This research is devoted to the development of hypothalamus structures target influence methods by means of projecting the stable optical stimuli onto retinoreflexogenic zones. Computer system of retino-hypothalamic functional system research (CS RHFS) will enable implementation of a specific method of hypothalamus condition diagnostic technique, which will, in turn, enable efficient techniques development for a specific treatment for such serious illnesses as alcoholism, drug addiction, infertility, ischemic heart disease, diabetes mellitus, essential hypertension, bronchial asthma etc.[1]
There is a direct relation between retina and hypothalamus. Hypothalamus regulates activity of nearly all organs and systems. And that is the reason why it is the basic target for diagnostic and therapeutic influence in case of patients with neuropathological syndromes. Affection or abnormality of hypothalamus function underlies these diseases. Specialists of Donetsk State Medical University and Family Medical Problems Research Institute deal with developments in the sphere of searching the means of target diagnostic and therapeutic influence on separate hypothalamus structures. Researches target is detection of retino-reflexogenic zones, which are situated on the retina, impacting on which by an optical stimulus, we stimulate structures of hypothalamus, which, in turn, regulates separate organs or structures activity. Computer system of the human retino-hypothalamic functional system analysis was developed to solve this task. Computer system of the human retino-hypothalamic functional system analysis, developed within the framework of this research, is a constituent part of a bigger project of the human retino-hypothalamic functional system research in organic states medicine and gynecology. Among a considerable list of diseases, conditioned by hypothalamus function affection or abnormality, this research makes an emphasis on the development of the new methods of treating patients, whose marriage is infertile, in borderline cases medicine sector.
- Search of adequate optical stimuli parameters, which cause hypothalamus reactions.
- Mapping retino-reflexogenic zones.
CS RHFS will enable:
- considerable increase of the efficiency of diagnostics, treatment and prevention of a whole range of different diseases;
- significant reduction of these diseases diagnostic and treatment costs;
- decrease of economic losses due to temporary inability to work and disability;
- substantial growth of the level of population psychophysical adaptability to negative social and psychological exposure.
For the search of stimulating optical stimulus (SOS) parameters, which cause hypothalamus reactions, the human retino-hypothalamic functional system (RHFS) neuronet model was developed within the framework of this research. This model enables forecasting and formation, along with genetic algorithm, of the expected appropriate organism response to retina stimulating effect in accordance with a preselected criterion. With this amendment, the model is able to substitute for the system under consideration (RHFS) [4].Deviation from the expected reaction under SOS influence as compared to the background reaction can be a criterion of decision efficiency assessment. The use of neuronet simulation is offered as a method of RHFS model implementation, since neuron nets are successfully applied when solving forecasting problems. Hypothalamus structures are impacted upon with the help of projecting a great number of light conduits, located on a contact lens, onto retino-reflexogenic zones, by means of light-emitting diodes. The controlling system enables regulation of coordinates, spectral distribution, energy density and SOS exposure (action time). Search of optical stimuli optimal parameters is an NP-complete task having high dimensionality. Since the number of light conduits is quite big, computational complexity of such a procedure is obvious. Computational procedure complexity can be calculated, according to the following formula:
where "m" is the number of light conduits;
"С" is the number of operating regimes of one light conduit.
The number of one light conduit operating regimes can be determined according to the following formula:
where Kf is the number of controlled parameters;
"i" is a parameter's sequence number;
"ni" is the number of i-parameter levels.
In the case under analysis, object's complexity makes 2048 for one light conduit. Then, overall object's complexity is:
For the search of optical stimuli optimal parameters, it is necessary to solve the combinatorial optimization problem. There are several approaches to it. Exhaustive, heuristic and evolutionary methods can be used as algorithms. The heuristic method can also be used in the case of this class of problem. The use of a certain rules set, made up by an expert group of a particular problem area, underlies such algorithms. The main disadvantage of this method is its subjectivity. That is, the rules applied are made up by experts, who rely on their own experience, knowledge and intuition. Therefore, an algorithm, developed on the basis of these rules, can work with certain errors implicit in it. Since an assigned task belongs to multiparametric functions combinatorial optimization, none of the listed methods will enable arrival at an optimal solution during an acceptable period of time. A more efficient evolutionary method (genetic algorithms) was used to solve this problem. A modified GA was developed for searching optical stimuli optimal parameters. The issues, related to individual encoding, objective function and reproduction operator selection, development of problem-oriented crossing-over and mutation operators were solved, and algorithm stopping criterion was chosen with the aim of the modified GA realization. It is offered to use neuronet model of the human RHFS as an objective (Fitness) function, which enables solution quality assessment. Findings: modified GA was developed in the course of theoretic and practical research. Joint use of GA and NS as an objective function enables arrival at a suboptimal solution during an acceptable period of time. The obtained results will be used for the development of efficient methods of specific treatment for a number of severe diseases.
This work is not completed yet. Final completion: Dec. 1, 2011. Full text of the work and materials on the subject can be obtained from the author after that date.