RU2318201C1 - Method for microscopic examination of a specimen - Google Patents

Abstract

FIELD: automated analysis of detection of pathologies in examined specimens of a material, for example, in oncomorphology.

SUBSTANCE: in the method, macroscopic examination of macro-preparation, meant for production of a specimen, is performed, for which purpose digital images of macro-preparation are produced together with enlarged images of its marked macro-sections to determine the specimen excision location. The specimen is excised and its microscopic examination is performed, for which purpose automatic scanning is performed, as a result of which digital images of the specimen are produced, specimen image zones are marked for their further detailing and by means of objective with high magnification automatic scanning of marked zones of specimen images is performed. Produced images are described by formalized signs from a given set and on basis of resulting combination of signs those images are selected from the base of standard images, which have similar combinations of signs, after that resulting images of marked zones of specimen images are compared to selected standard images, on basis of results of which condition of the specimen is evaluated.

EFFECT: increased information capacity and objectivity during microscopic examinations of specimens, ensured high quality of examination material and trustworthiness and objectivity of produced results, ensured multi-functionality of examinations and correct micro-anatomical diagnosis not only concerning differential diagnostic of benignancy or malignancy of process, but also concerning histogenesis of given neoplasm.

18 cl, 15 dwg, 3 tbl

Description

The proposed solution relates to methods for the study and analysis of materials using optical and computer tools and can be used, in particular, for the analysis and identification of pathologies of the studied samples of material, for example, in oncomorphology.

A known method of microscopic examination of a sample (RU 2267997 C1, publ. 20.01.2006), in which the sample is preliminarily obtained, the sample obtained is prepared for microscopic examination, and the sample is microscopically examined to obtain a fine needle aspiration biopsy of the nodular thyroid gland under ultrasound control followed by a cytological examination of the obtained punctate, while additionally the liquid part of the punctate is examined by the method of wedge-shaped dehydrate cations, then a drop of the obtained supernatant is dried on a glass slide for 18-24 hours at room temperature and with minimal air movement, and then the structure of the obtained sample is analyzed in the form of a facies under a microscope and, in the presence of arcuate cracks in the peripheral zone of the facies, a conclusion is made about the benign nature node, and if a network of "broken" cracks is detected in the peripheral zone of the facies, it is about malignant degeneration of the node. This method does not provide high quality of the studied material and, accordingly, the reliability and objectivity of the results obtained, does not provide multifunctionality of studies using a number of criteria that help the pathologist in making the correct histological diagnosis, not only in relation to the differential diagnosis of benign or malignant process in the thyroid gland, but also and in determining the histogenesis of this neoplasm.

A known method of microscopic examination of a sample (RU 2234099 C1, publ. 08/10/2004), which can be adopted as a prototype and in which the sample is preliminarily obtained, the sample obtained is prepared for microscopic examination, and a microscopic examination of the sample is carried out using histological sections neoplasm tissues, the images of which are obtained using a computer analyzer, the nuclei of small lymphocytes are isolated on the resulting image, their ploidy is determined, ayut her standard, which is carried out using a computer to determine the mean ploidy nuclei zones germ cell tumors, and the value of this index is performed diagnostics stages of carcinogenesis. However, this solution also does not have a sufficiently high information content and objectivity and does not allow the pathologist with any unclear histological picture in the studied thyroid preparation to determine the nature of the process, evaluate its malignancy, establish histogenesis, and the high quality of the studied material and, accordingly, the reliability and objectivity of the results obtained, multifunctionality of studies using a number of criteria that help the pathologist is not ensured anatomist staged correct histological diagnosis not only for the differential diagnosis of benign or malignant process, but also in identifying the histogenesis of this neoplasm.

The technical result achieved by the implementation of the claimed invention is to increase the information content and objectivity during microscopic examination of samples, in particular, to ensure, with any unclear histological picture in the studied thyroid preparation, determine the nature of the process, with an assessment of its malignancy, to establish histogenesis, to ensure the high quality of the studied material and, accordingly, the reliability and objectivity of the results obtained, to ensure multifunctionality and studies using a number of criteria to help the pathologist in making the correct histological diagnosis, not only in relation to the differential diagnosis of benign or malignant process, but also in determining the histogenesis of this neoplasm.

The technical result is achieved due to the fact that in the method of microscopic examination of a sample, including preliminary preparation of the sample,

a) conduct a macroscopic study of a macro product intended to obtain a sample using a macro image preparation system including a digital camera and a positioning and lighting device, and a computer connected to it, during which

b) conduct a survey photograph of the macro product, as a result of which digital images of the macro product are obtained,

C) carry out the allocation of the marking of the macro-section, intended to obtain a sample on a digital image of the macro, and

g) receive an enlarged digital image of the selected macro-stage, while the indicated operations c) and d) are repeated until the place of cutting the sample is determined, and

d) cut out the sample from the macro,

e) carry out the preparation of the obtained sample for microscopic examination,

g) carry out a microscopic examination of the sample using a computer-imaging system for acquiring sample images containing a microscope, a digital camera, a matching module for joining a digital camera and a microscope, during which

h) position the sample on the microscope stage,

i) produce an initial review view, for which a 4x lens is installed and focus is carried out,

j) conduct an automatic scan, as a result of which digital images of the sample are obtained, which are displayed on a computer screen,

k) distinguish by marking the zone of images of the sample for their further detail

m) carry out the change of the microscope lens to a lens with high magnification and its focusing,

m) conduct automatic scanning of the selected areas of the image of the sample, resulting in an enlarged digital image of the selected areas of the image of the sample, which are displayed on the computer screen,

o) describe the obtained images using formalized features from a given set of formalized features,

p) based on the combination of formalized features obtained, from the set of reference images contained in the base of reference images, those reference images that have similar parameter values and similar combination of signs are selected, after which

c) comparing the obtained images of the selected areas of the image of the sample with the selected reference images,

r) the results of which judge the state of the sample.

In this case, automatic scanning during microscopic examination is carried out by means of predetermined coordinate displacements with restriction of the extreme positions of the subject table by displacement drives connected to the computer, and the disks of the movement of the subject table are made in the form of a block of gear motors consisting of stepper motors with gears, while for research use the MIKMED-2 microscope and the coordinate table KS-30.

The device for positioning and lighting a macro-image acquisition system contains an object table for installing a macro-preparation, on which markings are made with a step of 1 cm in two mutually perpendicular directions, and barriers are installed on the object table from the rear edge and sides, while a macro preparation is placed on the object table blot with a napkin.

Photographing a macro product is carried out in a limited volume enclosed on all sides with a lockable opening for access inside, and the walls of the fence are made of easily washable material and have an internal reflective surface and an external light-absorbing surface, while the positioning and lighting device of the macro image preparation system contains a subject a table for installing a macro preparation, on the basis of which a means for positioning a digital camera and light sources is rigidly fixed a tripod consisting of three vertical guide-rails, which are arranged in the horizontal section in a triangle, wherein the light sources are arranged symmetrically about the optical axis of the digital camera. At the same time, the macro preparation is photographed under diffuse lighting by placing translucent shutters made of white translucent easily washable material inside the limited volume between the light sources and the object table, using halogen lamps that are located above the place of installation of the digital camera as light sources.

As a macro preparation, it is possible to use surgical material containing a thyroid neoplasm, the isolated macroareas are a thyroid neoplasm and / or its characteristic areas, for example, a tumor node or a capsule of a node, and a histological tissue section of the thyroid neoplasm stained with hematoxynline and eosin or picrofuchsin according to Van Gieson.

, где L=1 см, l - то же расстояние в пикселях изображения, а коэффициент формы рассчитывают по формулеIn addition, measurements are made of the macroscopic sections of the thyroid neoplasm, and the following parameters are measured: linear dimensions along the major and minor axes of inertia L, perimeter P, area S, shape coefficient K, while before starting the measurements, a scale factor is determined to bring the measurement results to metric length units and area:

, where L = 1 cm, l is the same distance in pixels of the image, and the shape factor is calculated by the formula

The results of macroscopic and microscopic examinations are used for histological diagnosis of thyroid neoplasms, for which purpose data on the probability of a diagnosis in the presence of each of the combinations of signs available in the selected reference images are additionally displayed on a computer screen and use this data when comparing the obtained images with Selected reference images and diagnosis.

Additionally use a system consisting of computer-assisted laboratory and consultation points, each of which contains a computer, and a computer connected to image acquisition systems, respectively, of a macro preparation and a sample, is used as a laboratory computer,

after the survey photograph of the thyroid neoplasm at the laboratory assistant’s point, the digital images of the thyroid neoplasm obtained are sent through a computer communication network to the consultation point, at which

carry out the selection of the marking of the macro-section, designed to obtain a tissue slice, on a digital image of a thyroid tumor,

and send to the laboratory station an indication of obtaining an enlarged digital image of this macro-section,

and after receiving an enlarged digital image at the laboratory station, the macro section is sent to a consultation center,

however, the indicated operations of marking the corresponding macro-section and obtaining its enlarged digital image are repeated until the place of the tissue cut is determined at the consultation point, after which an instruction is sent to the laboratory assistant to perform the tissue cut,

at the same time, digital images of a tissue section obtained during the process of automatic scanning are simultaneously displayed on computer screens of laboratory and consultation points,

after which, at the consultation point, marking of the zones of images of the tissue section is marked out for labeling for further detailing, and instructions are sent to the laboratory station for automatic scanning and obtaining enlarged digital images of the selected zones of images of the tissue section,

which display simultaneously on the screens of computer monitors of the laboratory and consultation points, after which operations c) -d) are performed at the consultation point, as well as histological diagnosis of thyroid neoplasms.

When carrying out at least one extraction operation by marking the corresponding macro-section, the location of the thyroid neoplasm incision is additionally marked and the direction about the corresponding incision and its enlarged digital image are sent to the laboratory assistant.

In addition, constant two-way audio communication is carried out between the laboratory and consultation centers, the patient’s initial data are generated and sent to the consultation center, and patient data are used to control the receipt of images and the transfer of images to a computer at the laboratory.

At the consultation point, an information package is additionally formed containing the results of measurements of P, L, S values and a description of the thyroid neoplasm with the attached received images and a formatted protocol for placement in the database. At the same time, measurements are taken, a description of the thyroid neoplasm is formed, and a protocol is drawn up at the consultation center during the cutting of a tissue section at the laboratory station and its preparation for microscopic examination. If necessary, at a consultation point, when receiving an image of a macro-section of a macro preparation, remote control of the system of obtaining images of a macro-preparation is performed to position the macro-section of a macro-preparation.

When determining the diagnosis, an informational expert decision support system is additionally initiated, including a database containing color reference images of thyroid neoplasms and tissue sections of thyroid neoplasms and their formalized descriptions according to a reference set of signs, and program modules for managing this database, using which carry out a detailed description of the histological picture of the obtained digital images, which are compared with the existing reference combinations of attribute values for the production of the final histologic diagnosis.

Additionally, an integrated training system is initiated, made in the form of an electronic textbook containing a set of diagnostic data in the form of texts, photographs, video clips and / or computer graphics.

The proposed method allows the use of an extensive database of histological images with detailed and formalized (uniform, standard) and comprehensive descriptions given by experts in this field. The approach proposed for the first time to make a histological diagnosis using computer technology - an expert system - significantly distinguishes this invention from the routine, traditional methods of morphological diagnosis of pathological processes.

To ensure these results, preparatory operations, technology and means of introducing and processing information features, means of providing analysis and presentation of the resulting information for decision-making are no less important.

The present invention solves the problem of creating a method that, using an interactive computer program - an expert system - assists the pathologist in the diagnosis and differential diagnosis of malignant and benign tumors and tumor-like processes of the thyroid gland.

Next, the above operations of the method are considered and the “necessity and sufficiency” of the combination of the claimed essential features is illustrated by the example of a microscopic examination of tumors and tumor-like processes of the thyroid gland - one of the most difficult localizations from the point of view of histo- and cytological diagnostics.

The invention is illustrated by drawings, where

figure 1 - structural stages of an automated remote histological rapid diagnosis;

figure 2 is a structural diagram of the technical means of automated histological diagnostics (ATLANT complex);

figure 3 - image of the macro thyroid gland;

figure 4 - image of a micropreparation of the thyroid gland;

5 is an illustration of a poorly prepared preparation for shooting;

6 - examples of defective micropreparations;

Fig.7 is a script for the preparation of a macroanalysis system for remote diagnostics;

Fig.8 is a scenario of remote control cutting of operational material;

Fig.9 is a macroanalysis scenario;

figure 10 - scenario for the preparation of micropreparations;

11 - preparation of a microanalysis system for automated diagnostics;

Fig - script remote search for informative fields on the MP;

Fig.13 is a microscopic analysis;

Fig - software structure of the expert system;

Fig - sequence of work with an expert decision support system.

Positions on the presented figures, in particular, are indicated:

figure 1 - at the first level of detail:

1 - operational material (OM);

2 - study of OM;

3 - study of micropreparation (MP);

4 - determination of the volume and tactics of the (surgical) operation. At the second level of detail;

5 - preparation of OM;

6 - preparation of a macroscopic analysis system;

7 - remote video control cutting macro preparation (OM);

8 - macroanalysis (OM);

9 - preparation of a micropreparation, (MP);

10 - preparation of a microanalysis system;

11 - remote search for informative fields on (MP);

12 - microanalysis.

In figure 2.

The technical support of the ATLANT complex is conditionally divided into four subsystems:

13 - computer (laboratory assistant);

14 - macro image preparation input system;

15 - image input system of micropreparations;

16 - communication system;

17 - a digital camera (camera);

18 - positioning and lighting device;

19 - camera (television camera);

20 - matching module;

21 is a microscope;

22 - a pairing device;

23 is a computer consultant.

On Fig:

24 - user interface;

25 - database replenishment subsystem;

26 - data analysis subsystem;

27 - user training subsystem;

28 - subsystem for editing features and nosological forms;

29 - a database of images and descriptions;

30 - database of signs and nosology.

According to the requirements of the International Histological Classification of Thyroid Tumors (World Health Organization, Geneva, 1988), benign and malignant neoplasms are distinguished among epithelial tumors, including carcinomas (cancers - follicular, papillary, medullary, anaplastic (undifferentiated) and mixed). Recently, there has been a tendency to single out a group of low-differentiated carcinomas, which include the so-called insular cancer. In the group of papillary cancer, a follicular variant of papillary cancer is distinguished, which prognostically, biologically and morphologically differs from follicular cancer. Diagnostic difficulties are also presented by some forms of medullary cancer, differential diagnosis of follicular cancer and benign follicular adenoma, etc. In the usual practice of the morphologist in the diagnosis of these forms, subjectivity is often allowed, the selection of some signs and ignoring others, etc. The absence of uniform, standard criteria and “one hundred percent” marker features does not allow an accurate and timely diagnosis to be made, and, consequently, an adequate treatment.

Necessary clarification on graphic materials.

Figure 1 shows the structural steps of an automated remote histological rapid diagnosis, carried out as a result of using the proposed method.

The first (upper) level of detail reflects the main stages of an automated histological rapid diagnosis (Fig. 1) and includes 4 stages: taking a macro preparation, which is used as an operative material (OM) (1), an OM study (2), a micropreparation study (MP) ) (3) and analysis of the results obtained, on the basis of which the volume and tactics of a surgical operation can be determined (4).

At the second level of detail, the process is divided into 8 stages, including the selection of surgical material (macro preparation) and its preparation for automated research (5), performance testing and tuning of the macroscopic analysis subsystem (6), the selection of an informative zone for cutting, remote video control of cutting a macro preparation (7) ) a highly qualified specialist consultant, the study of macropreparations remotely, using communication channels, automated measurement of the morphological properties of pat of biological sites and the formation of a formalized conclusion on a macrodrug: macroanalysis (8), preparation by a consultant of a micropreparation, (9) including cutting out material for freezing in a microtome, fixing, staining and imprisoning, checking the performance and tuning of a microanalysis system (10), remote study of a micropreparation specialist consultant using a programmable microscope stage (11), comparing research results with databases using an information system, accumulator uyuschey structured knowledge and experience of leading experts in the field of histological diagnosis - microanalysis (12).

The claimed method provides for the diagnosis of thyroid tumors using the information-measuring complex of decision support ATLANT. In normal practice, the histological diagnosis of thyroid tumors is based on visual macro- and microscopic examination of specially prepared histological preparations (macro- and micropreparations). Diagnostic value is the analysis of tissue structures and details of cellular morphology in the images of the preparations obtained under the microscope, provided that the informative zone of the surgical material is correctly selected for their preparation. The application of the claimed method allows to increase the accuracy and speed of decision-making on histological diagnosis during the rapid diagnosis in complex cases of diseases during surgical operations. It can be used in any healthcare institution in which histological diagnostics (both planned and express) is carried out with the appropriate automation equipment (ATLANT computer complex).

Automated histological diagnostic tools include two physician workstations (AWPs). The first workstation (conditionally laboratory point) is equipped with the ATLANT information-measuring complex, auxiliary devices (freezing with a microtome for preparing thin sections 5-Ymkm), materials and tools for cutting and concluding preparations (surgical knife, tweezers, scissors, ruler, tray, board, gloves, boxes, coverslips and slides), as well as chemical reagents for fixing and staining (formalin, alcohol, hematoxylin, eosin water, carbol-xylene, xylene, balm). The second workplace (probationary consultation point) is equipped with a computer. Both stations are connected by a computer and audio communication system.

The technical support of the ATLANT complex is conditionally divided into four subsystems: a computer (laboratory technician’s computer) (13), macro image preparation input system (14), micropreparation image input system (15), communication system (16) (see Fig. 2).

The macro image input system (14) includes: a digital camera (camera) (17), a photobox with a special positioning and lighting device (18), a foot pedal to control the capture and transfer of images to a computer. The micropreparation image input system (15) consists of a camera (camera) (19), a matching module (20) for docking the camera and a microscope (21), equipped with a motorized stage drive and navigation control module. The movement of the microscope stage is controlled both from the computer and from the positioning joystick. Both input systems are connected to a computer equipped with a video input device and special software (interface device) (22). The communication system (16) includes an optical modem with access to a fiber optic communication line. It should be noted that for the full functioning of the complex, a prerequisite is the connection to the fiber-optic network of the computer of the consultant (23), equipped with special software.

The ATLANT complex also includes an informational expert decision support system for thyroid tumors and a training system in the form of an interactive computer film.

Requirements for the key characteristics of the complex equipment:

1. The processor is no worse than the Pentium IV with a clock frequency of 3 GHz, with at least 1 GB of RAM, and with at least 120 GB of long-term memory.

2. TRUE COLOR monitor, 19 inches diagonal, resolution 1280 × 1024.

3. A digital camera with a photo frame size of at least 1280 × 1024.

4. Camera (camera) in the microsystem: work with the image in real time (the frame change time is not more than 40 ms, the image format is not less than 768 × 576).

5. Biological microscope with a special motorized drive and a set of x4, x10, x20, x40 lenses.

7. The bandwidth of the communication line is at least 2 Mbps.

The main stages that are carried out when implementing the inventive microscopic examination method will be shown using an example of a sample as a histological section of thyroid tumor tissue stained with hematoxinline and eosin or picrofuxin according to Van Gieson.

The studied macro- and micropreparations (Figs. 3, 4) must be prepared with a quality sufficient for histological examination in accordance with the Instruction for the study of biopsy and cytological material approved by the Ministry of Health.

The preparation of the macroanalysis system includes 6 procedures (Fig.7). At this stage, turn on the power, adjust the lighting, set the recommended shooting mode (distance to the camera, zoom), download macro analysis software. Then lay out the operating material (macro preparation) in accordance with the markings on the subject of the photobox. Macro preparations before placement on the subject table of the macro image preparation system (photobox) should be blotted with a napkin, as spilled liquid can be interpreted in the image as part of the drug (figure 5). In order to verify the correctness of the images entered from the photodetector, make a control survey of the operational material with a visual assessment of the images on the monitor in comparison with the real sample. Check the physical connection to the fiber optic network and the TCP IP protocol settings. A sufficient condition for a consultation is to display the pages of the "consultant" interface when referring to the corresponding link.

For preliminary obtaining a histological tissue section, a macroscopic study of a macro product, which is an operating material containing a thyroid tumor, is performed using a macro image preparation system including a digital camera and a positioning and lighting device, and a computer connected to it, during which the surgical material is selected and preparing it for automated research, health checks and settings adjustment of a macroscopic analysis system, selection of an informative zone for cutting, remote video control of cutting a macro product by a highly qualified specialist consultant.

A macro photograph is surveyed, as a result of which digital images of the macro product are obtained.

Marking of at least one macro-site, which is a neoplasm of the thyroid gland and / or its part, is carried out on a digital image of a macrodrug, and an enlarged digital image of each macro-site is obtained.

At the same time, these operations are repeated until they receive images of all the macro sites selected for the study and determine the place for obtaining the sample, a histological section.

Measure the macro-sites of thyroid neoplasms.

The method for making measurements (SVI) of pathological macroregions of thyroid gland micropreparations using an automated macroscopic analysis system includes the following main steps:

1. Definition of initial data, such as the purpose of the SVI, requirements for the accuracy of measurements, conditions for performing measurements.

2. The formulation of the measurement problem and the description of the measured values.

3. The choice of method and means of measurement, auxiliary and other means.

4. The establishment of the sequence and content of operations in the preparation and execution of measurements, as well as the processing of intermediate and final results.

5. Organization and conduct of an experiment to assess the accuracy of the SVI.

6. The establishment of the attributed characteristics of the measurement error, taking into account the source data for the SVI.

7. Development of procedures and standards for monitoring the accuracy of the obtained measurement results.

8. Development of a draft document on the SVI.

9. Metrological examination of the document on the SVI.

10. Certification of SVI.

The scope of application indicates the scope, the name of the measured values, the characteristics of the measured quantities (ranges), the characteristics of the measurement objects. The requirements for measurement error are established taking into account all components (methodological, instrumental, introduced by the operator). The measurement conditions are set in the form of nominal values or the boundaries of the ranges of possible values of influencing quantities.

Initial data for the development of SVI

Appointment SVI. This measurement method is intended for morphological studies of macro preparations on their color images with automated remote histological express diagnostics using the ATLANT information-measuring complex for decision-making support.

The objects of measurement are pathological (tumor and non-tumor) tissues of the thyroid gland, removed during surgical operations. The choice of surgical material is carried out by the operating surgeon. Before performing measurements, the operating material (macro preparation) does not require special processing.

Scope - in automated remote diagnostics in laboratories of histological express diagnostics of clinics.

Measured physical quantities, as well as their ranges (table. 1):

Table 1 Measured physical quantity of a measured quantity Units Change range one Distance between reference points cm 0.5-10 2 The perimeter of the selected object cm 3-30 3 Area of the selected object cm ² 0.8-78 four Shape factor dimensionless 1-2 * Reference points - points characterizing, according to the expert, the boundaries of the informative area.

Formulation of the measurement problem

The set of measured characteristics is determined in accordance with the requirements of the description of biopsy and surgical material. Measurements are carried out on color digital images using a special program that is part of the automated remote histological diagnostic software. At the first stage, the most informative image is selected, on which pathological areas are clearly visible. Before taking measurements, the scale and the measured value are entered. Then the user interactively marks the boundaries of the areas whose parameters are automatically calculated and entered into a formalized macroscopic conclusion.

When obtaining images, the material under study is located on the photobox object table, on which a scale grid is applied with a step of 1 cm. Using the scale grid, the system is checked for performance before starting measurements and visual control is made by the user during measurements.

, где L=1 см, l - то же расстояние в пикселях изображения, а коэффициент формы рассчитывают по формуле

.As a rule, four types of measurements are taken on the image of a thyroid gland macro product: linear dimensions along the major and minor axes of inertia L, perimeter P, area S, shape coefficient K, and before starting the measurements, a scale factor is determined to bring the measurement results to metric units of length and area:

, where L = 1 cm, l is the same distance in pixels of the image, and the shape factor is calculated by the formula

.

Measurement Error Requirements

In the absence of automated measuring instruments for the morphological characteristics of macro preparations in medical practice, these measurements are carried out manually using a ruler. The certified ruler error in determining the distance between two points is 1 mm. In real conditions, the experiments showed that the doctor determines the linear size of the tumor (which in most cases is 3-4 cm) with an error of 3 mm. In this case, the area is determined by the product of the dimensions of the axes of inertia, thus, the round shape is approximated by a square, which leads to an increase in the error.

Using an automated system of macroscopic analysis, it is possible to increase the accuracy of the measured parameters. For the same typical tumor sizes, measurement system errors are set at the following level (Table 2):

table 2 Measured physical quantity System error (in%) one Distance between reference points 3 2 The perimeter of the selected object 3 3 Area of the selected object 3 four Shape factor 5

Operations to prepare for measurements.

Measurement preparation operations are aimed at obtaining a digital informative image of sufficient quality. Image quality is considered sufficient for measurements if:

- on it pathological areas are clearly visible, the sizes of which are required to be determined,

- the image is focused, there is no blur and obvious optical distortion,

- there are no smudges in the frame field that distort the shape of the preparation.

After the completion of macroscopic studies, the macrodrug proceeds to a microscopic examination of the sample — a histological section of the tissue. The sample is obtained by cutting a histological tissue section, then the resulting tissue section is prepared for microscopic examination. It should be borne in mind that the process of studying micropreparations under a microscope is affected by compliance with the norms of cutting, freezing and staining of tissue sections. The preservation of the real picture of the tumor on the preparation subjected to treatment and the adequacy of the assessment of the observed structures largely depend on this. Figures 6a, 6b, 6c and 6d show a number of examples of images of correctly prepared histological sections and preparations with artifacts, i.e. with defects that are interferences with computer diagnostics (wrinkles, air bubbles, dirt, foreign objects).

On figa presents a properly prepared drug. Image of the thyroid gland of normal structure, x10 magnification, slice thickness 5 μm, hematoxylin-eosin stain.

On figb shows a drug prepared poorly. The image of the thyroid gland is normal in structure, x10 magnification, slice thickness 15 μm, hematoxylin-eosin stain, the sample was overexposed in hematoxylin, and because of the large thickness of the slice it is impossible to focus the drug, and the nuclei overlap each other, in addition, there is a fold in the drug.

On figv also shows a drug prepared poorly. The image of the thyroid gland is normal in structure, x10 magnification, slice thickness 5 μm, hematoxylin-eosin staining, however, the drug was overexposed to eosin.

6g also shows a poorly prepared preparation. The image of the thyroid gland is normal in structure, x10 magnification, slice thickness 10 μm, hematoxylin-eosin stain, but under the coverslip there is an air bubble that does not allow the image to be correctly focused.

After receiving the slice, it is microscopically examined using a computer-imaged sample imaging system containing a microscope, a digital camera, a matching module for joining a digital camera and a microscope, during which a tissue section is positioned on the microscope object table, and an initial review is performed to conduct which set a 4x lens and focus it.

An automatic scan is carried out, as a result of which digital images of a tissue section are obtained, which are displayed on a computer screen, highlighted with a marking of the image section of a tissue section for further detailing, the microscope lens is changed to a high-magnification lens and focused, and the selected image zones are automatically scanned tissue section, resulting in enlarged digital images of the selected areas of the tissue section images that are displayed on the monitor screen computer, measure the geometric parameters of these selected zones of images of tissue sections, describe the images using formalized features from a given set of formalized features, according to the combination of formalized features selected from the set of reference images contained in the base of reference images, those reference images that have similar values of parameters and similar combinations of features, after which the obtained images are compared with the selected mi reference images, the results of which are used for histological diagnosis of thyroid neoplasms.

Automatic scanning during microscopic examination is carried out by means of predetermined coordinate displacements with restriction of the extreme positions of the subject table by displacement drives connected to the computer, and the disks of the movement of the subject table are made in the form of a block of gear motors consisting of stepper motors with gears, using a microscope to carry out studies MIKMED-2 and object table coordinate KS-30.

The device for positioning and lighting a macro-image acquisition system contains an object table for installing a macro-preparation, on which markings are made with a step of 1 cm in two mutually perpendicular directions, and barriers are installed on the object table from the rear edge and sides, while a macro preparation is placed on the object table blot with a napkin.

Photographing a macro product is carried out in a limited volume enclosed on all sides, with a lockable opening for access inside, the walls of the fence made of easily washable material and have an internal reflective surface and an external light-absorbing surface, and the positioning and lighting device of the macro image preparation system contains a subject a table for installing a macro preparation, on the basis of which a means for positioning a digital camera and light sources is rigidly fixed a tripod consisting of three vertical guide-rails, which are arranged in the horizontal section in a triangle, wherein the light sources are arranged symmetrically about the optical axis of the digital camera. At the same time, the macro preparation is photographed under diffuse lighting by placing translucent shutters made of white translucent easily washable material inside the limited volume between the light sources and the object table, using halogen lamps that are located above the place of installation of the digital camera as light sources.

Additionally, when conducting a histological diagnosis of a thyroid neoplasm, data on the probability of a particular diagnosis in the presence of each of the combinations of signs present in the selected reference images are displayed on a computer monitor screen and these data are used when comparing the obtained images with the selected reference images.

When carrying out these studies, an additional system is used consisting of computer-assisted laboratory and consultation points, each of which contains a computer, and a computer connected to image-acquisition systems, respectively, of a macrodrug and a sample, is used as a laboratory computer.

Requirements for the laboratory assistant working at the laboratory:

1. Availability of access to work in the pathology laboratory, special training in labor protection, including qualification group II for electrical safety, the absence of contraindications for health reasons (in accordance with the instructions on labor protection for the staff of the pathology laboratory).

2. The ability to prepare histological preparations and work with a biological microscope.

3. Basic skills of working with a computer with the Windows operating system (turning on, turning off the power, launching applications on the desktop, switching between applications).

4. Skills of working with application software of the ATLANT complex.

5. The ability to work with the installation for photographing operational material (adjust the illumination of the operational material, select the exposure for photographing, conduct photography).

Requirements for a consultant working at a consulting point:

1. The qualifications of a pathologist.

2. Items 3, 4 (see above: 3. Basic skills of working with a computer running Windows (turning on, turning off the power, launching applications on the desktop, switching between applications).

4. Skills of working with application software of the ATLANT complex.

The laboratory room at which the automated analysis is carried out must be equipped in full compliance with the requirements for the laboratories in which the biopsy and surgical material is studied, it must be equipped with ventilation (Table 3).

Table 3

Measurement Conditions Ambient temperature from +18 to + 25 ° С Relative humidity from 60 to 80% Atmosphere pressure from 86 to 104 kPa

Between the laboratory and consultation points there is a constant two-way audio communication.

On Fig shows an example of a possible scenario for remote control cutting of operational material. After conducting a survey photograph of the thyroid neoplasm at the laboratory station, the received digital images of the thyroid neoplasm are sent via a computer communication network to the consultation center, where at least one macro-section is marked out on the digital image of the thyroid neoplasm, and an indication of receipt is sent to the laboratory assistant an enlarged digital image of each selected macro-section, and after receiving it at the laboratory e of the enlarged digital image of each macro section send it to the consultation point, while the indicated operations of marking the corresponding macro section with the marking and obtaining its enlarged digital image and transferring the relevant information from point to point are repeated until the images of all the macro sections selected for the study are received and determined at the consultation point, a place for cutting a tissue cut, after which an instruction is sent to the laboratory assistant to perform a tissue cut, and to nsultatsionnom said measuring step is performed macroportion neoplasm of thyroid gland and determine the presence or absence of these features in a given reference set of features (9). A micropreparation (tissue section) is prepared for research (Fig. 10), this process contains procedures for freezing a tumor piece in a microtome, fixing, staining in hematoxylin-eosin and imprisoning in a balm. The lab assistant as a result of these procedures receives a micropreparation for examination under a microscope. In this case, the microanalysis system is prepared for operation (Fig. 11). At this stage, the laboratory assistant turns on the power of the microscope, motorized drive and camera, loads the software of the microanalysis system. The first prerequisite for the performance of the microscopic analysis system is the absence of warnings about problems when the software is loaded.

Then carry out a remote search for information fields on a micropreparation (Fig), which includes an automatic scan of a tissue section.

Digital images of a tissue section obtained in the process of automatic scanning are simultaneously displayed on the computer monitors screens of the laboratory and consultation centers, after which the zones of the tissue section images are marked with marking for further detailing at the consulting center, and an automatic scanning and receipt are sent to the laboratory station enlarged digital images of the selected areas of the tissue section images, which are displayed simultaneously on the screen s laborantskogo consultation and computer monitors points, after which the advisory paragraph above operation is carried out.

When carrying out at least one extraction operation by marking the corresponding macro-section, the location of the thyroid neoplasm incision is additionally marked and the direction about the corresponding incision and receipt of its enlarged digital image are additionally sent to the laboratory, while the laboratory is additionally formed and sent to the consultation initial data on the patient.

At the consultation point, an information package is additionally formed containing the results of measurements of P, L, S values and a description of the thyroid neoplasm with the attached received images and a formatted protocol for placement in the database. Upon receipt of the image of the macro-section of the macro-preparation, if necessary, remote control is carried out of the macro-image acquisition system for positioning the macro-section of the macro-preparation.

Measurements, the formation of a description of the thyroid neoplasm and the execution of the protocol at the consultation point are carried out during the cutting of the slice at the laboratory station and the preparation of the slice for microscopic examination.

When determining the diagnosis, the consultant can turn to the information expert decision support system, which includes an integrated database containing color reference images of thyroid neoplasms and tissue sections of the thyroid neoplasms and their formalized descriptions of the reference set of signs, and software modules for managing the specified database ( Fig.13). The database includes color images of tumor preparations taken at various magnifications of x4, x10, x20, x40 lenses (types of lenses), standardized descriptions of them with signs marked on the images, as well as specially developed means of searching for similar cases of diseases that make it possible to take decisions to use the experience of leading histologists. In the ATLANT complex, 325 cases of tumors and tumor-like processes (over 2500 images) belonging to 15 nosological units served as material for creating a database of thyroid tumors.

At the final stage, the laboratory assistant prints out the diagnostic protocol and completes the session with the ATLANT decision support information and measurement complex.

It should be noted that in the process of work at any stage, if necessary, the laboratory assistant and consultant can turn to the built-in training system.

The proposed method for interactive microscopic analysis of thyroid preparations, mentioned above, is carried out in the software environment of the information expert decision support system (see Fig. 14 - software structure of the expert system) and includes seven main modules:

user interface (24), subsystems: updating the database (25), data analysis (26), user training (27), editing signs and nosological forms (28), a database of images and descriptions (29), a database of signs and nosology (thirty).

The objects of study in the structure of the expert system shown in Fig. 14 are color microscopic images of thyroid tumors in .bmp format, either directly entered into a computer from a camera connected to a microscope or downloaded from a file. A generalized methodology for working with an expert system is shown in Fig. 15. Three main modes are programmed in it: diagnostics, replenishment (editing) and training.

1. The replenishment (editing) mode includes two basic scenarios: dynamic editing of formalized lists of informative features and nosological forms and replenishment (a kind of training of the expert system) with new reference cases. This mode is available only for experts. Reference standards are morphologically confirmed cases of diseases described by experts. In accordance with the second scenario, an expert enters images of histological preparations from a file or from a camera, data forming a description of images in a given feature space, and also draws graphic markings corresponding to informative features on the images. Then these cases are stored in the knowledge base.

2. Training mode. It provides for three main scenarios: viewing (studying), training and knowledge control. In the training mode, it is allowed to view all the images available in the knowledge base for the selected nosological units, any formalized signs, their combinations, study the individual marked signs that are most characteristic of the selected nosology, as well as descriptions developed by leading experts in the histological diagnosis of thyroid tumors and compare them with their own. After mastering the principles of the organization of knowledge embedded in the base, it is recommended to proceed to the training scenario. This scenario includes the following steps: a description based on signs of a case of a disease randomly selected from the database, presented in a series of images, an estimated diagnosis, a comparison of your diagnosis and description with a reference. The knowledge control scenario (conditionally exam) differs from training in the function of automatically calculating the number of cases viewed and the number of correct answers.

3. Diagnostic mode. A diagnostic strategy using a decision support system can be built according to two main scenarios that can be used together or separately: an interactive electronic atlas or an expert system.

Being similar to the routine diagnostic method, the first mode has indisputable advantages: the ability to quickly search for images by signs or nosologies, the images are equipped with clarification of the diagnosis, comments, special markup that focuses on informative signs, the ability to display images and accompanying information on the printer, the function of viewing images with a given combination of features within certain nosological units and their comparison.

The second scenario (expert system) is the intellectual core of the decision-making support system under consideration. The task of the expert system is to determine the class (nosological form), to which the studied object (case of disease) most likely belongs. To do this, the user sets a combination of features in which he is most confident, and the system automatically searches for all cases with the same combination in the knowledge base. Next, you can see in percentage terms how these cases are distributed in the space of all nosological forms laid down in the database. There are also ways to narrow the diagnostic framework by tightening the search conditions by adding new features. The final decision rests with the user, and the system in this case acts as a "virtual consultation" of experts and can orient him within the diagnostic process.

Presentation of research results

The formation of standardized forms of documentation based on research results is a prerequisite for the effective automation of histological diagnosis of thyroid tumors. In the framework of diagnostics using the ATLANT decision support complex, using the specially designed computer forms, the following standardized conclusions are formed:

1. A formalized macroscopic conclusion, which contains a general description of the material indicating the part sent, the size of the material sent, the number and types of pathological processes in it, a description of each of the pathological processes. The conclusion includes color images of macro preparations with size marking, highlighted areas of interest and cut lines.

2. The protocol of telemedicine diagnostics containing patient data, clinical diagnosis, macroscopic description, histological diagnosis, the most characteristic images of a micropreparation for visual confirmation of the consultant’s conclusions, as well as service information (hospital number, name of laboratory and consultant, date, etc.) .

3. Printed forms of the conclusion of the conclusions of the expert system with the results of the search for similar cases of diseases with illustrating images and decoding of informative signs.

The proposed technical solution provides the following main results.

A method was developed for automated remote histological rapid diagnosis of tumors and tumor-like processes using the ATLANT information-measuring complex for decision-making support. The method is intended for automated remote histological rapid diagnosis of tumors in any healthcare institution in which histological diagnosis is performed (both planned and express). Technically, the complex represents a combination of two laboratory assistant and consultant workstations connected by a fiber-optic communication line, one of which, laboratory assistant's workstation, includes subsystems for inputting images of macropreparations and inputting images of micropreparations.

The implementation of the method is based on the following main stages: preparation and tuning of the macroanalysis system, remote control of the cutting of operational material, OM macroanalysis, micropreparation preparation, preparation and tuning of the microanalysis system, remote search for informative fields on the micropreparation, microanalysis using an expert information system. Stages are presented in a hierarchical structure with different levels of detail.

The results of automated diagnostics are presented in the form of standardized forms of documentation: a formalized macroscopic conclusion, a protocol for telemedicine diagnostics, printed forms for the conclusion of conclusions of an expert system.

2. A sequence of measurements of pathological areas of a macrodrug is presented in detail, which is a providing element for the implementation of an automated diagnostic method that regulates the initial data (measurement objects, measured values and their ranges), contains the formulation of the measurement problem, the requirements for measurement error, preparation procedures taking measurements.

3. The sequence of interactive microanalysis using the information subsystem of decision support for the thyroid gland, which includes three main modes: diagnosis, replenishment (editing) and training, is shown. The editing mode, which is accessible only to experts, includes two basic scenarios: dynamic editing of formalized lists of informative signs and nosological forms and replenishment with new reference cases. In the training mode, there are three main scenarios: viewing (studying), training and knowledge control. A diagnostic strategy using a decision support system can be built in two main scenarios: an interactive electronic atlas or an expert system. The expert system determines the most probable class of the input object and the coefficient of confidence of the result, as well as builds the most likely diagnostic solutions for the doctor with the ability to view and visually compare images of the input object and from the database.

Measured physical quantities, as well as their ranges (table. 1):

Table 1 Measured physical quantity Designation Units Measured value range one Distance between reference points L Cm 0.5-10 2 The perimeter of the selected object R Cm 3-30 3 Area of the selected object S Cm ² 0.8-78 four Shape factor TO dimensionless 1-2 * Reference points - points characterizing, according to the expert, the boundaries of the informative area. table 2 Measured physical quantity System error one Distance between reference points ± 3% 2 The perimeter of the selected object ± 3% 3 Area of the selected object ± 3% four Shape factor ± 5%

Table 3 Measurement Conditions Ambient temperature from +18 to + 25 ° С Relative humidity from 60 to 80% Atmosphere pressure from 86 to 104 kPa

Claims (18)

1. The method of microscopic examination of a sample, including preliminary obtaining the sample, while a) conduct a macroscopic study of a macro product intended to obtain a sample using a macro image preparation system including a digital camera and a positioning and lighting device, and a computer connected to it, during which b) conduct a survey photograph of the macro product, as a result of which digital images of the macro product are obtained, C) carry out the allocation of the marking of the macro-section, intended to obtain a sample on a digital image of the macro, and g) receive an enlarged digital image of the selected macro-stage, while the indicated operations c) and d) are repeated until the place of cutting the sample is determined, and d) cut out the sample from the macro, e) carry out the preparation of the obtained sample for microscopic examination, g) carry out a microscopic examination of the sample using a computer-imaging system for acquiring sample images containing a microscope, a digital camera, a matching module for joining a digital camera and a microscope, during which h) position the sample on the microscope stage, i) produce an initial review view, for which a 4x lens is installed and focus is carried out, j) conduct an automatic scan, as a result of which digital images of the sample are obtained, which are displayed on a computer screen, k) distinguish by marking the zone of the image of the sample for further detail, m) carry out the change of the microscope lens to a lens with high magnification and its focusing, m) conduct automatic scanning of the selected areas of the image of the sample, resulting in an enlarged digital image of the selected areas of the image of the sample, which are displayed on the computer screen, o) describe the obtained images using formalized features from a given set of formalized features, p) based on the combination of formalized features obtained, from the set of reference images contained in the base of reference images, those reference images that have similar parameter values and similar combination of signs are selected, after which c) comparing the obtained images of the selected areas of the image of the sample with the selected reference images, r) the results of which judge the state of the sample. 2. The method according to claim 1, characterized in that the automatic scanning during microscopic examination is carried out by means of predetermined coordinate displacements with restriction of the extreme positions of the subject table by displacement drives connected to the computer, and the disks of the movement of the subject table are made in the form of a motor gear unit, consisting stepper motors with gears, and for this, a MIKMED-2 microscope and a coordinate coordinate table KS-30 are used for research. 3. The method according to claim 1, characterized in that the positioning and lighting device of the macro-image acquisition system comprises an object table for installing the macro-preparation, on which markings are made with a step of 1 cm in two mutually perpendicular directions, and from the rear edge and sides on the subject fencing is installed on the table, and before being placed on the object table, the macro preparation is blotted with a napkin. 4. The method according to claim 1, characterized in that the photographing of the macro product is made in a limited volume enclosed on all sides, having a lockable opening for access inside, and the walls of the fence are made of easily washable material and have an internal reflective surface and an external light-absorbing surface, wherein the positioning and lighting device of the macro image preparation system comprises an object table for installing the macro preparation, on the basis of which the positioning means is fixed digital camera and light sources in the form of a tripod, consisting of three vertical guide rails, which in horizontal section are located at the vertices of the triangle, while the light sources are mounted symmetrically relative to the optical axis of the digital camera. 5. The method according to claim 4, characterized in that the macro photography is photographed under diffuse lighting by placing translucent dampers made of white translucent easily washable material between the light sources and the object table, using halogen lamps as light sources, which are located above the installation location of the digital camera. 6. The method according to any one of claims 1 to 5, characterized in that as a macro preparation using surgical material containing a thyroid neoplasm, the isolated macro-sections are a thyroid neoplasm and / or its characteristic sections, for example, a tumor node or a node capsule, in this case, a histological section of the tissue of the thyroid neoplasm stained with hematoxynline and eosin or picrofuxin according to Van Gieson is used as a sample. 7. The method according to claim 6, characterized in that they additionally measure the macro-sites of the thyroid neoplasm, and the following parameters are measured: linear dimensions along the major and minor axes of inertia L, perimeter P, area S, shape coefficient K, while before of measurements determine a scale factor to bring the measurement results to metric units of length and area

, where L = 1 cm, l is the same distance in pixels of the image, and the shape factor is calculated by the formula

. 8. The method according to claim 7, characterized in that the results of macroscopic and microscopic examinations are used for histological diagnosis of thyroid neoplasms, for which purpose data on the probability of a particular diagnosis in the presence of each of the combinations of signs available in the selected reference images, and use this data when comparing the resulting images with the selected reference images and establishing a diagnosis. 9. The method according to claim 8, characterized in that they additionally use a system consisting of computer-assisted laboratory and consultation points, each of which contains a computer, the computer connected to the image acquisition systems, respectively, of the macrodrug and the sample, used in the quality of the computer at the laboratory, and after conducting a survey of the thyroid tumor at the laboratory, the digital images of the thyroid neoplasm obtained Threats are sent via a computer communication network to a consultation center, where they mark out a macro-section intended for receiving a tissue section on a digital image of a thyroid neoplasm with a marking, and an indication to receive an enlarged digital image of this macro-section is sent to a laboratory, and after receipt at a laboratory the enlarged digital image of the macro section is sent to a consultation point, while the indicated operations of marking according to repeating macro-section and obtaining its enlarged digital image is repeated until the place of tissue cutting is determined at the consultation point, after which an instruction is sent to the laboratory assistant to perform tissue cutting, while digital images of the tissue cutting obtained during automatic scanning simultaneously depict on the screens of computer monitors of the laboratory and consultation points, after which at the consultation point they mark out zones tissue section for their further detailing, and an indication is sent to the laboratory station for automatic scanning and obtaining enlarged digital images of the selected tissue section images, which are displayed simultaneously on the computer monitor screens of the laboratory and consultation points, after which operations are performed at the consultation point -t), as well as histological diagnosis of neoplasm of the thyroid gland. 10. The method according to claim 9, characterized in that when carrying out at least one extraction operation by marking the corresponding macro-section, the location of the incision of the thyroid neoplasm is additionally marked and the direction about the corresponding incision and its enlarged digital image are sent to the laboratory assistant. 11. The method according to claim 9, characterized in that in addition between the laboratory and consultation points is a constant two-way audio communication. 12. The method according to claim 9, characterized in that the laboratory data additionally generated and sent to the consultation point the initial data on the patient. 13. The method according to claim 9, characterized in that the foot pedals are used to control the acquisition of images and the transfer of images to a computer at the laboratory station. 14. The method according to claim 9, characterized in that an additional information packet is formed at the consultation point containing the measurement results of the P, L, S values and a description of the thyroid neoplasm with the attached received images and a formatted protocol for placement in the database. 15. The method according to 14, characterized in that the measurement, the formation of a description of the thyroid neoplasm and the execution of the protocol at the consultation point is carried out during the cutting laboratory tissue cutting section and its preparation for microscopic examination. 16. The method according to claim 9, characterized in that, at the consultation point, upon receipt of the macro-image of the macro-preparation, if necessary, remote control is performed of the macro-image acquisition system for positioning the macro-preparation of the macro-section. 17. The method according to claim 9, characterized in that, when determining the diagnosis, an information expert decision support system is further initiated, including a database collection containing color reference images of thyroid neoplasms and tissue sections of the thyroid neoplasms and their formalized descriptions according to the reference set of signs , and software modules for managing the specified database, with the help of which a detailed description of the histological picture of the digital images is carried out i, which is compared with the existing reference combinations of the values of the characters for the final histological diagnosis. 18. The method according to claim 9, characterized in that it further initiates an integrated learning system, made in the form of an electronic textbook containing a set of diagnostic data in the form of texts, photographs, video clips and / or computer graphics.

Images