<<
>>

Experimental researches on processing of video shots of ultrasonic

4.3.1 Into structure of the equipment of an experimental research enter (a drawing 4.3).

PVEM with the minimum complete set:

- The processor intel core I3 with clock frequency not less than 2800 MHz;

- Operative memory not less than 4 Gb;

- Video card Radeon ™ HD 6470M;

- USB a socket;

- CD/DVD the disk drive;

- The monitor not below 19 inches;

- The keyboard;

- The manipulator of type "mouse";

The scanner with the target permission to 600 points on inch.

Software:

- MS Windows-7/8/10;

- Matlab R2017b with built in package Image Processing Toolbox

Drawing 4.4 - the Experimental equipment for check of methods and working capacity of algorithms of processing of the visualised information, an estimation of efficiency of identification, forecasting and quality Ш1Р at diyoagnostike prostate diseases

The equipment is equipped:

- Database with standards of condition PZH in a normal condition and razyolichnymi sample kinds of diseases, orderliness of standards is compounded with the structural model resulted in chapter 1 (a drawing 1.12);

- Database of patients (group of the questioned patients with various kinds of diseases of 98 persons);

- The knowledge base where descriptions atypical (neyostandartnyh) situations in diagnostics of diseases the prostates presented in the form of simplified frames (including from external sources of publications) are placed and ordered.

The equipment (a drawing 4.4) is capable to realise two operating modes:

- Practical regimen for the doctor of the urologist;

- Research regimen for the expert-scientist.

Thus ON assumes realisation of algorithms primary (a drawing 4.2) and secondary (a drawing 4.3) processings at use of base of standards and anketiyorovanija parametres, and in atypical situations of the knowledge base with possibility of its replenishment.

The scheme of algorithm for the doctor-urologist is presented in a drawing 4.5. In this regimen the confidence of correct functioning of algorithms of primary and secondary processing, formation PPR [61, 62, 63], acceptances diayognosticheskogo decisions, and in a non-standard (atypical) situation - effektivyonost uses of frames in the knowledge base and possibility of its replenishment is checked. OrgaYOnizatsija bases of standards it is constructed according to architecture of conceptual model of the investments resulted in chapter 1 (a drawing 1.10, 1.11, 1.12).

Drawing 4.5 - Algorithm of work of the equipment for the doctor-urologist

The second regimen is provided for the expert-scientist, and also - for proyogrammista for the purpose of studying of completeness of a database (base of standards) and base znayony, perfection in them of reference and frame descriptions, replenishment of data and use as auxiliary bases of the questioned data, confidence estimations at forecasting of development of disease and efficiency of treatment if the entrance information used by the equipment, assumes video sequence presence as result of inspection during opredeyolennogo time of the separate patient. The algorithm of work of the researcher - the programmer is resulted in a drawing 4.6.

Drawing 4.6 - Algorithm of work of the equipment at installation of programs and operating modes (beginning)

Drawing 4.6 - Algorithm of work of the equipment at installation of programs and operating modes (termination)

In the second regimen quality of classification (raspoznavayonija), working capacity in realisation of algorithms of forecasting of development zayobolevanija and (or) efficiency of its treatment is checked also at use and research of an approved technique.

Results of experiments have objectively shown working capacity razrayobotannyh methods and a way at complex processing visualised inyoformatsii with application PPR as diagnostic decision according to medical classification and the indications accepted as initial and formayo
lizovannye at construction of architecture of conceptual model (drawings 1.10, 1.11 and 1.12).

4.3.2 As an experimental material the complete set reyoalnyh video shots (computer images) ultrasonic of a prostate of patients with a normal current condition of a prostate and with various kinds of a pathology (various situations) has been selected.

The information from video shots was scanned by means of computer skayonera type HP LaserJet 1120, and to data file after scanning for razlichyonyh groups (sets) of shots following kinds of transformation according to realised basic functions and stages of initial and secondary data processing in SNNR were applied:

- Scaling of video shots, ghost of files of current data to uniform scale for simplification of the subsequent transformations (drawings 4.7 - 4.10)

Drawing 4.7 - Scaling of 1st video shot

110

Drawing 4.8 - Scaling of 2nd video shot

Drawing 4.9 - Scaling of 3rd video shot

Drawing 4.10 - Scaling of 4th video shot

- Masking of the basic video shots by their prodifferentiated masks for the purpose of augmentation of a sharpness of transitions and borders of fragments, especially the image foreground (a drawing 4.11)

Drawing 4.11 - Masking of the basic video shots

- Allocation of the foreground by selection of a zone of restriction of borders of the image of a prostate in the form of a rectangle, a triangle and a circle (drawings 4.12 - 4.15);

Drawing 4.12 - Allocation of the foreground in the form of a rectangle

Drawing 4.13 - Allocation of the foreground in the form of a triangle

Drawing 4.14 - Allocation of the foreground in the form of a circle

Drawing 4.15 - Applying of a grid of polar co-ordinates on ultrasonic

Prostate

- Masking of the foreground (applying) by means of Fermi's spiral for the purpose of reading of the current information in the form of uniform prostranstvennoyojarkostnoj (spatially-colour) realisation for the subsequent statisticheyoskoji spectral processing (a drawing 4.16).

Drawing 4.16 - Masking of the foreground (applying) by means of Fermi's spiral

As settlement indicators of quality [194] diagnostic reshajuyoshchih rules were used: diagnostic sensitivity (DCH), diagnostiyocheskaja specificity (DS), diagnostic efficiency (DE) a solving rule.

These indicators were calculated according to distributions of results konyotrolnyh tests and resulted in table 4.1.

Table 4.1 - Results of check tests

The surveyed Results of operation of rules In total
The positive The negative
n IP LO ИП+ЛО
n LP IO ЛП+ИО
In total ИП+ЛП ЛО+ИО ИП+ЛП+ЛО+ИО

Where r - number of a class of investigated disease; - quantity of images in control sample in an investigated class of diseases; n - quantity izobyorazheny without a pathology; IP - truly positive take peer kolicheyostvu class images ωrправильно classified by a surveyed rule; LP - false positive take peer to quantity izobrazheyony of a class j, wrongly carried by a solving rule to klassuj; LO - is false negative result: quantity of images of a class j r, wrongly carried by a solving rule to a class j; IO - truly negative reyozultat: quantity of images of a class j correctly classified reshayojushchim a rule.

For the resulted designations in table 4.1 calculation of indicators of quality is carried out according to expressions:

115

Let's conditionally consider, that to 1 class without patoyologii, pictures of ultrasonic concern 2 class - with a pathology.

In table 4.2 results of work of model intellectual siyostemy, realising structure a drawing 4.1, on control sample are resulted.

Table 4.2 - Results of check tests of diagnostics

DCH = 0,92, DS = 0,87, DE = 0,89

Results of experiments objectively show real possibility of their purposeful processing for the subsequent formation of support of decision-making according to system criterion function.

<< | >>
Scientific source Razumova Xenia Viktorovna. INTELLECTUAL SYSTEM of PROCESSING of the VISUALIZED INFORMATION FOR ESTIMATION And FORECASTING of the CONDITION of the PROSTATE. The DISSERTATION on competition of a scientific degree of a Cand.Tech.Sci. Kursk - 2018. 2018

Other medical related information Experimental researches on processing of video shots of ultrasonic:

  1. Introduction
  2. Methods of an estimation of a microbiocenosis
  3. 2.2. Research methods
  4. 2.2.1 Tool methods
  5. modern representations aboutan aetiology and a pathogenesis perinatalnoj encephalopathies
  6. THE TABLE OF CONTENTS
  7. INTRODUCTION
  8. a general characteristic
  9. materials and methods of experimental researches
  10. statistical processing of results of research
  11. THE CONCLUSION
  12. 1.2. A role of vegetative regulation in formation of an arterial hypertensia
  13. 2.3 Statistical methods of research
  14. THE MAINTENANCE
  15. Design of experimental researches.