criteria for early detection of psychosomatic disorders...
TRANSCRIPT
CRIteRIA FoR eARLY DeteCtIon oF PsYCHosoMAtIC DIsoRDeRs BY VIBRAIMAGe PARAMeteRs 141
DOI: 10.25696/ELSYS.VC2.EN.25
CRITERIA FOR EARLY DETECTION OF PSYCHOSOMATIC DISORDERS BY VIBRAIMAGE PARAMETERS
T. M. Novikova1, A. F. Bobrov 2, A. A. Kosenkov 2, V. Y. Shcheblanov2
1 Central health-sanitary unit № 91, FMBA of Russia, Lesnoy city, [email protected] state Research Center — Burnasyan Federal Medical Biophysical Center of Federal Medical
and Biological Agency (sRC–FMBC) of Russia, Moscow, Russia
Abstract: It is shown that vibraimage technology is a promising tool for the development of criteria for early diagnosis of psychosomatic disorders, in the tasks of improving the medical sup-ply of workers in unfavorable conditions. The developed formalized criteria and decisive rules allow recognizing presence/absence of hypertensive disease, diseases of the gastrointestinal tract, musculoskeletal exchange processes. Increase of accuracy of formalized criteria and decisive rules of early diagnostics of psychosomatic disorders is connected with application of artificial neural networks.
Keywords: Vibraimage technology, psychosomatic disorders, early detection criteria, decisive rules, artificial neural networks, working in adverse conditions.
Psychosomatics (Greek: “psyche” — soul, “soma” — body) is a branch of medicine and psychology that studies the influence of psychological (mainly psychogenic) factors on the occurrence and subsequent dynamics of somatic diseases (Malkina-Pykh, 2005). The basis of any psychosomatic illness is the body’s response to emotional experience, accompanied by functional changes and pathological changes in the body. Scientists have been interested in the problem of the relationship of psychological and somatic components for centuries since the time of Aristotle and Hippocrates. A sufficiently large number of models and theories of psychosomatic diseases have been developed. Medical statistics show that up to 70% of patients seeking medical care suffer from psychosomatic diseases. The most common diseases include hypertension, coronary heart disease, asthma, various types of dermatitis, diabetes mellitus, gastric ulcer and duodenal ulcer and cancer.
In total, scientists describe more than 40 psychosomatic diseases, and this number is constantly increasing due to environmental change (economic, social, environmental instability), deterioration of the psychophysiological and mental state of a person (fears, aggression, anxiety, low stress tolerance, exhaustion, etc), as well as hereditary diseases (Medvedev, 2013). Particular attention is paid to the early diagnosis of psychosomatic diseases with the medical support of workers whose work activities are associated with harmful and dangerous working conditions. In particular, in the medical care of nuclear workers. This is necessary for the timely implementation of rehabilitation and recreational activities aimed at preserving the health of personnel and ensuring the safety of the enterprise (Ivanov, Fedotov, 2016).
The development of express methods and criteria for the early detection of psychosomatic disorders is a topical scientific and practical task in assessing the health of workers at hazardous industries. A promising direction for solving this problem is the use of vibraimage technology (Minkin, 2017). The theoretical possibility of
24
T. M. Novikova, A. F. Bobrov, A. A. Kosenkov, V. Y. Shcheblanov
T. M. Novikova, A. F. Bobrov, A. A. Kosenkov, V. Y. Shcheblanov142
its use for the medical diagnosis of various mental diseases, diseases of the nervous system and pathology of the vestibular apparatus is presented in (Blank et al., 2013). It is based on the assumption that micromovements of the human head should be informatively associated with certain pathological conditions. In the development of this area in the Russian Scientific Center of Radiology and Surgical Technologies, studies have been conducted to determine the parameters of patients vibraimage with a histologically verified diagnosis of prostate cancer (Blank et al., 2013).
Similar studies were conducted in a group of healthy people who underwent research for the presence of cancrophilia syndrome, which showed a negative result. As a result of studies, a significant difference was found between the experimental and control groups in a number of parameters of vibraimage. This makes it possible to conduct operational screening for the early detection of people with prostate cancer (Blank et al., 2013).
The purpose of this study was to develop criteria for the early detection of certain types of psychosomatic disorders in terms of vibraimage parameters.
Materials and MethodsThe object of the research was the staff of one of the enterprises of the State
Corporation Rosatom. During periodic medical examinations and psycho-physiological examinations, the health status of 282 employees of the enterprise was examined. Psychosomatic disorders were grouped into the following groups: the presence/absence of any type of disorder; the presence/absence of hypertension (GB); the presence/absence of diseases of the gastrointestinal tract; the presence/ absence of diseases of the musculoskeletal system; the presence/absence of diseases of metabolic processes. In the course of psychophysiological examinations using VibraMED program (Vibraimage PRO, 2019), the parameters of vibraimage were recorded. The results of the study were analyzed using the methods of multivariate statistical analysis (Kim et al., 1989), implemented in the program STATISTICA v.8.0.
Research resultsIn order to select the informative parameters of vibraimage that reliably differentiate
the presence/absence of certain types of psychosomatic disorders, and to build the decisive rules for their formal identification, a stepwise discriminant analysis was used (Kim et al., 1989). The analysis is carried out using all parameters of vibraimage stored in VibraMED program database.
The table shows vibraimage parameters, ranked in order of information content for various types of psychosomatics. The ranking was carried out by the value of the Wilks’ lambda criterion (Λ). The designations of indicators correspond to those stored in the database: X — mean values, Vi — variability, S — standard deviation of the indicator. Because currently there are no data on the psychophysiological interpretation of all recorded parameters of vibraimage, a meaningful analysis of the results obtained is very difficult. The table also shows the magnitude of the errors
CRIteRIA FoR eARLY DeteCtIon oF PsYCHosoMAtIC DIsoRDeRs BY VIBRAIMAGe PARAMeteRs 143
of the 1st (probability of making a decision about the absence of a psychosomatic disturbance if there is one) and 2nd (the probability of making a decision about the presence of a psychosomatic disturbance if there is none) kind of developed formal decision rules. Error values of the 1st type do not exceed 10%, 2nd type — 20%.
It is important to emphasize that the constructed formalized decision rules for the identification of psychosomatic disorders have limitations, not acknowledging which leads to an increase in these errors. The result was generally as expected, since vibraimage parameters reflect the systemic response of the body to internal and external stimuli and factors (Bobrov et al., 2018) and it would be too optimistic to expect a high accuracy of identification of a specific type of psychosomatic disorder. One of the promising studies in this area is the use of artificial neural networks (ANN) for the identification of psychosomatic disorders, which, in contrast to the methods used for linear discriminant analysis, are nonlinear mathematical methods. In particular, the use of multilayer perceptron in information technologies of occupational medicine (Bobrov A. F., 2003).
Currently, various ANN learning algorithms have been developed for automatic selection of the neuron weights of all layers of a multilayer perceptron. Algorithms of learning implement the principle of learning with and without a teacher. The setting up of the ANN is carried out for each example from the training sample until the ANN begins to recognize all these examples with the required accuracy. Only after that, ANN is considered ready for use for the recognition of real objects shown. A tuned and trained ANN is able to effectively recognize the new objects presented to it, attributing them to one of the classes whose recognition it was able to train.
Tablethe results of certain types of psychosomatic diseases recognition
using formalized decision rules
№type of
psychosomatic disease
Leading informative indicators Recognition results using formalized decision rules
Parameter designation
MagnitudeWilks’
criterionp-level
Average % correct
classification
type 1st error,
%
type 2nd error,
%
1Psychosomatic diseases in general
Charm (P17)-X 0,55 0,000
87,6 7,5 17,3
P1-X 0,47 0,001
self-Regulation (P18)-Vi 0,46 0,005
Balance (P16)-Vi1 0,45 0,027
s1-Vi 0,45 0,063
Inhibition (F6)-X 0,45 0,068
T. M. Novikova, A. F. Bobrov, A. A. Kosenkov, V. Y. Shcheblanov144
№type of
psychosomatic disease
Leading informative indicators Recognition results using formalized decision rules
Parameter designation
MagnitudeWilks’
criterionp-level
Average % correct
classification
type 1st error,
%
type 2nd error,
%
2 Hypertonic disease
P2-X 0,472 0,000
95,9 3,0 9,1
F5 (fast)-Vi 0,449 0,000
F4-X 0,418 0,000
s5-X 0,411 0,000
s1-Vi 0,404 0,002
s5-Vi 0,404 0,002
A4-Vi 0,402 0,003
F4-Vi 0,398 0,009
3Diseases of the gastrointestinal tract
Charm (P17)-Vi 0,49 0,000
95,9 2,8 16,7P4-X 0,47 0,005
Balance (P16)-Vi 0,47 0,009
F1-X 0,47 0,030
A4 (fast)-X 0,46 0,048
A1-X 0,46 0,058
self-Regulation (P18)-X 0,46 0,059
energy (P8)-X 0,46 0,086
4Diseases of the musculoskeletal system
F5 (fast)-Vi 0,58 0,000
93,4 2,5 10,3
stress (P6)-Vi 0,54 0,000
F6-Vi 0,45 0,000
A3-Vi 0,44 0,001
F2-Vi 0,43 0,003
F7-X 0,43 0,006
suspect (P19)-Vi 0,43 0,008
suspect (P19)-X 0,41 0,103
Table (continued)
CRIteRIA FoR eARLY DeteCtIon oF PsYCHosoMAtIC DIsoRDeRs BY VIBRAIMAGe PARAMeteRs 145
№type of
psychosomatic disease
Leading informative indicators Recognition results using formalized decision rules
Parameter designation
MagnitudeWilks’
criterionp-level
Average % correct
classification
type 1st error,
%
type 2nd error,
%
5 Metabolic diseases
P1-X 0,41 0,000
95,8 4,3 4,0
F4-X 0,41 0,000
s1-Vi 0,40 0,005
Aggression (P7)-X 0,39 0,011
Aggression (P7)-Vi 0,39 0,012
Balance (P16)-X 0,39 0,026
Conclusions:1. Vibraimage technology is a promising tool for developing criteria for the early
diagnosis of psychosomatic disorders in the task of improving medical care for people working in hazardous conditions.
2. The developed formalized criteria and decision rules allow recognizing the presence/absence of hypertension, gastrointestinal tract disease, musculoskeletal system, and metabolic processes within the limits of the established limitations with an acceptable level of errors.
3. Improving the accuracy of formalized criteria and decisive rules for the early diagnostics of psychosomatic disorders is associated with the use of artificial neural networks.
References:
1. Blank M. A. et al. (2013), Prostate cancer screening method. Patent RU2515149, Electronic resource http: // http://psymaker.com/downloads/RU2515149.pdf
2. Blank M. A., Blank O. A., Metelev V. V., Minkin V. A., Kiselev D. G. (2013). Application of vibraimage technology for medical diagnostics, Elsys Corp., publishing 2013, http://psymaker.com/Bibliography/RU/2013/OncoVibra.pdf
3. Bobrov A. F. (2003). Information technologies in occupational medicine, Occupational medicine and industrial ecology, 2003, № 9, p. 20–26, (Russian language).
4. Bobrov A. F., Scheblanov V. Yu. (2018). Vibraimage technology: a new paradigm in psychophysiological surveys of personnel of enterprises and objects of the nuclear industry, Proceedings of the 1st International Open Science Conference, Modern Psychophysiology. Vibraimage Technology, SPb, Renome, 2018, p. 15–24, (Russian language).
Table (end)
T. M. Novikova, A. F. Bobrov, A. A. Kosenkov, V. Y. Shcheblanov146
5. Ivanov V. V., Fedotov D. D. (2016). Dynamics of mutual relations of psychophysiological adaptation and psychosomatic disorders in the process of professional activity of staff in an enterprise with harmful and dangerous working conditions, Medicine of Extreme Situations, № 3, 2016, p. 77–87.
6. Joe. Kim, CH. U. Mueller, W. R. Cleck and others (1989). Factor, discriminant and cluster analysis, trans. from English;; by ed. I. S. Enyukova. — M., Finance and Statistics, 1989. — 215 p. (Russian language).
7. Malkina-Pykh I. G. (2005). Psychosomatics: Reference practical psychologist, M, Publishing house Eksmo, 2005, 992 p.
8. Medvedev V. E. (2013). Psychosomatic diseases of the concept and the current state of the problem, Journal Archive of internal medicine, № 6 (14), 2013.
9. Minkin, V. A. (2017). Vibraimage. SPb., RENOME. DOI: 10.25696/ELSYS.B.EN.VI.201710. Schelkanova E. S. (2018). Vibraimage and express-diagnostics of psychophysiological state
of persons in dangerous professions, Proceedings of the 1st International Open Science Conference, Modern Psychophysiology. Vibraimage Technology, SPb, Renome, 2018, p. 19–25, (Russian language).
11. Vibraimage PRO (2019). Emotion Recognition and Behavior Detection System Manual, Elsys Corp. publishing, http://www.psymaker.com/downloads/VI10_ManualEng.pdf