Sample freezing and slicing system and slicing method for pathology department

Abstract

The invention discloses a pathology department sample freezing and slicing system, and relates to the technical field of sample slicing, and the pathology department sample freezing and slicing system comprises a central control module, a liquid nitrogen injection module, a freezing table, a sensor module, a slicing control module, an image monitoring module, a sample limiting control module and a slicing collection module. The whole freezing and slicing system realizes high automation and intelligence, reduces time and errors of manual operation, and realizes rapid sample freezing and accurate slicing operation. The invention also discloses a pathology department sample freezing and slicing method. The method comprises the following steps: step 1, pre-treating a sample; 2, freezing the sample; step 3, slicing operation; 4, slice quality judgment and parameter adjustment; and 5, collecting and storing the slices, and ensuring the stability and accuracy of the sample in the freezing and slicing processes and the accuracy of the slice positions by adopting a composite fixing mode of combining vacuum adsorption and magnetic fixation and a high-precision laser positioning system.

Description

Technical Field

[0001] The present invention relates to the technical field of sample sectioning, and specifically to a pathological sample cryosectioning system and a sectioning method. Background Technique

[0002] Cryosectioning is a method of quickly cooling tissue to a certain hardness under low-temperature conditions and then sectioning it. Because its production process is faster and simpler than paraffin sectioning, it is widely used in rapid pathological diagnosis during surgery. There are many types of cryosections, such as cryosectioning method with a low-temperature constant-temperature freezer and carbon dioxide cryosectioning method. Cryosections can quickly help doctors diagnose the pathology of patients. However, when performing cryosections, the pathological tissue needs to be frozen first. The existing freezing method is to freeze the pathological tissue through a freezing constant-temperature box. Such a freezing method involves operations of placing and taking, and it is easy to cause contamination of the frozen pathological tissue.

[0003] The invention patent with the publication number CN115655776A discloses a pathological sample cryosectioning system and a sectioning method, including a base. A feeding mechanism is arranged on the upper surface at one end of the base. Among them, a pressing mechanism is arranged inside the feeding mechanism. Among them, a sliding mechanism is arranged on the upper surface at the other end of the base. Among them, a cooling mechanism is arranged at the upper end of the sliding mechanism. Among them, a sectioning mechanism is arranged inside the cooling mechanism.

[0004] The existing cryosectioning system and sectioning method generally drive a runner to rotate when a rotating shaft rotates, drive a notch to rotate when the runner rotates, drive a U-shaped rod to rotate with the notch, drive a rotating plate to rotate with the U-shaped rod, and drive a cutting knife to rotate counterclockwise around the rotating shaft when the rotating plate rotates, so as to cut the pathological tissue with the cutting knife. When the runner rotates one circle, the cut pathological section adheres to the glass slide. Then, the threaded sleeve is removed from the fixed card slot, the rotating plate is rotated to the lower end of the notch, and then the glass slide is pushed out of the clamping groove, thus quickly completing the cutting of the section. This method focuses on quickly sectioning the sample, but the fast sectioning means that it is impossible to judge whether there are quality problems such as uneven thickness, cracks, and breakages in the section during the sectioning process. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention provides a pathological sample cryosectioning system and a sectioning method, which solve the existing problems.

[0006] To achieve the above objectives, the present invention is realized through the following technical solutions: A pathological sample cryosectioning system, including:

[0007] A central control module, which uses a high-performance industrial computer and is built-in with a PID control algorithm program;

[0008] Liquid nitrogen injection module, which adopts a high-precision liquid nitrogen injection valve and is connected to a liquid nitrogen storage tank through a high-pressure stainless steel pipe

[0009] Freezing stage, made of aluminum alloy, with a microchannel structure machined inside. Liquid nitrogen circulates in the microchannels to achieve efficient heat exchange. Its surface is treated with a special anti-slip coating, and the coating is made of polytetrafluoroethylene material;

[0010] Sensor module, including a temperature sensor, a pressure sensor, a flow sensor, an image sensor, a vacuum gauge, and a laser range finder sensor;

[0011] Section control module, used to control two high-torque stepper motors on linear guides. The two linear guides respectively control the movement of the cutting knife in the horizontal and vertical directions;

[0012] Image monitoring module, which is connected to the central control system through a dedicated image acquisition card and is equipped with an image analysis model based on deep learning algorithms. It is trained according to a large number of image data sets containing sections of different qualities and can accurately identify quality problems such as the thickness uniformity, cracks, and breakages of the sections;

[0013] Sample limit control module, which adsorbs and limits the sample by controlling a vacuum adsorption device and further limits the sample by a magnetic fixing device;

[0014] Section collection module, including a section collection box with a three-layer buffer structure inside, used to collect the samples after sectioning.

[0015] Preferably, the temperature sensor adopts a K-type thermocouple temperature sensor, with a measurement range of -200°C to 1200°C, a measurement accuracy of up to ±0.1°C, a response time of less than 5ms, and is installed inside the freezing stage and on the surface of the sample through a high-temperature-resistant ceramic sleeve to ensure accurate measurement of the temperature of the sample and the freezing stage. The central control module adjusts the opening of the liquid nitrogen injection valve according to the data fed back by the temperature sensor and the preset freezing curve to achieve precise control of the freezing rate of the sample.

[0016] Preferably, the pressure sensor and the flow sensor are installed on the high-pressure stainless steel pipe to monitor the pressure and flow of liquid nitrogen in real time and transmit the data to the central control module in real time.

[0017] Preferably, the image sensor adopts a high-resolution industrial camera, model MV-CE050-10GC, with a frame rate of 60fps and a resolution of 2560×1920 pixels, and is signal-connected to the image acquisition card on the image monitoring module.

[0018] Preferably, the vacuum gauge is installed on the vacuum adsorption device to monitor the vacuum degree in real time and transmit the data to the central control module.

[0019] Preferably, the step angle of the stepper motor is 0.005°, which is connected to a precision lead screw through a coupling. The lead of the lead screw is 0.5 mm, and the accuracy can reach ±0.01 mm. The linear guide uses a high-precision ball guide, and its straightness error is less than ±0.005 mm / m. The slicing knife uses a high-hardness alloy blade.

[0020] Preferably, the model of the laser distance sensor is LDM4X, with a measurement accuracy of up to ±0.05 mm and a measurement range of 0 - 500 mm. It is installed on a movable bracket, and the bracket is driven by a motor to scan on the freezing table, and the measurement data is transmitted to the central control system. The central control system constructs a three-dimensional model of the sample based on these data, accurately determines the position and contour of the sample, and provides accurate positioning information for the slicing operation.

[0021] Preferably, the first layer of the slicing collection box is a sponge cushion layer with a thickness of 5 mm, which is used to buffer the impact force when the slices fall; the second layer is a silica gel cushion layer with a thickness of 3 mm, which further reduces the collision between the slices and the collection box; the third layer is an anti-static coating, which prevents static electricity adsorption during the collection of slices and affects the integrity of the slices.

[0022] The present invention also discloses a method for cryosection of pathological specimens in the pathology department, including the following steps:

[0023] Step 1: Sample pretreatment. Put the pathological tissue sample into the sample preservation solution, take it out before cryosection, rinse it with normal saline and dry it, place it on the freezing table of the sample fixing and positioning module, and fix it through a vacuum adsorption and magnetic fixing device.

[0024] Step 2: Sample freezing. Preset a freezing curve according to the sample characteristics, start the sample freezing module, accurately control the freezing rate through a flow control valve and a temperature sensor, monitor the temperature change in real time, and dynamically adjust the liquid nitrogen injection amount by using a PID control algorithm.

[0025] Step 3: Slicing operation. After the sample freezing is completed, the laser positioning system obtains the position and contour information of the sample, and the central control system controls the slicing module to slice. The intelligent image monitoring system collects the slice images in real time and evaluates the slice quality through a deep learning algorithm.

[0026] Step 4: Judgment of slice quality and parameter adjustment. If the deep learning algorithm detects abnormal slice quality, the system issues an alarm and pauses slicing, automatically adjusts the slicing parameters according to the evaluation results, such as slicing speed, cutting angle of the knife, etc., and continues slicing after passing the inspection.

[0027] Step 5: Slice Collection and Preservation. After slicing, the slices automatically slide down into the collection box. After the collection box is full, it is transferred to the storage box of the low-temperature drying and preservation system for preservation.

[0028] Preferably, during the training process of the deep learning algorithm, a large number of image datasets containing normal slices and various types of slices with quality problems are used to adjust the parameters of the convolutional neural network through the backpropagation algorithm to minimize the cross-entropy loss function between the prediction result and the true label.

[0029] Beneficial Effects

[0030] The present invention provides a cryostat sectioning system and a sectioning method for pathological specimens. Compared with the prior art, it has the following beneficial effects:

[0031] 1. For the cryostat sectioning system and sectioning method for pathological specimens, the entire cryostat sectioning system realizes high automation and intelligence, reduces the time and error of manual operation, enables rapid sample freezing and precise sectioning operation, greatly shortens the sample processing time, can meet the urgent needs of rapid pathological diagnosis during surgery, provides strong support for doctors to formulate surgical plans in a timely manner, improves the efficiency of medical work. At the same time, a composite fixation method combining vacuum adsorption and magnetic fixation, as well as a high-precision laser positioning system, ensure the stability and accuracy of the sample during freezing and sectioning, avoid sample displacement and damage, guarantee the accuracy of the section position, and improve the usability and diagnostic value of the sections.

[0032] 2. For the cryostat sectioning system and sectioning method for pathological specimens, by precisely controlling the sample freezing rate, it effectively avoids the damage of ice crystals to the tissue structure of the sample, ensures the integrity of the sections and the clarity of cell morphology. The high-precision sectioning mechanism and the intelligent image monitoring system based on the deep learning algorithm can accurately identify and correct section quality problems, ensure the uniformity and stability of the section thickness, significantly improve the imaging effect of the sections under the microscope, provide clearer and more accurate pathological information for pathologists, and help improve the accuracy of pathological diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a schematic diagram of the system connection of the present invention;

[0034] Figure 2 is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0035] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0036] Referring to Figure 1-2 , the present invention provides two technical solutions:

[0037] The first implementation mode: A pathological sample cryostat system, comprising:

[0038] A central control module, which uses a high-performance industrial computer and is built-in with a PID control algorithm program;

[0039] A liquid nitrogen injection module, which uses a high-precision liquid nitrogen injection valve and is connected to a liquid nitrogen storage tank through a high-pressure stainless steel pipe. The pressure sensor and the flow sensor are installed on the high-pressure stainless steel pipe to monitor the pressure and flow rate of the liquid nitrogen in real time and transmit the data to the central control module in real time. The measurement accuracy of the pressure sensor is ±0.05 MPa, and the measurement accuracy of the flow sensor is ±0.03 L / min.

[0040] A freezing table, which is made of aluminum alloy and has a micro-channel structure processed inside. Liquid nitrogen circulates in the micro-channels to achieve efficient heat exchange. Its surface is treated with a special anti-slip coating, and the coating uses polytetrafluoroethylene material. The size of the freezing table can be customized according to actual needs, generally 200 mm in length × 150 mm in width × 50 mm in height, which can meet the freezing requirements of different sizes of samples.

[0041] A sensor module, including a temperature sensor, a pressure sensor, a flow sensor, an image sensor, a vacuum gauge, and a laser ranging sensor;

[0042] The temperature sensor uses a K-type thermocouple temperature sensor, with a measurement range of -200°C to 1200°C, a measurement accuracy of up to ±0.1°C, a response time of less than 5 ms, and is installed inside the freezing table and on the sample surface through a high-temperature resistant ceramic sleeve to ensure accurate measurement of the temperature of the sample and the freezing table. The central control module adjusts the opening of the liquid nitrogen injection valve through a control signal according to the data fed back by the temperature sensor and the preset freezing curve to achieve precise control of the freezing rate of the sample.

[0043] The laser ranging sensor is model LDM4X, with a measurement accuracy of up to ±0.05 mm, a measurement range of 0 - 500 mm. It is installed on a movable bracket, and the bracket is driven by a motor to scan on the freezing table, and the measurement data is transmitted to the central control system. The central control system constructs a three-dimensional model of the sample based on these data, accurately determines the position and contour of the sample, and provides accurate positioning information for the slicing operation.

[0044] The slicing control module is used to control two high-torque stepper motors on linear guides. The two linear guides respectively control the movement of the cutting knife in the horizontal and vertical directions. The step angle of the stepper motor is 0.005°, and it is connected to a precision lead screw through a coupling. The lead of the lead screw is 0.5 mm, and the accuracy can reach ±0.01 mm. The linear guides use high-precision ball guides, and their straightness error is less than ±0.005 mm / m. The slicing knife uses a high-hardness alloy blade.

[0045] The image monitoring module is connected to the central control system through a dedicated image acquisition card. It is equipped with an image analysis model based on deep learning algorithms, which is used to train according to a large number of image data sets containing slices of different qualities, and can accurately identify quality problems such as the thickness uniformity, cracks, and breakages of the slices. The image sensor uses a high-resolution industrial camera, model MV-CE050-10GC, with a frame rate of 60 fps and a resolution of 2560×1920 pixels, and it is signal-connected to the image acquisition card on the image monitoring module.

[0046] During the training process, the parameters of the convolutional neural network (such as convolutional kernel weights, fully connected layer weights, etc.) are continuously adjusted through the backpropagation algorithm to minimize the loss function between the prediction result and the true label. After sufficient training, the model can accurately identify quality problems such as the thickness uniformity, cracks, and breakages of the slices. Once a quality problem is detected, the system immediately issues an alarm and sends an adjustment instruction to the central control system to automatically adjust the slicing parameters.

[0047] The sample limit control module adsorbs and limits the sample by controlling a vacuum adsorption device. The vacuum gauge is installed on the vacuum adsorption device to monitor the vacuum degree in real time and transmit the data to the central control module. The vacuum adsorption device uses a rotary vane vacuum pump with an air extraction rate of 15 L / min and an ultimate vacuum degree of up to -0.1 MPa. The vacuum pump is connected to the vacuum adsorption holes on the freezing table through a vacuum-resistant rubber pipe. The adsorption holes are evenly distributed on the surface of the freezing table, with a hole diameter of 2 mm. During the vacuum adsorption process, the vacuum degree is monitored in real time through the vacuum gauge to ensure the stability of the vacuum adsorption.

[0048] The sample is further limited by a magnetic fixing device, which consists of 8 electromagnets with adjustable magnetic strength and is distributed around the freezing table. The magnetic strength of the electromagnets can be adjusted within the range of 0 - 500 mT, and the precise control of the magnetic strength is achieved by adjusting the current magnitude. The control circuit of the electromagnets is connected to the central control system. According to the shape and size of the sample, the central control system automatically adjusts the magnetic strength and distribution of the electromagnets to ensure the firm fixation of the sample.

[0049] The section collection module includes a section collection box with a three - layer buffer structure inside, which is used to collect the samples after sectioning. The first layer of the section collection box is a sponge cushion layer with a thickness of 5 mm, which is used to buffer the impact force when the sections fall; the second layer is a silica gel buffer pad with a thickness of 3 mm to further reduce the collision between the sections and the collection box; the third layer is an anti - static coating to prevent static adsorption during the collection of sections and affect the integrity of the sections.

[0050] The second implementation method: A method for cryosection of pathological specimens in the pathology department, including the following steps:

[0051] Step 1: Sample pretreatment. Put the pathological tissue sample into the sample preservation solution, take it out before cryosection, rinse it with normal saline and dry it, place it on the freezing table of the sample fixation and positioning module, and fix it through vacuum adsorption and the magnetic fixing device.

[0052] Step 2: Sample freezing. According to the sample type, size and tissue characteristics, preset a freezing curve in the central control system. Start the sample freezing module. The high - precision flow control valve precisely adjusts the liquid nitrogen injection flow according to the preset curve. The temperature sensor monitors the sample temperature in real - time. The central control system uses the PID control algorithm to dynamically adjust the opening of the flow control valve to ensure that the sample is evenly cooled to the required temperature at the preset freezing rate.

[0053] Step 3: Sectioning operation. After the sample freezing is completed, the laser positioning system scans to obtain the sample position and contour information and transmits it to the central control system. Combining the preset section thickness and path, control the stepping motor and the sectioning knife of the sectioning module to perform sectioning. During the sectioning process, the intelligent image monitoring system collects section images in real - time and inputs them into the trained deep - learning model for analysis and evaluation.

[0054] Step 4: Slice Quality Judgment and Parameter Adjustment. The deep learning model judges whether there are quality problems such as uneven thickness, cracks, and breakages in the slices based on the image analysis results. Once quality anomalies are detected, the system immediately issues an alarm and automatically pauses the slicing operation. According to the quality assessment results output by the model and combined with the preset parameter adjustment strategy, the slicing parameters are automatically adjusted, such as the slicing speed, the cutting angle of the knife, etc. For example, if the model detects uneven slice thickness, the system will appropriately reduce the slicing speed and finely adjust the cutting angle of the knife. After ensuring that the quality of subsequent slices meets the requirements, the slicing operation is continued.

[0055] Step 5: Slice Collection and Preservation. After slicing is completed, the slices automatically slide down into the collection box. After the collection box is full, it is transferred to the storage box of the low-temperature drying preservation system for preservation. The temperature in the box is precisely adjusted between -20°C and -80°C, and the fluctuation is controlled within ±1°C. It is equipped with an efficient desiccant and an ultraviolet sterilization device that is started regularly to ensure that the slices are not contaminated and damaged during preservation.

[0056] In this embodiment, the formula of the PID control algorithm is:

[0057]

[0058] where u(t) is the control quantity at time t, that is, the opening adjustment signal of the flow control valve; K p is the proportional coefficient, which determines the response intensity of the controller to the current temperature deviation; e(t) is the temperature deviation at time t, that is, the difference between the preset temperature and the actual measured temperature; K i is the integral coefficient, which is used to eliminate the steady-state error of the system; K d is the differential coefficient, which predicts the temperature change trend according to the change rate of the temperature deviation, adjusts the control quantity in advance, and enhances the stability of the system.

[0059] During the freezing process, the temperature of the sample is continuously monitored to ensure the stability and accuracy of the freezing process. By adjusting the values of K p , K i , K d , the freezing control effect for different samples can be optimized. For example, for samples that are more sensitive to temperature changes, the value of K d can be appropriately increased to more quickly respond to the change of temperature deviation and prevent excessive temperature fluctuations from damaging the samples.

[0060] In actual debugging, the optimal value ranges of K p , K i , K d for different types of samples are determined through multiple experiments. For example, for liver tissue samples, through experiments, it is determined that the value of K p is between 2 - 3, and the value of K iThe value ranges from 0.1 to 0.3, K d When the value ranges from 0.5 to 1, it can achieve an ideal freezing rate control and effectively avoid the formation of ice crystals.

[0061] In this embodiment, a deep learning model based on a convolutional neural network (CNN) is constructed before slicing. The basic structure of the model includes multiple convolutional layers, pooling layers, and fully connected layers. In the convolutional layer, convolution operations are performed on the input image using convolutional kernels of different sizes (such as 3×3, 5×5, etc.) to extract image features. Let the input image be, the convolutional kernel be, and the output feature map be. Then the convolution operation formula is:

[0062] O(i, j) = ∑ m，n I(i + m, j + n) × K(m, n);

[0063] Among them, (i, j) is the pixel position in the output feature map, and (m, n) is the pixel position in the convolutional kernel. By stacking multiple convolutional layers, features at different levels of the image can be extracted, from simple edges, textures to complex structural features.

[0064] The pooling layer is mainly used to downsample the feature map, reducing the amount of data while retaining important features. Common pooling methods include max pooling and average pooling. Taking max pooling as an example, assuming the pooling window size is 2×2, then each element in the output feature map is the maximum value within the pooling window, and the formula is expressed as:

[0065] O(i, j) = max m，n I(2i + m, 2j + n);

[0066] Among them, the value range of (m, n) is (0, 0), (0, 1), (1, 0), (1, 1).

[0067] After passing through multiple convolutional layers and pooling layers, the obtained feature map is flattened and input into the fully connected layer. The fully connected layer performs a linear transformation on the features through the weight matrix W and the bias b, and then undergoes a non-linear transformation through an activation function (such as the ReLU function: f(x) = max(0, x)), and finally outputs the evaluation result of the slice quality, such as whether the slice thickness is uniform and whether there are cracks. The calculation process of the fully connected layer can be expressed as:

[0068] y = f(Wx + b);

[0069] Among them, x is the input feature vector, and y is the output evaluation result.

[0070] Train the model on a large image dataset containing slices of different qualities, and continuously adjust the model parameters (such as convolutional kernel weights, fully connected layer weights, etc.) through the backpropagation algorithm to minimize the loss function between the prediction result and the true label (such as the cross-entropy loss function: where y i is the true label, is the predicted probability), so that the model can accurately identify slice quality problems

[0071] Meanwhile, the content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0072] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0073] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A frozen section system and sectioning method for pathological specimens, characterized in that, Including: A central control module, which uses a high-performance industrial computer and has a PID control algorithm program built in; A liquid nitrogen injection module, which uses a high-precision liquid nitrogen injection valve and is connected to a liquid nitrogen storage tank through a high-pressure stainless steel pipe A freezing stage, which is made of aluminum alloy and has a microchannel structure processed inside. Liquid nitrogen circulates in the microchannels to achieve efficient heat exchange. Its surface is treated with a special anti-slip coating, and the coating uses polytetrafluoroethylene material; A sensor module, including a temperature sensor, a pressure sensor, a flow sensor, an image sensor, a vacuum gauge, and a laser range finder sensor; A slicing control module, which is used to control two high-torque stepper motors on linear guides. The two linear guides respectively control the movement of the cutting knife in the horizontal and vertical directions; An image monitoring module, which is connected to the central control system through a dedicated image acquisition card and has an image analysis model based on a deep learning algorithm. It is used to be trained according to a large number of image data sets containing slices of different qualities and can accurately identify quality problems such as the thickness uniformity, cracks, and breakages of the slices; A sample limit control module, which adsorbs and limits the sample by controlling a vacuum adsorption device and further limits the sample by a magnetic fixing device; A slice collection module, including a slice collection box with a three-layer buffer structure inside, which is used to collect the samples after slicing.

2. The cryostat sectioning system for pathological samples according to claim 1, wherein: The temperature sensor uses a K-type thermocouple temperature sensor, whose measurement range is -200°C to 1200°C, the measurement accuracy reaches ±0.1°C, the response time is less than 5ms, and it is installed inside the freezing stage and on the surface of the sample through a high-temperature-resistant ceramic sleeve to ensure that the temperature of the sample and the freezing stage can be accurately measured. The central control module adjusts the opening of the liquid nitrogen injection valve through a control signal according to the data fed back by the temperature sensor and the preset freezing curve to achieve precise control of the freezing rate of the sample.

3. The cryostat sectioning system for pathological samples according to claim 1, characterized in that: The pressure sensor and the flow sensor are installed on the high-pressure stainless steel pipe to monitor the pressure and flow of liquid nitrogen in real time and transmit the data to the central control module in real time.

4. A pathological sample cryosection system according to claim 1, characterized in that: The image sensor uses a high-resolution industrial camera, model MV-CE050-10GC, with a frame rate of 60fps and a resolution of 2560×1920 pixels. It is signal-connected to the image acquisition card on the image monitoring module.

5. A pathological sample frozen section system according to claim 1, characterized in that: The vacuum gauge is installed on the vacuum adsorption device to monitor the vacuum degree in real time and transmit the data to the central control module.

6. The cryostat sectioning system for pathological samples according to claim 1, wherein: The step angle of the stepper motor is 0.005°, and it is connected to a precision lead screw through a coupling. The lead of the lead screw is 0.5mm, and the accuracy reaches ±0.01mm. The linear guide uses a high-precision ball guide, and the straightness error is less than ±0.005mm / m. The slicing knife uses a high-hardness alloy blade.

7. A pathological sample frozen section system according to claim 1, characterized in that: The laser range finder sensor model is LDM4X, with a measurement accuracy of ±0.05mm and a measurement range of 0 - 500mm. It is installed on a movable bracket and scans on the freezing stage by driving the bracket with a motor, and transmits the measurement data to the central control system. The central control system constructs a three-dimensional model of the sample based on these data, accurately determines the position and contour of the sample, and provides accurate positioning information for the slicing operation.

8. The cryostat sectioning system for pathological specimens according to claim 1, wherein: The first layer of the slice collection box is a sponge cushion layer with a thickness of 5 mm, which is used to buffer the impact force when the slices fall; the second layer is a silica gel cushion layer with a thickness of 3 mm, which further reduces the collision between the slices and the collection box; the third layer is an anti-static coating, which prevents static electricity adsorption during the collection of slices and affects the integrity of the slices.

9. A method for frozen section of pathological samples, based on the pathological sample frozen section system according to any one of claims 1-8, characterized in that, It includes the following steps: Step 1: Sample pretreatment. Put the pathological tissue sample into the sample preservation solution, take it out before cryosection, rinse it with normal saline and dry it, place it on the freezing table of the sample fixation and positioning module, and fix it through the vacuum adsorption and magnetic fixation device. Step 2: Sample freezing. Preset the freezing curve according to the sample characteristics, start the sample freezing module, precisely control the freezing rate through the flow control valve and temperature sensor, monitor the temperature change in real time, and dynamically adjust the liquid nitrogen injection volume using the PID control algorithm. Step 3: Sectioning operation. After the sample freezing is completed, the laser positioning system obtains the sample position and contour information, the central control system controls the sectioning module to section, and the intelligent image monitoring system collects the section images in real time and evaluates the section quality through the deep learning algorithm. Step 4: Section quality judgment and parameter adjustment. If the deep learning algorithm detects abnormal section quality, the system issues an alarm and pauses sectioning, automatically adjusts the section parameters according to the evaluation results, such as sectioning speed, cutting angle of the knife, etc., and continues sectioning after passing the inspection. Step 5: Section collection and preservation. After sectioning, the sections automatically slide into the collection box. After the collection box is full, it is transferred to the preservation box of the low-temperature drying preservation system for preservation.

10. A method for cryosection of pathological specimens according to claim 9, characterized in that: In the training process of the deep learning algorithm, a large number of image data sets containing normal sections and sections with various quality problems are used to adjust the parameters of the convolutional neural network through the backpropagation algorithm to minimize the cross-entropy loss function between the prediction result and the true label.

Images