inaccuracy - data error (see details: Quality of service information. Terms and definitions: GOST R 51170-98. - Official publication. - M.: IPK Standards Publishing House, 1998. - P. 74 (6).

A known method of controlling the psychophysical reaction of a person and a device for its implementation (see RF patent No. 2216269, IPC A61B5 / 04, A61B5 / 16, publ. 20.11.2003). The method measures the shift of the resonant frequency of the oscillatory circuit of the generator of electrical vibrations, made with the possibility of smoothly adjusting the frequency by placing at a distance of 20-50 cm from the test piezoelectric transducer included in the oscillatory circuit of tuning the frequency of the generator. This shift is measured with a change in the psychophysical state of the test subject, stimulated by verbal influence, creating a positive or negative attitude, while the psychophysical reaction is judged by the magnitude and nature of the shift in the resonant frequency.

A device for recording a psychophysical reaction includes an electric oscillation generator configured to smoothly change the frequency, a quartz based piezoelectric transducer with an intrinsic resonant frequency in the range of 17-20 MHz, included in the oscillator frequency tuning circuit of the generator and placed at a distance of 20-50 cm from the subject, and a frequency meter for measuring the frequency of the oscillatory circuit. The device is made with the possibility of shielding from external electromagnetic radiation.

The disadvantages of this method are:

- use only to determine the psychophysical reaction at different levels of a person’s state;

- the impossibility of its use in the field of law (law enforcement and the system of legal relations), forensics, procedural and non-procedural studies, forensic examinations related to the assessment of reported data;

- the impossibility of using it to establish contactless motor activity in assessing the reliability / inaccuracy (error / error) of the reported data;

- the inability to use the method for conducting a direct expert assessment carried out by an experienced person;

- only one piezoelectric transducer;

- the impossibility of using other proximity sensors operating on other physical phenomena and processes;

- restrictions on the placement of the sensor at a distance of 20-50 cm;

- only verbal impact on the subject;

- the need to create a positive or negative attitude;

- use only to determine the psychophysical reaction at different levels of a person’s state.

A known method of non-contact optical measurement of vibration parameters of biological objects (see RF patent No. 2546714, IPC G01H9 / 00, publ. 04/10/2015). In the method, a video signal of the test object is formed using a video camera, a subsequent digitization of the specified signal is performed using an analog-to-digital converter, a digital video image is processed using a computer, and information on the absolute vibration parameters of the points of the test object on the video image is obtained.

The disadvantages of this method are:

- lack of an integrated approach to the study of the object;

- the inability to use to establish contactless motor activity when establishing the reliability / inaccuracy (error / error) of the reported data;

- obtaining information only about the absolute vibration parameters of the entire (integral) object;

- use only for vibration diagnostics, measuring vibration parameters of an object;

- signal generation based only on a video image;

- lack of analysis of motor activity in various anatomical and physiological zones of a person.

A known method and device for analyzing human behavior (see RF patent No. 2292839, IPC A61B5 / 04, A61B5 / 16, publ. 02/10/2007). The method includes the following steps: one or more measurements or observations of a person; encoding measurements or observations to form multiple channels; and channel analysis using artificial intelligence to derive information related to human psychology. The device comprises detection means for making one or more measurements or observations of a person, encoding means for encoding measurements or observations with the formation of multiple channels, and automatic machine classification means configured to analyze channels and output information related to human psychology. In this case, the automatic machine classification tool contains an artificial intelligence tool or a statistical classification tool. The invention allows to quickly obtain objective information on the analysis of human behavior.

The disadvantages of this method and device are:

- the impossibility of using it to establish contactless motor activity when establishing the reliability / inaccuracy (error / error) of the reported data;

- the impossibility of its use in the field of law (enforcement and legal relations), forensic science, procedural and non-procedural studies related to the assessment of reported data (use only for analysis of human behavior related to human psychology);

- the need to use the artificial intelligence method, artificial neural network, - machine learning, genetic algorithm and fuzzy logic;

- receiving the initial signals based on the use of the camera and microphone only in the device.

A known method for determining the falsity of transmitted information on the dynamics of the parameters of non-verbal behavior of a person (see RF patent No. 2506048, IPC A61B5 / 16, publ. 02.10.2014). In the method, a structured interview is conducted, including blocks of questions of a neutral and control nature, while in response to questions, parameters of non-verbal behavior of a person are recorded. Then, the dynamics of non-verbal behavior is studied with the selection of its most informative parameters, grouped on the basis of factor analysis into a group of vegetative reactions, a group of mimic reactions, a group of pantomime reactions, revealing a tendency for changes in each parameter and / or group of parameters characteristic of the person being studied. Then, the parameters of non-verbal behavior of subscribers in the process of communicative interaction are recorded by means of an infocommunication system, by which they conclude that the transmitted information is false / true by comparing the received reaction with the response to questions of a neutral and / or control block of questions recorded during a structured interview.

The disadvantages of this method are:

- the impossibility of using it to establish the reliability / inaccuracy (error / error) of the reported data based on an analysis of the dynamics of motor activity in a non-contact manner;

- the lack of a standardized automated calculation of the number of motor reactions, taking into account the differentiation according to their severity based on IT technologies;

- subjectivity in assessing the severity of reactions based on the assignment of points by the person conducting the study;

 - the possibility of use only to establish the falsity / truth of the transmitted information;

- the lack of the possibility of presenting incentives by demonstrating visual images on the subject.

A known method of microkinetic psychosemantic detection (see RF patent No. 2425629, IPC A61B5 / 16, published on 08/10/2011). The method presents with an exposure no longer than 40 ms, with a different interval in a masked unconscious form of two or more categories of semantic stimuli combining synonymous or similar in meaning words, short phrases or images, the subjective significance of which the experimenter is interested in, and one control category from semantically neutral stimuli in the form of a random set of numbers. Unrecognized behavioral responses are recorded in response to unconscious stimuli, followed by a statistical comparison of these reactions in order to identify the subjective significance of the unconscious stimuli by which operator readiness is maintained. In this case, the response to unconscious stimuli is recorded in the form of micromotion superimposed on the stereotypical movements of the subject's hand in response to masking stimuli. The movement is recorded with sampling from 1 to 40 ms, followed by software processing of the motion image: at the points of start, end of movement, upper and lower peaks, intermediate peaks, change of direction of movement in the vertical and horizontal planes, speed and spectral composition of vibrations in vertical and horizontal planes. In addition, the thesaurus of unconscious stimuli may include a category of semantically neutral stimuli in the form of randomly generated meaningless series of numbers presented between semantically significant stimuli in order to avoid the effect of interference and prolonged aftereffect of stimuli. The source of information about each unconscious reaction of a test subject can be a two-dimensional vector of several hundred point coordinates that fix unique hand trajectories in response to a stimulus. Based on the obtained micromodulations, hand movements can compare reactions in response to stimuli grouped according to belonging to certain categories or to individual subthreshold stimuli, thereby revealing their unique semantic groups.

The disadvantages of the method include:

- the impossibility of its use for the analysis of non-contact motor activity of various anatomical zones of human motor activity, various severity and direction in establishing the reliability / inaccuracy (error / error) of the data reported;

- the impossibility of using in the field of law (law enforcement and the system of legal relations), forensics, procedural and non-procedural studies, forensics related to the assessment of reported data;

- application only for revealing the subjective significance of unconscious semantic stimuli, combining synonymous or close in meaning words (short phrases) or images with which operator readiness is maintained;

 - the use as a control category of stimuli of semantically neutral stimuli in the form of a random set of numbers;

- registration of responses only in the form of micromotions (micromodulations of arm movements) superimposed on the stereotyped movements of the subject’s arm in response to masking stimuli with a sampling rate of 1 to 40 ms.

Closest to the claimed method is a method for diagnosing falsity of the reported information on the dynamics of parameters of non-verbal behavior of a person based on a structured interview (see RF patent No. 2415645, class IPC A61B5 / 16, publ. 04/10/2011), while a structured interview includes blocks of questions : neutral, control and verification character. When answering questions, parameters of non-verbal behavior of a person are recorded, then the dynamics of non-verbal behavior is studied with the selection of its most informative parameters, grouped on the basis of factor analysis into a group of vegetative reactions, a group of mimic reactions, a group of pantomime reactions, identifying the trend of changes in each person a parameter and / or a group of parameters by which a conclusion is made about the falsity / truth of the information being reported by comparing the reaction to questions of the test block with a reaction to questions of a neutral and / or control block of questions.

The disadvantages of the prototype method include:

- the impossibility of its application in the field of law when using special knowledge in law enforcement and the system of legal relations, as well as forensic science, examination (including forensic examination) and other branches of knowledge related to the assessment of data and information;

- the inability to use to assess the reliability (infallibility) of data (information, information) or their inaccuracy (error);

- the impossibility of a random arrangement of stimulus blocks in the stimulus system and stimuli in blocks, the use of several fragments based on different formats with a different number of stimulus blocks;

- the absence of verified signs of reaction identified on the basis of a population study;

- the inability to systematize the results on the basis of the types of response and the production of calculations of the reliability of conclusions based on the criteria of large statistical groups;

- the impossibility of producing automated fixation of reactions and their automated assessment, including a combination of direct expert assessment carried out by a competent person in the video image with an automated assessment based on the claimed device with a generalized conclusion on the totality of the estimates made;

- the inability to include in the analysis of a verified sign that reflects the motor function of the vocal muscle.

A device for diagnosing falsity of reported information on the dynamics of parameters of non-verbal behavior of a person (see RF patent No. 2577934, IPC A61B5 / 0205, A61B5 / 0255, publ. 03.20.2016). The device comprises a first video camera installed with the ability to remove video information from the patient’s face, a second video camera installed with the ability to remove video information from the patient’s feet, a third video camera installed with the ability to remove video information from the patient’s body, a highly sensitive microphone installed near the patient’s mouth. Camcorders and a highly sensitive microphone are connected to the electronic unit through blocks of significant differences. The electronic unit is configured to analyze the information received from the units of significant differences with a judgment on the falsity of the information provided. The block of significant differences is made with the ability to trace the dynamics of changes in the parameters of non-verbal behavior of the patient according to the algorithm specified in the device.

The disadvantages of this device are:

- this device is a contact, which limits its use in law enforcement;

- fixation of motor activity is determined based on the use of a coordinate manipulator (computer mouse, joystick, trackball, etc.) when practicing a special target drawn on a computer screen;

- the impossibility of using it to establish contactless motor activity in assessing the reliability / inaccuracy (error / error) of the data reported (used only to make judgments about the falsity of the information provided);

- the lack of an assessment algorithm that operates on the basis of verified features identified in population studies and large statistical groups;

- the use of a large number of cameras (at least three).

- the impossibility of using in the field of law (law enforcement and the legal relations system), forensics, procedural and non-procedural studies, forensic examinations related to the assessment of reported data.

- the presence in the device of a predetermined algorithm according to which the dynamics of parameter changes is monitored, without the possibility of its adaptation (change) taking into account specific tasks set for solving by an expert;

- the absence of other technical channels for obtaining information about the investigated object.

The technical problem of the claimed group of inventions is the development of a method and device for differentiated assessment with the establishment of reliability (error-free) and inaccuracy (error) of the reported data (information, information) based on non-contact fixation of motor reactions against a background of diverse stimulus in the process of monitoring in automatic mode.

The technical result of the invention is to increase the objectivity of evaluating the information provided while simplifying by eliminating routine labor.

The technical result is achieved by the fact that in the method of evaluating the reported information, including the impact on the subject with blocks of stimuli of a neutral, control and verification nature, fixing the parameters of motor reactions to the stimulus, studying the dynamics of changes in parameters grouped on the basis of factor analysis into a group of reactions, identifying the trends for each parameter and / or group of parameters and the assessment of the reported information by comparing responses to stimulus blocks of the verification and control x A character, according to the invention, fixes motor reactions by scanning at least one anatomical zone online, using three groups of verified reaction parameters as reaction groups: multilevel motility, facial motility, physiologically determined motility, and four types of trends can be identified responses identified on the basis of dominant trends: maximum and minimum response parameters to control stimulus blocks, ascending and descending parameters of reactions to stimulus blocks from neutral to test, while the stimulus is performed by sets of blocks, after exposure to a block of a neutral character, blocks of a neutral, control and test character are randomly generated with a different number of blocks, the reliability is evaluated automatically taking into account the types of response : with the prevalence of reaction parameters to control stimulus blocks over reaction parameters to test stimulus blocks Character of make a conclusion about the reliability of the reported information, with the predominance of reaction parameters on the nature of the verification blocks incentives over reaction parameters on the control units based incentives conclude unreliability of the information reported.

Scanning of motor reactions is carried out using non-contact motion sensors.

Fixation of motor reactions is carried out by audio fixation.

The stimulus effect is carried out for no more than 90 minutes in the aggregate of all blocks, while the periods of stimulus influence of each set of neutral, control, and verification blocks are chosen to be of the same or different duration, while for different block durations, the number and severity of motor reactions are calculated per unit time for each block of incentives.

Additionally, video reactions of motor reactions are carried out using a video camera.

In addition, reliability is assessed by expert methods based on the results of video recording of motor reactions, when reaction parameters to control stimulus blocks prevail over reaction parameters to test stimulus blocks in two blocks, a conclusion is drawn on the reliability of the reported information, if reaction parameters to test stimulus blocks prevail over reaction parameters to control blocks of stimuli, conclude that the information provided is unreliable, in the absence of the prevalence of signs in two groups determines the prevalence of stimuli of reaction parameters to blocks of stimulus blocks of a control nature over the parameters of reactions to blocks of stimulus blocks of a verification nature and vice versa for the totality of signs, if there is a prevalence of more than 33.3%, they conclude that categorization is made, with values from 16.6% up to 33.3% make a probabilistic conclusion, with values of 16.6% or less it is not possible to draw a conclusion.

The technical result is also achieved by the fact that the device for evaluating the reported information for implementing the method comprises an active unit based on a modular design including a microprocessor and at least one proximity sensor connected to a microprocessor programmed with the ability to process signals from the proximity sensor (s) ( o), an automated assessment unit of the reported information, to one of the inputs of which an active unit is connected.

Contactless sensors can be made in the form of ultrasonic (ultrasonic) sensors and / or motion sensors and / or an audio module.

The device may further comprise a passive unit for video recording of motor reactions, connected to another input of the unit for automated evaluation of the reported information.

The invention is illustrated by drawings, where:

- in FIG. 1 - presents the format of the stimulus system (IS → ATP → NSIS), video recording of motor reactions of the limbs of the victim in a criminal case against a background of diverse stimulus effects on the subject of assessing the reliability (error-free) of the reported data (forensic report No. 224/18, dated 04.06.2018 d);

- in FIG. 2 - the episodes of video recording of micromotor reactions of the defendant in a criminal case against a background of diverse stimulus effects, consisting of two fragments of IP → ATP → NSIS1 and IP → ATP → NSIS2, are presented in accordance with two issues to be resolved (forensic report No. 114/10 , dated May 13, 2010);

- in FIG. 3 - a fragment of the interface is presented, a report scheme based on the results of an automated assessment in accordance with the format of the stimulus system, including two fragments on the first question (A1) and the second question (A2);

- in FIG. 4 - a General view of the inventive device in the process of recording signals from the active module of the scanning device;

- in FIG. 5 is a diagram of the installation, the dotted line indicates the possibility of increasing the number of ultrasound sensors and (or) other proximity sensors;

- in FIG. 6 is a fragment of the contactless registration of complex motor acts, accompanied by a change in the body layout based on the ultrasonic sensor of the active module;

- in FIG. 7 - results of automated calculation in the form of histograms;

- in FIG. 8 - analog-to-digital conversion (conversion of a signal from a sound pressure wave to a digital signal).

The positions in the drawings indicate:

1 - ultrasonic sensors;

2 - webcam;

3 - microphones of the audio module with one-way orientation;

4 - switching unit, including a processing board (microprocessor) and a multi-channel sound card;

5 - processing unit (personal computer - PC);

6-14 - periods of change in motor activity.

The subject is affected by stimulus blocks, which can be a structured interview or a demonstration of static images (photographs, drawings, diagrams, etc.), or video images (real or simulated based on IT technologies), or by influencing the auditory analyzer (questions presenter, samples of audio recordings, systematized sounds, etc.).

Moreover, three groups of stimuli are used in the stimulus system: neutral, control, and verification.

The first group is the indifferent stimuli of IP. At the beginning of the stimulus system, at least one block is always located. When recording reactions to this group of stimuli, reactions that are as close to zero as possible are recorded.

The second group is the socially positioned stimuli of the PCA, which reflect reaction indicators when answering questions of social importance in relation to the functioning of the subject in society.

The third group is directly situational-informational incentives of NSIS, affecting the socially significant (most often in the legal sense) situation and the functioning of the person in relation to whom the NIA is implemented.

When recording reactions to each of the indicated groups of stimuli, reactions are recorded in accordance with the types of response identified on the basis of a population study.

Taking into account the established in practice contact assessment of the reported information on the basis of multifunctional biotechnological measuring and computing complexes [MBIVK (polygraphs)], the stated categories of incentives can be designated respectively as neutral (N), control (K), test (P), or indicated by some otherwise, preserving the original semantic content.

The number of these blocks, as well as the number of stimuli included in the blocks in the format of the stimulus system, can be different, for example, the stimulus system can consist of one fragment with a sequential comparison of motor reactions.

The stimulus effect is carried out by sets of blocks. The basic model of the stimulus system is represented by a single combination of stimulus blocks. For example, IP → ATP → NSIS (see Fig. 1).

It is possible to implement a stimulus system on the basis of a different combination of stimulus blocks: after exposure to an IS block (neutral), the blocks are implemented: ATP (control) and NSIS (test) in a chaotic manner with a different number of blocks. For example, IP → NSIS1 → SPS → NSIS2 (see Fig. 2).

The stimulus system may include other combinations of blocks and different relative positions relative to each other, for example, two IP blocks, two ATP blocks and two NSIS blocks, and other combinations of blocks determining the format of the stimulus system are possible.

Blocks of stimuli can be of the same or of different durations. Comparison of reactions is always carried out only in the sequence IS → ATP → NSIS.

For practical purposes, the assessment of the information communicated is of greatest importance only to those approaches that are based on verified features that allow us to differentiate the patterns of authenticity (error-free) and unreliability (inaccuracy) of the reported information that underlie the proposed method of contactless evaluation. The scientific justification of the proposed method is confirmed by an experimental study to verify the signs of motor response to diverse stimuli in the process of communicating reliable (error-free) / unreliable (erroneous) relevant information and a mathematical analysis of the obtained statistical data is carried out in accordance with the sample.

To calculate the sample size, a special calculator was used (see: https://blog.anketolog.ru/2015/12/vyborka/), the sample structure corresponded to the general population, which was considered as indicators reflecting the number of court sentences that entered into legal force for certain types of crimes for the period from 2005 to 2014 in various districts, with the exception of several federal districts: the Far Eastern, Crimean, North Caucasian. [see: The practice of considering applications for early release of convicts in Russian courts / Analytical report (version for regulatory authorities); under the editorship of O.M. Kiyutsina, IPSO. - SPb., 2016. - 102 сhttp: //i-pso.ru/2016/04/05/299/)]. The total sample index was 814 primary studies on motor activity per unit time against the background of multidirectional stimulation with a confidence interval of 99.7 + 5%, of which 13 were murders; 31 - intentional infliction of grievous bodily harm; 4 - violent sexual crimes (rape); 19 - robbery; 61 - robbery; 253 - theft; 95 - drug offenses; 6 - violations of traffic rules that resulted in the death of a person; 329 - other. The age of the subjects included in the study at different periods of criminal proceedings without differentiation by procedural status ranged from 6 to 69 years, the average was 37 years. At the same time, it was believed that the subjects, regardless of the procedural status and adjudication, can report both error-free and erroneous data in the process of implementing a structured interview. Statistical processing of the results was performed using the statistical software package STATISTICA (data analysis) software system), version 6 of Stat Soft, Inc., USA (2001). The representativeness of the sample (the correspondence of the characteristics of the sample to the characteristics of the population or the general population as a whole) was determined based on the statistical significance of the result at the p-level (p <0.05).

The factor analysis of the data made it possible to distinguish three factors: the first factor is “Multilevel Motility” (PM), the second factor is “Physiologically driven Motility” (POF), and the third factor is “Facial Motility” (FM).

The first factor included the following parameters: multidirectional movements (RD) of the first level (RD 1 - motility in the head area); the second level (RD 2 - motility in the shoulder girdle, including the forearm); third level (RD 3 - brush motility, fingers); fourth level (RD 4 - lower limb motility). The first factor absorbed multilevel multidirectional body movements, including changes in the general layout of the object under study in space, as well as combinatorially determined movements associated (CODS) with the content side of communication and combinatorially caused movements that are not connected (CODSC) with the content side.

The second factor (physiologically driven motility) included: blinking (blinking - blinking (BL); respiratory-forced movements (RFD), including respiratory-reflex movements (RRD); peristaltic movements (PD), visually detectable; autonomic movements (VD - movements language); the number of changes in the pace of speech (CITR); the number of latent periods (CLP); the number of pauses of hesitation (CPC).

The third factor (facies motility) included: upper-local facies movements (VLFD - forehead, eyebrow); mid-local facies movements (SLPD - nose, nose); lower-local facies movements (NLFD - the area of the mouth, lip movements); generalized facies movements (GFD, they are also emotional ones - smile, smirk, grin).

The distribution of indicators by significance within each factor is presented in table 1.

Table 1.

Factor Analysis Results

1. Multilevel motility of 58.7%: Taxiway 1 0.99 Taxiway 2 0.99 Taxiway 3 0.99 Taxiway 4 0.99 Mrd 0.99 CODNS with the content side of communication 0.98 KODS with the content side of communication 0.96

Physiologically driven motility 36.9%: BL 0.99 RFD (RRD) 0.99 PD 0.98 VD 0.98 KITR 0.93 KLP 0.90 KPH 0.87

Facial motility 4.1%: VLFD 0.99 SLFD 0.98 GFD 0.97 NLFD 0.96

According to the results of the factor analysis, multilevel motility (RM), physiologically determined motility (POF) and facies motility (FM) were regarded as verified groups of signs of motor activity.

In addition to determining the main factors, reference values were determined for the four universal types of response identified under standardized stimulus effects, which did not depend on a particular factor or parameter and were manifested by characteristic dynamics, differing only in the level (severity) of reactions.

The first two types of response (A, A1) reflected the dynamics of motor reactions when reporting error-free data, the second two types (B, B1) reflected the dynamics of motor reactions when reporting error data.

The first type (A-type) was manifested by the dynamics of the response associated with a decrease in the level of reactions during control stimulation compared to neutral and an increase in reactions during control stimulation when compared with the control.

The second type (A1-type) was characterized by dynamics due to an increase in the level of response according to the results of control stimulation when compared with the index of neutral stimulation and a decrease in the level of response when exposed to test stimuli when compared with the control reaction.

The third type (B-type) was characterized by a higher level of responses to control stimuli when compared with neutral stimuli and an even higher level of response according to the results of test stimulation when compared with control stimulation.

The fourth type (B1-type) differed from the previous one by the reverse dynamics, the highest level of response was observed against the background of neutral stimulation exceeding the level of the control. At the same time, the level of reactions to test stimuli was less than to control ones.

The dynamics of individual parameters was characterized by a number of specific features affecting both the level of response to diverse stimulation and the reliability of the comparison in calculating the number of motor acts per unit time.

Thus, the RD1 parameter in the framework of the first factor in the A-type response was characterized by the following dynamics of indicators in terms of the ratio of neutral, control and verification stimulation, respectively, 29.01 ± 1.20, 21.89 ± 0.81 and 28.43 ± 0.89 , if there is a statistically significant difference in all cases of comparison (p <0.05).

The same indicator for the A1 type was represented by the following dynamics: 26.55 ± 0.97; 33.19 ± 1.00 and 25.65 ± 0.90, if there is a statistically significant difference in all cases (p <0.05).

The B-type according to the parameter RD1 on neutral stimulation was 26.69 ± 0.78, respectively, on the subsequent 30.19 ± 0.83 and 31.96 ± 0.88. Moreover, a statistically significant difference in comparing the results of neutral and verification stimulation was not obtained (p> 0.05).

The same indicator for the B1 type of response reflected the following dynamics: 37.15 ± 0.99, then 31.38 ± 0.84 and 23.65 ± 0.84. Moreover, in all cases of comparison, a statistically significant difference was obtained (p <0.05).

When analyzing the response changes according to the RD2 parameter, the following picture was obtained: by the A-type 2.46 ± 0.17, then 1.43 ± 0.12 and 2.67 ± 0.20; by the A1 type 3.90 ± 0.31, then 5.94 ± 0.40 and 4.27 ± 0.34; B-type 2.04 ± 0.13, respectively 3.19 ± 0.19 and 4.61 ± 0.26; by B1-type 4.01 ± 0.29 then 2.78 ± 0.24 and 1.70 ± 0.14. In all cases, for the considered parameter, with the exception of comparing the results of the neutral series with the B-type test, a statistically significant difference was obtained (p <0.05).

When analyzing the response changes according to the RD3 parameter, the result is obtained: by the A-type 12.57 ± 0.49, then 10.97 ± 0.39 and 11.38 ± 0.41; by A1 type, 22.98 ± 1.28, then 23.27 ± 1.35 and 20.32 ± 1.04; B-type 10.97 ± 0.63, respectively 11.04 ± 0.73 and 13.98 ± 0.68; according to the B1 type, 10.48 ± 0.87, then 9.90 ± 0.71 and 7.40 ± 0.47. Moreover, in all cases, according to the considered parameter, with the exception of comparing the results of the neutral series with the test according to the B-type, a statistically significant difference was obtained (p <0.05).

The RD4 parameter was characterized by the following numerical values: by the A-type 12.24 ± 0.69, then 10.87 ± 0.45 and 11.09 ± 0.69; by A1-type 10.68 ± 0.73, then 11.71 ± 0.98 and 9.95 ± 0.71; by B-type 9.45 ± 0.52, respectively 10.81 ± 0.68 and 12.82 ± 0.89; according to the B1 type, 12.05 ± 0.53, then 10.26 ± 0.55 and 8.40 ± 0.35. In all cases, according to the considered parameter, with the exception of comparing the results of the neutral series with the test according to the B1 type, a statistically significant difference was obtained (p <0.05).

The mrd parameter was characterized by the following numerical values (at p <0.05): by the A-type 6.99 ± 0.46, then 5.00 ± 0.29 and 7.30 ± 0.39; by A1-type 5.08 ± 0.14, then 6.10 ± 0.22 and 5.45 ± 0.21; in the B-type 2.25 ± 0.20, respectively 3.58 ± 0.19 and 4.90 ± 0.28; according to the B1 type, 5.96 ± 0.26, then 4.60 ± 0.20 and 3.12 ± 0.16.

KODS-parameter: by the A-type 6.80 ± 0.64, then 5.03 ± 0.41 and 6.60 ± 0.49; by A1 type 2.13 ± 0.23, then 3.60 ± 0.29 and 2.51 ± 0.27; by B-type 2.48 ± 0.27, respectively 3.55 ± 0.32 and 4.88 ± 0.35; by B1-type 3.00 ± 0.35, then 2.56 ± 0.29 and 1.90 ± 0.28. Moreover, in all cases, according to the considered parameter, a statistically significant difference was obtained (p <0.05).

The CODN parameter was characterized by the dynamics of indicators with a statistically significant difference for all comparison options (p <0.05) and amounted to 1.39 ± 0.10 for the A-type, further 0.98 ± 0.06 and 1.35 ± 0, 08, by A1-type 0.31 ± 0.05, then 0.53 ± 0.04 and 0.31 ± 0.03, by B-type 0.55 ± 0.02, respectively 0.79 ± 0, 07 and 0.98 ± 0.07, according to the B1 type 0.50 ± 0.04, then 0.33 ± 0.03 and 0.25 ± 0.03.

The BL parameter within the second factor was characterized by the following features of the dynamics of indicators according to the results of multidirectional stimulation: by the A-type 30.66 ± 1.06, then 28.04 ± 1.04 and 30.14 ± 1.13; by A1 type, respectively, 23.22 ± 0.87, then 24.97 ± 0.93 and 22.92 ± 0.80; by B-type - 23.72 ± 1.14, 25.64 ± 1.22 and 26.9 0 ± 1.32; by B1-type - 27.14 ± 0.91, then 26.18 ± 1.02 and 25.09 ± 0.89. Moreover, in the case of comparing the results of neutral stimulation with verification, a statistically significant difference was not confirmed by B and B1 types of response (p> 0.05).

The dynamics of the indicators according to the RFR / RRD parameter is represented by the following picture: by the A-type - 2.40 ± 0.56, then 1.94 ± 0.49 and 2.50 ± 0.55; A1-type - 0.63 ± 0.04, then 0.95 ± 0.06 and 0.65 ± 0.05; B-type - 1.05 ± 0.27, 1.88 ± 0.34, 2.20 ± 0.39; for the B1 type — 0.68 ± 0.04, 0.43 ± 0.03, 0.30 ± 0.02, a statistically significant difference in the indicators was confirmed for all types of response (p <0.05).

The PD parameter is represented by the following characteristics of the dynamics of indicators according to the results of multidirectional stimulation: by the A-type 0.99 ± 0.03, then 0.71 ± 0.03 and 0.84 ± 0.02; by A1 type, respectively 0.46 ± 0.02, then 0.62 ± 0.02 and 0.53 ± 0.03; B-type 0.40 ± 0.02, 0.60 ± 0.02, 0.72 ± 0.02; by B1-type 1.04 ± 0.04, 0.76 ± 0.02; 0.54 ± 0.01, in all of the above types of response and comparison options, a statistically significant difference was obtained (p <0.05). According to the fourth type of response (B1), no statistically significant difference was found when comparing the results of neutral and verification stimulation (p> 0.05).

The VD parameter was represented by the following dynamics of indicators with a statistically significant difference for all comparison options (p <0.05): for the A-type 1.40 ± 0.10, then 0.85 ± 0.05 and 1.28 ± 0 , 09; by A1 type, 0.47 ± 0.76, then 0.66 ± 0.82 and 0.42 ± 0.66; by B-type 0.45 ± 0.05, respectively 0.68 ± 0.07 and 0.78 ± 0.09; according to the B1 type, 1.09 ± 0.18, then 0.86 ± 0.13 and 0.61 ± 0.10.

The VLFD parameter within the third factor was characterized by the following features of the dynamics of indicators according to the results of multidirectional stimulation: by the A-type 6.96 ± 0.47, then 5.23 ± 0.43 and 6.67 ± 0.50; by the A1 type, 6.43 ± 0.56, then 7.68 ± 0.60 and 6.00 ± 0.58; according to the B-type 2.40 ± 0.11, respectively 3.17 ± 0.13 and 4.38 ± 0.18; by B1-type 6.18 ± 0.37, then 5.53 ± 0.30 and 4.03 ± 0.24. According to this parameter, in all cases, with the exception of the results of neutral and verification stimulation according to the fourth type of response, a statistically significant difference was obtained (p <0.05).

The SLPD parameter was characterized by the following characteristics of the dynamics of indicators according to the results of multidirectional stimulation: by the A-type 7.19 ± 0.46, then 6.06 ± 0.40 and 7.12 ± 0.49; by the A1 type, 6.40 ± 0.40, then 7.39 ± 0.47 and 5.93 ± 0.41; by B-type 4.11 ± 0.48, respectively 5.41 ± 0.54 and 6.32 ± 0.57; according to the B1 type, 6.71 ± 0.41, then 5.69 ± 0.37 and 4.25 ± 0.31. According to this parameter, in all cases and in all variants of comparison, a statistically significant difference was obtained (p <0.05).

NLFD-parameter was represented by the following dynamics of indicators with a statistically significant difference for all comparison options (p <0.05): A-type 5.54 ± 0.22, then 4.31 ± 0.22 and 5.40 ± 0, 22; by the A1 type, 5.75 ± 0.45, then 6.90 ± 0.46 and 5.98 ± 0.41; by B-type 3.56 ± 0.15, respectively 4.48 ± 0.17 and 5.60 ± 0.20; according to the B1 type, 7.79 ± 0.30, then 6.79 ± 0.28 and 5.25 ± 0.23.

The GPD parameter was characterized by the following dynamics of indicators with a statistically significant difference for all comparison options (p <0.05): A-type 2.38 ± 0.56, then 1.74 ± 0.49 and 2.53 ± 0.55 ; A1-type 0.63 ± 0.04, further 1.12 ± 0.07 and 0.65 ± 0.05; B-type 2.04 ± 0.13, respectively 3.19 ± 0.19 and 4.61 ± 0.26; according to the B1 type, 0.55 ± 0.04, then 0.43 ± 0.03 and 0.31 ± 0.02.

The essence of the proposed method is associated with the implementation of situational information analysis (SIA) as a universal methodological basis for a non-contact assessment of the reliability (accuracy) of the reported data

The algorithm for implementing SIA data includes the development of a stimulus system [structured interview, structured conversation (test)] or some other form that underlies subsequent communication with the subject, acceptable in a particular case. For this, the built-in text editor of the program for automated situational-information analysis DetectInfo-sensor-channel (DSC), certificate of state registration No. 2018663519 of 10.30.2018), authors Ivanov L.N., Potapov R .V., Azizov M.V.

Each problem to be resolved on the basis of SIA in the stimulus system is given a specific period of stimulus exposure, designated as a "fragment", a "comparison (comparison) zone", or indicated in some other way. Each fragment consists of a set of diverse stimuli, designated as a “stimulus block” (“stimulus group”) or designated in some other way. A fragment or set of fragments is related to the concept of “stimulus system”, or “research program”, or is indicated in some other way. A fragment always correlates with the assessment of the reported data on any one issue that matters to the initiator of the study, in FIG. Figure 3 shows the report scheme for two fragments (two issues to be resolved).

Before the study, proximity sensors, a Web camera for recording motor manifestations against the background of stimulus exposure and microphones are placed in relation to the location of the object. Communicators are located face to face, recording management is carried out by a specialist on the basis of a personal computer (PC) and appropriate software (software).

The subject is exposed to stimulus blocks IS (neutral), ATP (control) and NSIS (test) for no more than 90 minutes in the aggregate of all blocks included in the stimulus system. The periods of stimulus exposure of each set of blocks of a neutral, control and verification nature are chosen of the same or different duration.

The parameters of motor reactions to the stimulus effect are recorded by scanning various anatomical zones of the test person online using contactless sensors made in the form of ultrasonic sensors and / or motion sensors and / or an audio module (active units of the device), including microphones with one-way focus for audio recording and a sound card reflecting the motor function of the vocal muscle, as well as a video camera (passive device unit).

Evaluation of the reported information is carried out according to the parameters grouped on the basis of factor analysis by comparing the response to stimulus blocks of a test and control nature with identifying the trend of changes for each parameter and / or group of parameters.

Three groups of verified reaction parameters are used as reaction groups: multilevel motility, facies motility, physiologically determined motility.

Identification of trends is carried out according to four types of response, formed on the basis of dominant trends: the maximum and minimum parameters of reactions to stimulus blocks of a control nature, the increase and decrease of parameters of reactions to stimulus blocks from IP (neutral) to NSIS (test), while the stimulus the impact is carried out by sets of blocks, after exposure to a block of ICs (neutral) in nature, they are impacted by blocks of ATP (control) and NSIS (test) character in a chaotic manner with different numbers Twomey blocks included in the stimulus system.

Preparation of recorded reactions for analysis.

On the basis of the implemented stimulus system, samples are obtained for comparative studies and the initial material is prepared for reference. Regardless of the channel for receiving information about the studied object, both when using technical devices and direct expert evaluation, when calculating the reactions in the first block of the stimulus system, the signal recording section from the beginning of the stimulus in the amount of 25% of the duration is stopped.

For various durations of stimulus blocks included in the fragment, an automated (software) calculation of the number of reactions per unit time for each anatomical zone included in the analysis is performed.

With direct visual expert assessment carried out by a competent person, for each analyzed regional anatomical zone (with automated processing of results) for each block of stimuli in accordance with the formula: Р / Тs * Сonst, where:

P is the number of signs for each parameter;

Тs - duration of the corresponding block of questions in seconds;

Const = 60 sec.

The reliability assessment is carried out automatically taking into account the types of response: when the response parameters to the control stimulus blocks prevail over the reaction parameters to the test stimulus blocks, the information is reported to be reliable, when the reaction parameters to the test stimulus blocks prevail over the reaction parameters to the stimulus blocks control nature conclude that the information provided is unreliable.

In the production of process studies, the automated assessment is supplemented by an assessment of the reliability carried out by an expert according to the results of video recording of motor reactions, when the reaction parameters for control stimulus blocks prevail over the reaction parameters for check stimulus blocks in two blocks, the conclusion is drawn on the reliability of the information reported, with the reaction parameters prevailing on blocks of stimulus blocks of a verification nature over the parameters of reactions to blocks of stimulus blocks of a control nature RA, they conclude that the reported information is unreliable, in the absence of predominance of signs in two groups, the prevalence of stimuli of reaction parameters to control stimulus blocks is determined over reaction parameters to test stimulus blocks and vice versa for the whole set of signs, if there is a prevalence of more than 33.3%, a categorical conclusion, with values from 16.6% to 33.3% they make a probabilistic conclusion, with values of 16.6% or less it is not possible to draw a conclusion.

The substantiation of conclusions, taking into account the degree of categoricality based on the results of automated processing, differs from a direct expert assessment of the source data. This is due to the fact that the comparison of the dominance of the control or test subjects takes into account not only the frequency of response to stimuli, but also the severity of motor manifestations in individual local anatomical zones.

The presence of a statistically significant difference (SDR) in the automated assessment when comparing data sets allows us to increase the conclusion drawn from the results of a direct expert assessment in terms of categoricality by one level.

To implement the proposed method, a device is used, a general view of which is presented in FIG. 4. The device is made in the form of a modular design, contains active and passive blocks, the general diagram of the device is shown in FIG. 5.

The active unit contains proximity sensors 1, made in the form of ultrasonic sensors and / or motion sensors and / or microphones 3 of the audio module.

Ultrasonic sensors 1 of the device are connected to a microprocessor located in the switching unit 4 and then connected to the processing unit 5 (PC) for registering reflected rays from the object of study by measuring the distance from the sensor (sensors as ray sources) to various anatomical and physiological regions of the subject (Fig. . 5).

In addition to ultrasonic vibrations, it is possible to use other types of rays (directional light, laser, x-ray). Depending on the types of tasks, sensors based on the piezoelectric effect, magnetostrictive and others can be used.

The possibility of increasing the number of ultrasonic sensors or connecting other non-contact sensors, the number of which, as well as options for combining them with other non-contact sensors operating on various physical principles, is determined by the conditions of a specific study, the conditions in which a situational information analysis (SIA) of data is carried out, and the specifics tasks assigned to the specialist (indicated by a dotted line in FIG. 5).

Using an active module using non-contact sensors, generalized ones are fixed that are associated with the inclusion of several motor groups in the motor reaction, leading to a change in the position of the subject's body in space and (or) large-motor reactions in the anatomical zones of local fixation.

A graphical display of physical activity recorded online is shown in FIG. 6, where: period 6 reflects the initial position of the subject; periods 7, 9 indicate an increase in the distance between the sensor and the object of study; periods 8, 10, 12, 14 indicate retention of distance; periods 11, 13 - to approach the object of study to the sensor. Thus, FIG. 6 reflects the sequential process associated with a change in the general layout of the body in space, which is due to multilevel multidirectional body movements (MRI). The result of automated calculation in the form of histograms is shown in FIG. 7.

Automated calculation of motor manifestations according to the specified module is oriented both to counting the number of movements associated with changing the distance from the sensor to the local anatomical zone of the subject, and the severity of motor manifestations against the background of diverse stimulus effects in accordance with the claimed method.

The second approach involves the use of this module based on artificial intelligence.

The module, based on the passive method (video module), uses the existing ambient light and represents the shooting of an object using various video cameras (Web-cameras). The camera simultaneously serves for formal video recording of the course of the stimulus effect and for obtaining samples of a subsequent comparative study.

On the basis of the video module, the application of the stimulus system developed for the task of studying and obtaining samples for comparative research is fixed. The same video is used for automated assessment of motor manifestations based on IT-technologies.

The basis of the video module is a Web camera with a resolution of at least 2 megapixels (without interpolation of 1920x1080 pixels) and a microphone, an input of at least USB 2.0, high-definition video recording of 1080p or other parameters, or a video camera with performance characteristics allowing direct (expert) analysis of motor manifestations from the display and (or) automated analysis. Automated counting using the video module of a scanning device differs from a direct expert assessment of the source data. This is due to the fact that the comparison of the dominance of motor reactions to stimuli takes place taking into account not only their frequency of occurrence, but also the severity of manifestations in individual local anatomical zones.

To assess fine motor skills in the claimed method, a computer-based program for automated situational-information analysis SIADI was used (Ivanov L.N., Sidorenko D.A. Program for automated situational-informational analysis of personified information DetectInfo [SIA Detectinfo (SIADI)]. Certificate of state registration of the program for computers No. 2018614437 of 08/06/2018).

After selecting the anatomical and physiological zone for assessing motor activity, a frame-by-frame cycle of processing the initial video was automatically performed with the differentiation of foreground and background image elements (image element, pel - from English pel - short for pi element element (pix element) or pi capture cell or eliz - Russian a variant of the term pixel (pixel) - from the English pixel - the smallest logical element of a two-dimensional digital image in a raster graphic, i.e. an indivisible object). Subtracting the background included calculating the reference image, subtracting each new frame from this image and the threshold value of the result. Thus, binary image segmentation was formed in the projection of the studied anatomical zone. After reading the number of image elements on each frame, they were saved according to the blocks of grouping diverse stimuli in accordance with the format of the stimulation period (the image in the frame as a square of 1 cm ² , obtained from a distance of 2 meters, corresponded to 25 points, the change in the number of points in dynamics reflected motor activity, its increase or decrease). After completing the calculation of the video, the output received an array of data in accordance with the format of the stimulus system, which saved the number of image elements on each frame in a given calculation area. The data obtained were divided into two arrays, including local maxima and local minima. After that, the number of amplitudes and their severity were calculated.

Under stimulus exposure, the dominance of motor reactions (ATP> NSIS or NSIS> ATP) can occur both due to a change in the number of reactions and due to a change in severity. This means that the calculations of the reliability of the conclusions of the study should take into account both the frequency of their occurrence and their severity. The number of automated measurements in relation to a single projection anatomical zone can be in the tens or hundreds, which fundamentally changes the attitude to the calculation of the reliability of the result in specific studies, which are based in SIA using large statistical groups. Other options for the operation of the video module include the use of the principle of pattern recognition in relation to various anatomical and physiological zones of motor activity and other approaches that allow you to evaluate the dynamics of the motor function of the subject in the process of stimulus exposure.

The second approach to the functioning of the video module is associated with pattern recognition by supporting anatomical and physiological zones and counting the number of motor manifestations in them.

The third is associated with the use of a module that operates on the basis of artificial intelligence.

The operation of the audio module of the device is associated with a discrete time function and a finite set of possible values of binary digital signals (bit stream). The general scheme of comparative analysis for this module is fully consistent with the claimed method by comparing samples of the subject’s voice obtained at different periods of stimulus exposure.

The passband for the transmission channels of sound messages of the subject is limited to a frequency of 4-5 kHz and (or) may be different. The sampling frequency is 8-10 kHz and / or may be different. The signal-to-noise ratio is maintained at the highest achievable level.

First, an analog-to-digital conversion is performed, i.e. converting a signal from a sound pressure wave to a digital signal (FIG. 8).

To implement SIA data based on the use of an audio module, it is necessary to record speech in the RAM of a PC or computer storage medium. The device uses narrow-band [speech (with a frequency range from 300 to 3000 Hz)] microphones with one-sided orientation. The distance from the microphone to the subject’s face is determined by the nature of the study (process / non-process), the properties of the microphone, and the conditions under which recording is performed.

For each communicator, a separate microphone can be used, for which the device has a multi-channel sound card (ADC) with the ability to configure digitization parameters. The voice recording of a communication specialist is used as a model for the subsequent use of a multi-band expander. To reduce the complexity of subtraction when superimposing the voices of communicators, the device provides the ability to disable recording when voicing questions by a specialist.

To obtain sound characteristics by various parameters for comparison in accordance with the requirements of the method and to obtain clear spectral characteristics of sound received from the subject’s microphone, it is cleaned from noise. The input discrete audio signal is processed by filters according to the formula, where Xi is the set of discrete values of the audio signal.

On the basis of the known characteristics of the voice, those that are not related to them are canceled, for this the Kaiser window function is used

After filtering the spectrum, the Hanning window is superimposed.

After filtering the noise in the signal, the start and end of the recording are determined. The beginning of the audio recording will be characterized by a burst of signal, which corresponds, in our case, to the subjects uttered the first word at the beginning of communication, if you look with X ₀ . Accordingly, if you look down from Xn, then the burst will correspond to the end of the audio recording.

When the specialist’s voice is superimposed on the audio recording from the subject’s microphone, based on software designed to determine the similarity and difference between the signals, the specialist’s voice is subtracted. The output generates a file with a difference that reflects only the audio recording of the subject's voice in the communication process. If artifacts are saved by overlaying votes, these sections of the audio recording are excluded from the analysis.

To complete the preparation of the file for comparative analysis according to the claimed method, the initial part is excluded from the audio recording corresponding to the first stimulus block, an entry is left in the array of discrete signal values suitable for comparative analysis by stimulus blocks.

Analysis of the audio recording, reflecting motor manifestations associated with the functioning of the subject’s vocal muscle, occurs according to various parameters.

Amplitude (dB), reflects the change in voice volume. In our case, the amplitude with respect to the digitized signal is the sample value. We measure the discrete values of the level (amplitudes) in the comparable blocks of the stimulus system. Analyzing the set of discrete values of the audio recording level in each of the blocks of the stimulus system, we get a picture of the dynamics of indicators for each block of IP, ATP, NSIS. The resulting data array is subjected to analysis with the calculation of reliability indicators in accordance with the claimed method.

Peak amplitude - that is, the absolute value of the maximum of the obtained discrete values of the sound level. To calculate the mean square value (RMS) of the amplitude, we square the obtained values, calculate the average value of the resulting set, and then extract the root from it, if there are N values and x (i) is the amplitude of the i-th discrete value, then

.RMS amplitude =

.

When measuring amplitudes corresponding to different phases, instead of the amplitude and phase, two amplitudes corresponding to different phases are measured. One of these phases is represented by the cosine, the other by the sine.

Power. To determine the total power (PM) over a period of time, the squares of samples are added up, the average value of the squares of the discrete values is proportional to the average power.

Power spectral density (PSD) is a function that describes the distribution of signal power depending on the frequency, that is, the power per unit frequency interval. The PSD signal stores information only about the amplitudes of the spectral components (phase information is lost), therefore, all signals with the same amplitude spectrum and different phase spectra have the same power spectral density. Evaluation of the PSD can be performed by the Fourier transform method, which assumes obtaining a spectrum in the frequency domain by means of the fast Fourier transform (FFT) or by another method.

Relative power in our case means the ratio of power in comparison by stimulus blocks according to the claimed method. Based on this parameter, two sounds are compared corresponding to the compared blocks of stimuli; for this, the ratio of their power is taken.

The following approach is used to compare IP with ATP, and then ATP with NSIS. Suppose that X [1..N] and Y [1..N] are arrays of numbers of the same size N containing spectral powers corresponding to periods of the stimulus being compared. Then the measure of similarity between them is calculated by the following formula:

where Mx and My are the mathematical expectations for the arrays X [] and Y [], respectively, calculated by the following formula:

In addition, the average power spectral density (power gain / attenuation), the ratio of the average amplitude and average power, and sound loss are estimated. To do this, using a digital sound processing algorithm, it is established whether there is a difference (dB) at the output, i.e. it is established how much the sound corresponding to the IC block is weaker / stronger than the SPS block, the period of the SPS block is compared accordingly with the NSIS block.

Frequency - due to fluctuations in the vocal folds (hertz / sec.). Sound is decomposed into sine waves, this allows you to build the frequency spectrum of sound. The frequency spectrum of a sound wave is a graph of the amplitude versus frequency. To measure the amplitude of one frequency, the existing signal is multiplied by a sinusoid of the same frequency, the obtained samples are added up. When writing in symbolic form, the samples have the values s ₀ , s ₁ , ..., s _t , .... The variable t is the reference number that replaces the time value. The amplitude of the frequency f is measured in a first approximation, when calculating the following sum:

A _f =

The values of t and f do not match the accuracy of time and frequency. Summation gives a number proportional to the amplitude. Repeating these calculations for different values of f allows you to measure the amplitude of all frequencies in the signal.

The spectral density of the aperiodic function reflects the characteristic features that appear as a result of frequency analysis. Here, using a pair of Fourier transforms, the relationship between the aperiodic function of time and its complex spectrum is determined.

In addition, the frequency and duration of pauses and other audio parameters are measured.

The second approach to the functioning of this module is associated with the use of artificial intelligence.

The third - is associated with spatial orientation, based on at least three microphones located in space with respect to the sound source, which allows you to track the 3D position of the object due to changes in its position (relative positioning patterns) with respect to microphones at various stages of stimulus exposure.

The fourth option is associated with the use of one speech microphone with a manual and (or) automated homing system for the object of study.

In addition, a breaker is provided that allows you to mute the subject’s microphone while voicing verbal stimuli by the leading communicator

Regardless of the selected system for evaluating the results, based on a direct analysis of the materials carried out by an expert on the basis of their recognition by verified signs or automatically based on the claimed device, after receiving reaction samples, they are differentiated according to the type of reaction for each symptom (in the case of a direct assessment) and (or) according to regional anatomical and physiological zones allocated for automated analysis using the inventive device or some about another device that operates according to the canons of a comparative analysis of reactions obtained on the basis of stimulus effects, with the allocation of A / A1-types of response, respectively, reflecting the dynamics of motor reactions when reporting reliable (error-free) data, or B / B1 types indicating inaccuracy ( error) of the data, which allows us to determine the dominant trend and make a final conclusion.

When comparing the scheme: IS → ATP → NSIS, each type of response was characterized by the principle of "more or less" (corresponding to an increase, decrease in the level of reactions).

A-type (more-less-more). It was manifested by the dynamics of the response associated with a decrease in the level of reactions according to ATP compared with IP and an increase in reactions to NSIS when compared with ATP.

A1-type (less-more-less). It was characterized by dynamics due to an increase in the level of response to ATP when compared with IP and a decrease in the level of response to NSIS when compared with ATP.

B-type (less-more-more). It was characterized by a higher level of reactions to ATP when compared with IP and an even higher level of response according to the results of NSIS when compared with ATP stimulation.

B1-type (more-less-less). The highest level of response was observed against the background of IP, exceeding the level of ATP, while the level of reactions to NSIS was less than that of ATP with respect to the fragment that corresponds to the issue to be resolved on the basis of SIA.

Depending on the tasks (procedural or non-procedural), the method allows using one of the assessment systems, directly expert, carried out by a knowledgeable person or automated, and their combination to formulate conclusions. An automated assessment can use one of the device modules, some combination of them, or the entire set of modules of the claimed device.

When implementing NIA based on a combination of a direct visual assessment with an automated assessment using the inventive device, the conclusion based on the results of the study takes into account each of them, while the conclusion based on the results of a direct (visual) assessment is based on the coincidence of two intermediate conclusions according to the groups of signs, or the conclusion is carried out on the basis of calculating the difference in the dominance of the control and verification topics on the basis of all informative parameters and is compared with dartizirovannoy scale.

The combination of a direct expert assessment carried out by a competent person with automated registration and an assessment based on the claimed device is mandatory if the study is of a procedural nature and the report is used as evidence in legal proceedings.

The presence of a statistically significant difference (SDR) in the compared data sets reflecting motor activity under the influence of diverse stimulus effects is an essential point in substantiating the conclusions. This is due to the fact that the presence of SDR in an automated assessment affects the categoricality of conclusions drawn from the results of an assessment made directly by a forensic expert, by increasing the categoricality of the latter by one step.

When initiating a situational information analysis (SIA), which reflects the confirmation of the result with the calculation of the probability and reliability of the conclusions, computer programs designed for statistical data processing are additionally used, or special programs with integrated statistics packages are used.

Examples of using the claimed method in the framework of expert procedures in criminal cases.

Example 1

Consider, as an example, a fragment of an automated assessment based on a passive module of the results of facial expression from expert opinion No. 223/18 related to the investigation of the suspect P.M.S., suspected of committing a crime in accordance with Part 2 of Art. 162 of the Criminal Code.

Based on automated data processing, two arrays of numbers were allocated. The first one reflected the dominance of the control subjects, where the average value, taking into account the number and severity of reactions, was 12.1 ± 1.83. The second one was related to the verification topic and amounted to 8.3 ± 2.22. The performed reliability calculation revealed a difference significant at the level of p <0.05, with a confidence interval from - 95,000% to + 95,000%.

The presence of a statistically significant difference (SDR) in the compared data sets reflecting motor activity under the influence of control or test stimuli is an essential point in substantiating the conclusions. This is due to the fact that the presence of SDR in an automated assessment affects the categoricality of conclusions drawn from the results of an assessment made directly by a forensic expert, by increasing the categoricality of the latter by one step.

In our example, the difference in the indicators reflecting the dominance of the control and verification topics in the group (facial motility) amounted to less than 16.6% of the number of informative signs. Therefore, it was not possible to conclude from the results of a direct assessment of motor activity for the indicated group of signs. However, taking into account automated data processing, the group conclusion was regarded as probabilistic.

Example 2

From the decision on the appointment of a forensic examination to assess the reliability of the information provided: whether the victim B., born on September 8, 1964, reactions testifying to the fact that he passed on May 22, 2016, are identified during the assessment of the reliability (infallibility) of the reported information , Born on December 27, 1963, funds in the amount of 80,000 rubles for assistance in not bringing the company to administrative responsibility in the form of a fine for the violations committed?

From the expert’s opinion No. 289/17: the assessment of motor activity by video was carried out according to verified criteria by an expert method based on a visual analysis of the dynamics of motor activity and automatically. In the process of researching the recording of a structured interview, the dynamics of motor activity was monitored and recorded in three groups of signs: “multilevel motor skills” (first group); “Physiologically determined motility” (second group); “Facies motility” (third group).

Automated analysis of motor activity based on the DetectInfo LLC device (see Fig. 4) was performed according to two local anatomical zones (head region, area of primary location of the hands). The correlation of the number and severity of reactions was carried out programmatically by parameters reflecting motor activity in the ranges: 0-5 points; 6-10; 11-15, etc. up to 296-300. Statistical processing of the results was performed using the statistical software package STATISTICA (data analysis) software system), version 6 of Stat Soft, Inc., USA (2001).

In the course of the study, it was established: according to the first group of signs, more pronounced, systemic reactions to NSIS were obtained compared to ATP by the ratio of response indicators according to the RD4 parameter - 23.34-7.52-4.50; MRDT - 5.76-4.52-4.25; CODNS - 1.64-0.38-0.00. At the same time, more pronounced, systemic reactions to ATP stimuli were obtained compared to NSIS in parameters: RD1 - 24.33-21.82-26.88; RD2 - 0.49-0.00-0.38; RD3 - 10.19-5.83-13.25; KODS –0.33-0.56-3.38. Thus, the ratio of indicators indicating the fallibility / error-freeness of the reported data for the first group of signs was 3/4.

When assessing motor activity according to the second group of signs, more pronounced reactions to NSIS stimuli were obtained compared to ATP in terms of the ratio of response indicators to IS, ATP and NSIS questions with respect to parameters: PD - 0.16-0.19-0.50. At the same time, more pronounced, systemic reactions to ATP stimuli were obtained according to the parameters: BL - 44.88-39.50-42.63; RFD (RRD) - 1.15-0.38-1.13; VD - 2.79-1.69 - 2.87. Thus, the ratio of indicators indicating the error / error of the reported data in the second group of signs was 1/3.

In the third group, more pronounced, systemic reactions to ATP stimuli were obtained compared to NSIS in terms of parameters: VLFD - 2.79-1.88-2.13; SLPD - 0.82-1.50-1.38; NLFD - 5.10-3.01-5.38. Thus, the ratio of indicators indicating the error / error of the reported data for the third group of signs was 0/3.

The overall ratio of indicators indicating the reliability (error-free) inaccuracy (error) of the reported data from a structured interview based on an expert approach was 4/10 due to the predominant number of reactions to ATP. This indicates that B. has information on the circumstances of the case and reports reliable (error-free) data (see tables 2, 3, 4, 5, 6).

table 2

The results of an expert assessment of the dynamics of motor activity

Groups and subgroups of motor activity Measurement indicators for blocks of questions: IP THX NSIS 1. Different level motility 1.1. Taxiway 1 24.33 21.82 26.88 1.2. Taxiway 2 0.49 0.00 0.38 1.3. Taxiway 3 10.19 5.83 13.25 1.4. Taxiway 4 23.34 7.52 4,50 1.5. MRDT 5.76 4,52 4.25 1.6. CODES 0.33 0.56 3.38 1.7. CODNS 1,64 0.38 0.00 2. Physiologically driven motility 2.1. BL 44.88 39.50 42.63 2.2. RFD (RRD) 1.15 0.38 1.13 2.3. PD 0.16 0.19 0.50 2.4. VD 2.79 1,69 2.87 3. Facial motility 3.1. VLFD 2.79 1.88 2.13 3.2. SLFD 0.82 1,50 1.38 3.3. NLFD 5.10 3.01 5.38 3.4. GFD 0.16 0.00 0.00

Table 3

Results of automated assessment of motor activity

in frequency and severity of signs (projection of the head)

№№ Range ATP dominance Range NSIS dominance IP THX NSIS ∑ NKP IP THX NSIS ∑ NKP 0-5 47.62 43,23 50.29 16 20 7.71 8.87 9.21 25.79 6-10 13.02 18.73 16.17 47.92 51 55 2.74 4.93 6.30 13.97 11-15 9.77 11.34 9.21 30.32 116 120 0.85 1,64 3.07 5.56 21-25 5.65 10.02 8.57 24.24 176 180 0.68 0.82 1.13 2.63 26-30 5.31 7.72 7.60 20.63 216 220 0.85 0.82 0.64 2,31 31-35 5.13 8.21 4.36 17.88 221 225 0.51 0.65 1.13 2.29 36-40 3,59 7.72 5.01 16.32 241 245 0.51 1.15 1.29 2.95 41-45 3.76 5.91 5.01 14.68 46-50 2.91 7.56 4.20 14.67 56-60 3.08 5.42 4.04 12.54 61-65 3.08 6.24 3.07 12.39 66-70 2.23 4.76 3.07 10.06 71-75 2.06 5.09 4.36 11.51 76-80 2,56 4.27 2.74 9.57 81-85 1.88 4.76 2.74 9.38 86-90 2.23 5.09 1.94 9.26 91-95 2.23 3.94 2.10 8.27 96-100 2.40 2.79 1.77 6.96 101-105 1.19 3.12 0.80 5.11 106-110 2.22 2.79 1.94 6.95 111-115 2.05 2.63 1.77 6.45 121-125 1.19 2.63 1,61 5.43 126-130 1.37 2.63 2.10 6.1 131-135 1,02 2,30 1.94 5.26 136-140 1,02 2,30 1.45 4.77 141-145 0.51 1,64 1.13 3.28 146-150 0.85 3.45 0.80 5.1 151-155 1.19 1.97 1.77 4.93 156-160 0.68 1.97 1.13 3.78 161-165 0.51 2,30 0.64 3.45 166-170 1,02 1,64 0.48 3.14 171-175 0.68 2,30 0.97 3.95 181-185 0.85 1,64 0.97 3.46 186-190 0.85 0.82 2.10 3.77 191-195 0.68 2,30 0.80 3.78 196-200 0.85 1.97 1.13 3.95 201-205 1,02 2.63 2.10 5.75 206-210 1,54 1,64 1.29 4.47 211-215 0.85 0.82 1.29 2.96 226-230 0.51 0.32 0.48 1.31 231-235 0.00 1.15 0.80 1.95 236-240 0.34 0.82 0.80 1.96 246-250 0.34 0.32 0.97 1,63 251-255 0.51 0.82 0.48 1.81 256-260 1,02 0.49 0.64 2.15 261-265 0.68 1.31 0.64 2.63 266-270 0.51 0.82 0.48 1.81 271-275 0.34 0.32 0.80 1.46 276-280 0.68 0.49 0.64 1.81 281-285 0.34 1.31 0.32 1.97 286-290 0.51 0.82 0.48 1.81 291-295 0.68 0.49 0.80 1.49 296-300 1,02 0.32 0.48 1.82

Table 4

T-criterion for dependent samples subject to dominance ATP> NSIS

and NSIS> ATP

Variable n Average
value by
∑ IS-SPS-NSIS Std. Off different. t p Will trust.
- 95,000% Will trust.
+ 95,000% SPS> NSIS 53 36.90 NSIS> ATP 7 18.67 10.52 4,58 0.003754 8.49 27.95

Note: the difference is noted, significant at the level of p <0.05.

Thus, the results of automated analysis in relation to the designated local zone coincide with the conclusions obtained on the basis of expert evaluation.

Table 5

Results of automated assessment of motor activity

in frequency and severity of signs (projection of the location of the brushes)

No. Range ATP dominance Range NSIS dominance IP THX NSIS ∑ NKP IP THX NSIS ∑ NKP 11-15 34.94 36.65 32.34 103.93 0-5 84.28 62.95 61.77 209.00 21-25 20.72 37.31 29.11 58.03 6-10 49.85 36.82 35.57 122.40 26-30 17.30 25.47 20.86 63.63 16-20 27.40 30.08 30.08 87.56 31-35 14.38 22.84 18.76 55.98 41-45 8.73 13.80 14.23 36.76 36-40 10.96 17.58 15.84 44.38 46-50 6.85 12.82 14.55 34.22 51-55 6.16 10.84 10.18 27.18 56-60 4.11 7.23 7.76 19.10 71-75 2.74 6.08 5.49 14.31 61-65 4.96 7.72 7.92 20.60 76-80 3,59 5.42 3.71 12.72 66-70 4.11 6.57 6.79 17.47 81-85 1,54 4.60 4.20 10.34 86-90 1,54 2.63 3,55 7.72 96-100 2.74 2.46 3.07 8.27 91-95 2.05 2,30 3,55 7.90 106-110 1.37 1.31 2.10 4.78 101-105 2.05 1.80 1.77 5.62 111-115 1.19 0.98 2.42 4,59 151-155 0.34 0.65 1.13 2.12 116-120 1.19 0.98 1.13 3.30 186-190 0.51 0.16 0.16 0.83 121-125 1,02 0.65 1.94 3.61 191-195 0.00 0.00 0.32 0.32 126-130 0.85 0.65 1,61 3.11 201-205 0.51 0.16 0.00 0.67 131-135 0.85 0.49 1,61 2.95 206-210 0.51 0.00 0.00 0.51 136-140 0.51 0.00 1.77 2.28 231-235 0.17 0.00 0.00 0.17 141-145 0.68 0.32 0.64 1,64 246-250 0.00 0.16 0.16 0.32 146-150 0.68 0.00 0.97 1.65 276-280 0.34 0.16 0.00 0.50 156-160 0.85 0.00 0.97 1.82 296-300 0.17 0.00 0.00 0.17 161-165 0.34 0.32 0.97 1,63 166-170 0.51 0.00 0.48 0.99 171-175 0.68 0.00 0.48 1.16 176-180 0.51 0.00 0.48 0.99 181-185 1,02 0.00 0.80 1.82 196-200 0.17 0.00 0.16 0.33 211-215 0.00 0.00 0.16 0.16 216-220 0.00 0.32 0.32 0.64 221-225 0.17 0.00 0.16 0.33 226-230 0.17 0.00 0.16 0.33 236-240 0.00 0.16 0.00 0.16 241-245 0.17 0.00 0.16 0.33 251-255 0.00 0.00 0.16 0.16 256-260 0.00 0.00 0.16 0.16 261-265 0.17 0.00 0.16 0.33 266-270 0.34 0.00 0.16 0.50 271-275 0.00 0.00 0.16 0.16 281-285 0.17 0.00 0.16 0.33 286-290 0.00 0.00 0.00 0.00 291-295 0.00 0.00 0.00 0.00

Table 6

T-criterion for dependent samples subject to dominance ATP> NSIS

and NSIS> ATP

Variable n Average
value by
∑ IS-SPS-NSIS Std. Off different. t p Will trust.
- 95,000% Will trust.
+ 95,000% SPS> NSIS 40 23.61 NSIS> ATP 20 13.47 10.71 4.01 0,000903 4.80 15.46

Note: the difference is noted, significant at the level of p <0.05.

Thus, the results of automated calculation and analysis in the indicated zone coincide with the general conclusions obtained on the basis of expert evaluation, which indicates that B. has information on the circumstances of the case and reports reliable (error-free) data.

Conclusion: during the study of the injured B., 09.09.1964g., Reactions are revealed that indicate that they have been provided with reliable (error-free) data indicating that he is aware of the fact that on May 22, 2016, he was located in the front passenger seat in the passenger compartment a car near Z., born on December 27, 1963, located in the driver’s place, in his presence, he put money in the amount of 80,000 rubles between the pages of an object resembling a book (organizer, notepad, notebook, etc.), or under the same circumstances, located outside the cabin a car, put on the car seat through an ajar window the indicated amount of money (other options for action are not presented in the case file), which were intended for Z. for assistance in not bringing OJSC to administrative responsibility. The conclusion is categorical.

The invention improves the efficiency and achieve a high degree of reliability. The method does not require special conditions for the place and time,

provides the opportunity for direct examination of the subject at various stages of criminal proceedings, including field conditions or a court session; the possibility of indirect and remote study of the test subject based on the assessment of materials reflecting the parameters of motor activity.

The inventive method, in contrast to the existing ones, is considered as an open system, which gives a number of advantages in the form of the possibility of combining with any modern methods and devices for collecting and analyzing received data, including the use of existing and original software products.

An advantage of the inventions is non-contact technology; low cost of research; low cost of the device; use in various degrees of illumination; drawing markers on the object of study is not required.

The proposed method can be considered as a new type of forensic forensics with inclusion in the classification and standardization system by obtaining arrays of data that are subjected to statistical processing according to the canons of large statistical groups with the calculation of the probability of conclusions; elimination of errors of the 1st type based on the conclusions about the reliability (inaccuracy) of the reported information related to situational anxiety or due to personality traits and errors of the 2nd type associated with missing a target; establishing an objective picture of an event (fact, action, etc.) by determining the reliability (infallibility), inaccuracy (error) of the information communicated; high accuracy in the calculation of reactions; a combination of direct expert assessment carried out by a competent person with automated due to the inventive device.

Claims (10)

1. A method for evaluating the reported information, including the impact on the subject with blocks of stimuli of a neutral, control and verification nature, fixing the parameters of motor reactions to the stimulus, studying the dynamics of changes in parameters grouped on the basis of factor analysis into a group of reactions, identifying trends in each parameter and / or a group of parameters and the assessment of the reported information by comparing reactions to stimulus blocks of a verification and control nature, characterized in that they carry out fixation by scanning motor reactions, in which three groups of verified reaction parameters are used: multilevel motility in the head, shoulder girdle, including forearms, hands, fingers, lower limbs, and body movements, facial motility - movements of the forehead, eyebrows, nose, nose, mouth, lips, physiologically determined motility - blinking, respiratory-forced movements, vegetative, the number of changes in the rate of speech, the number of latent periods, the number of pauses of hetization, Trends are identified according to four types of response, identified on the basis of dominant trends: the maximum and minimum response parameters to control blocks of stimuli, the increase and decrease in reaction parameters to stimulus blocks from neutral to test, while the stimulus is carried out by sets of blocks, after Impacts by a block of a neutral nature effect the blocks of a neutral, control and verification character in a chaotic manner with a different number of blocks, The reliability check is carried out automatically taking into account the types of response: when the reaction parameters for control stimulus blocks prevail over the response parameter stimulus blocks the verification character concludes that the information is reliable, the reaction parameters for the verification stimulus block prevail over the response parameters control nature conclude that the information provided is unreliable. 2. The method according to p. 1, characterized in that the scanning of motor reactions is carried out using non-contact motion sensors. 3. The method according to p. 1, characterized in that the fixation of motor reactions is carried out by audio fixation. 4. The method according to p. 1, characterized in that it additionally carry out the video recording of motor reactions using a video camera. 5. The method according to p. 1, characterized in that the stimulus effect is carried out for no more than 90 minutes in the aggregate of all blocks, while the periods of stimulus effect of each set of blocks of neutral, control and verification nature are chosen of the same or different duration. 6. The method according to p. 5, characterized in that for different durations of the blocks, the number and severity of motor reactions per unit time are calculated for each block of stimuli. 7. The method according to p. 6, characterized in that the reliability assessment is carried out by expert analysis based on the results of video recording of motor reactions, with the prevailing reaction parameters for control blocks of stimulus blocks over the reaction parameters for block of stimulus blocks of a checking nature in two blocks conclude that the information being reported is reliable, with the predominance of reaction parameters to test stimulus blocks over reaction parameters to control stimulus blocks, the conclusion is made that the information mations, in the absence of predominance of signs in two groups, determine the predominance of stimuli of reaction parameters to blocks of stimulus blocks of a control nature over reaction parameters to blocks of stimulus blocks of a verification nature and vice versa for the whole set of signs, if there is a predominance of more than 33.3%, they conclude categoricality, with values from 16.6% to 33.3% make a probabilistic conclusion, with values of 16.6% or less it is not possible to draw a conclusion. 8. The device for evaluating the reported information for implementing the method according to claim 1, characterized in that it contains an active unit based on a modular design including a microprocessor and at least one proximity sensor connected to a microprocessor programmed with the ability to process signals from the contactless (s) ) of the sensor (s), an automated assessment unit of the reported information, to one of the inputs of which an active unit is connected. 9. The device according to claim 8, characterized in that the proximity sensors are made in the form of ultrasonic sensors and / or motion sensors and / or an audio module. 10. The device according to p. 8, characterized in that it further comprises a passive unit for video recording of motor reactions, connected to another input of the unit for automated evaluation of the reported information.

Images