Cavity tissue biopsy guiding device and method based on optical fiber imaging and state discrimination

Abstract

The application discloses a cavity tissue biopsy guiding device and method based on optical fiber imaging and state identification, and belongs to the technical field of optical fiber imaging, medical image analysis and endoscope biopsy guiding. The device comprises a light source module, a modulation module, an optical fiber probe, a biopsy forceps, a calibration module, an imaging module, a dynamic calibration module and a computer. The optical fiber probe is inserted into the biopsy forceps channel and extends out of the forceps opening area to scan and image the target tissue; the calibration module measures the transmission matrix, the dynamic calibration module performs dynamic calibration based on the reflection signal, and the computer completes image reconstruction, tissue state identification and sampling prompt generation. According to the spatial correlation between the imaging area and the biopsy area, the method outputs sampling prompt information to assist in target area biopsy sampling, and improves the pertinence, accuracy and reliability of biopsy positioning.

Description

Technical Field

[0001] This invention belongs to the technical field of fiber optic imaging, medical image analysis and endoscopic biopsy guidance, specifically relating to a cavity tissue biopsy guidance device and method based on fiber optic imaging and state discrimination. Background Technology

[0002] Clinical diagnosis of lesions in cavities typically relies on endoscopic observation and tissue biopsy. For cavities in the digestive, respiratory, and urinary tracts, and other cavities accessible endoscopically, pathological examination remains a crucial basis for lesion characterization and grading, and obtaining representative tissue samples is a prerequisite for pathological examination. For superficial cavities, endoscopic biopsy forceps sampling has become a routine diagnostic sampling method due to its minimal invasiveness and proven technique.

[0003] Currently, biopsy forceps are mostly used for sampling under the guidance of white light endoscopy. The sampling location mainly relies on macroscopic visual observation and the operator's experience and judgment. White light endoscopy mainly provides macroscopic morphological information of the tissue surface, making it difficult to identify early microscopic lesions, occult structural abnormalities, and lesion boundary areas at the microscopic scale before sampling. It also cannot directly reflect the fine structural features of the tissue and potential lesion risks. Therefore, in practical applications, it is often necessary to combine multi-point sampling methods to improve the detection probability.

[0004] In the prior art, there are already solutions that structurally integrate imaging probes with biopsy forceps. For example, invention patent CN113425339B discloses a flexible biopsy forceps with a hollow forceps channel, which accommodates an imaging probe by setting a channel inside the forceps body to achieve imaging-assisted biopsy operations. This type of solution mainly involves the structural cooperation between the imaging probe and the biopsy forceps; it does not explicitly describe the use of image feature-based tissue state discrimination or the use of the spatial correspondence between the fiber optic probe imaging area and the biopsy area for sampling guidance.

[0005] Therefore, for cavity tissue biopsy scenarios, it is still necessary to provide a biopsy guidance technology solution that can combine microscopic structural information, tissue state discrimination results, and the spatial correlation between the fiber optic probe imaging area and the biopsy area to improve the targeting and effectiveness of sampling and positioning. Summary of the Invention

[0006] To address the problems of existing cavity tissue biopsies, which mainly rely on macroscopic observation with white light endoscopy and operator experience, making it difficult to obtain microscopic structural information of the target area before sampling, and lacking real-time tissue condition discrimination and sampling guidance mechanisms, this invention provides a cavity tissue biopsy guidance device and method based on fiber optic imaging and condition discrimination.

[0007] This invention combines in-situ fiber optic imaging, tissue state discrimination, and spatial correlation guidance between the fiber optic probe imaging area and the biopsy area to form an integrated biopsy guidance mechanism that combines imaging, discrimination, and sampling, thereby improving the targeting and effectiveness of biopsy localization.

[0008] To achieve the above-mentioned objectives, an embodiment provides a cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination, comprising a light source module, a modulation module, a fiber optic probe, biopsy forceps, a calibration module, an imaging module, a dynamic calibration module, and a computer; The modulation module is used to modulate the incident beam provided by the light source module and couple the modulated beam into the fiber optic probe; The fiber optic probe includes a first end face and a second end face. The first end face is used to receive the modulated light beam. The fiber optic probe can be inserted from the entry of the biopsy forceps and the second end face can be extended from the distal jaw region. The sample signal and reflected signal collected by the second end face are respectively input to the imaging module and the dynamic calibration module. The calibration module is used to measure the transmission matrix of the fiber optic probe; The imaging module is used to receive the sample signal and generate tissue images; The dynamic calibration module is used to acquire the current reflection pattern and generate compensation terms or compensation matrices based on the changes in intensity, centroid position, correlation coefficient and / or phase correlation peak position between the current reflection pattern and the reference reflection pattern, so as to dynamically calibrate the pre-measured transmission matrix and / or modulation control parameters. The computer is communicatively connected to the calibration module, imaging module, and dynamic calibration module, respectively, and is used to realize transmission matrix calculation, image reconstruction, dynamic calibration of the transmission matrix, and real-time tissue status discrimination. It is also used to generate sampling prompt information and sampling guidance control information for spatial correlation guidance based on the spatial correlation between the imaging area of ​​the fiber optic probe and the biopsy area, combined with the tissue status discrimination results, so as to assist the operator in adjusting the instrument position and performing biopsy sampling near the target area.

[0009] Preferably, the light source module outputs a single longitudinal mode continuous laser whose wavelength is matched with the scattering and / or fluorescence characteristics of the target cavity tissue. Specifically, the wavelength is selected based on the scattering intensity, absorption intensity, penetration depth, autofluorescence response, or excitation spectrum of the target cavity tissue at different wavelengths, so as to enhance the image contrast between normal and abnormal tissues and improve the sensitivity of identifying abnormal states of the target tissue.

[0010] Preferably, the modulation module includes a digital micromirror array device, a pinhole filter device, and a polarization modulation device, used to perform amplitude modulation, phase modulation, and polarization modulation on the input beam; wherein the modulation pattern is generated by any one of off-axis holography, phase conjugation, compressed sensing, or deep learning methods.

[0011] Preferably, the fiber optic probe is a single multimode fiber whose light transmission band covers the laser band of the light source module, and its second end face is provided with a surface functional layer with high transmittance and low reflectance.

[0012] Preferably, the biopsy forceps includes an operating handle assembly, an insertion cannula assembly, and a forceps head assembly connected in sequence, and the fiber optic probe can extend into the forceps channel inlet of the operating handle assembly and extend out from the forceps jaw area of ​​the forceps head assembly; The insertion cannula assembly is a flexible cannula with a hollow channel extending along the axial direction, the proximal end of which is connected to the operating handle assembly, and the distal end of which is connected to the clamp head assembly. The clamp head assembly includes a first clamping flap and a second clamping flap, which are rotatably connected around the same axis to achieve opening and closing movements, and have a through clamping channel outlet at their distal end. The operating handle assembly includes a pull ring, a push-pull rod, and a clamping channel inlet; the push-pull rod is used to drive the clamping head assembly to open and close. The clamp entrance, the hollow channel inside the push-pull rod, the hollow channel of the clamp head assembly connection structure, and the clamp exit are connected in sequence to form a through clamp structure for accommodating the optical fiber probe (8).

[0013] Preferably, the spatial relationship between the imaging area of ​​the fiber optic probe and the biopsy area is formed by the assembly position relationship between the fiber optic probe and the biopsy forceps, and / or obtained through pre-calibration, structural limitation, operational calibration and combinations thereof.

[0014] Preferably, the computer performs real-time organizational status determination, including: The real-time acquired tissue images are preprocessed, and spatial and frequency domain features are extracted to construct feature vectors for tissue state discrimination. Based on a pre-trained tissue state discrimination model, the feature vectors are subjected to discriminative analysis, and the target tissue state category and its corresponding probability value are output as the tissue state discrimination result. The tissue state discrimination model includes a feature input layer, a feature fusion layer, a classification discrimination layer, and a probability output layer. The classification discrimination layer adopts one of support vector machine, random forest, multilayer perceptron, or lightweight convolutional neural network. The tissue state categories include one or more of normal tissue, inflammatory or benign lesion tissue, intraepithelial neoplasia tissue, and cancerous tissue.

[0015] Preferably, based on the spatial correlation between the fiber optic probe imaging area and the biopsy area, and combined with the tissue state discrimination result, sampling prompt information and sampling guidance control information for spatial correlation guidance are generated, including: Based on the spatial correlation and the organization status discrimination result, corresponding sampling prompt information is generated. The sampling prompt information may include one or more of the following: continue forward delivery, appropriate retraction, adjust orientation, maintain current position, perform sampling, or repositioning. The tissue state discrimination result is compared with a preset discrimination threshold to generate sampling guidance control information, specifically including: when the preset guidance conditions are met, outputting sampling suggestions, key observation or positioning adjustment prompts; when the preset guidance conditions are not met, outputting prompts to continue observation, re-imaging or repositioning.

[0016] Preferably, the computer is used to constrain the generation of sampling prompt information by combining image quality indicators and / or tissue state discrimination confidence indicators. When the image quality indicators and / or tissue state discrimination confidence indicators do not meet the preset conditions, the computer outputs a re-imaging prompt, a repositioning prompt, or a continued observation prompt.

[0017] Preferably, the imaging module includes a filtering device and a signal acquisition device for separating, detecting, and digitizing the sample signal collected in reverse from the second end face of the fiber optic probe; wherein the filtering device is a long-pass filter and / or a dichroic mirror.

[0018] To achieve the above-mentioned objectives, this invention also provides a method for using the aforementioned cavity tissue biopsy guidance device based on fiber optic imaging and state determination, comprising the following steps: S1. Start the biopsy guidance device and set the system parameters; S2. Adjust the modulation module so that the modulated beam is coupled into the fiber optic probe; S3. Insert the fiber optic probe into the inlet of the biopsy forceps and make the second end face of the fiber optic probe protrude from the jaw region at the distal end of the biopsy forceps; S4. Measure the transmission matrix of the fiber optic probe and establish a dynamic calibration relationship based on the reflected signal; S5. Insert the biopsy forceps into the cavity to be tested through the endoscope working channel, and perform macroscopic localization of the suspected area under white light endoscope; S6. Under the control of the computer, a modulation pattern is generated in the modulation module, so that the modulation beam forms multiple focal points in front of the second end face of the fiber optic probe, and is scanned and imaged according to a predetermined scanning method, which may be raster scanning or snake scanning. S7. During fiber optic imaging, the imaging module and the computer generate images in real time and make real-time judgments on the tissue status at the current imaging position; at the same time, the dynamic calibration module collects the reflection signal of the fiber optic probe in real time and performs dynamic calibration on the transmission matrix based on the reflection signal. S8. Based on the tissue condition determination results and the spatial relationship between the fiber optic probe imaging area and the biopsy area, generate corresponding sampling prompt information; S9. When the tissue is determined to be intraepithelial neoplasia and / or cancerous tissue, and its corresponding probability value is higher than the preset threshold, output a sampling prompt message to assist the operator in controlling the biopsy forceps to perform biopsy sampling near the target area; when the determination result does not meet the preset sampling conditions, maintain the observation state or prompt repositioning.

[0019] Furthermore, before generating sampling prompts or performing biopsy sampling, constraints can be imposed by combining image quality indicators and / or tissue condition discrimination confidence indicators to improve guidance reliability.

[0020] Compared with the prior art, the beneficial effects of the present invention include at least the following: (1) Based on the existing macroscopic observation of endoscopy, the present invention introduces fiber optic microscopy imaging and real-time tissue state discrimination function, which can provide microscopic structural information of the target area and lesion risk judgment before sampling, reduce the simple reliance on the operator's experience, and improve the pertinence of lesion identification before biopsy.

[0021] (2) This invention establishes a spatial association between the imaging area of ​​the fiber optic probe and the biopsy area, and converts the tissue state discrimination result into sampling prompt information, thereby realizing the transformation from imaging observation to biopsy guidance and improving the accuracy and effectiveness of sampling positioning; in addition, this invention adopts a structural integration method in which the fiber optic probe is inserted through the biopsy forceps channel and the biopsy forceps are sent into the cavity through the existing endoscope working channel, which has good system integration and clinical application compatibility.

[0022] (3) The present invention dynamically calibrates the transmission matrix by reflecting the signal and can combine the image quality index and the tissue state discrimination confidence index to constrain the sampling prompt information, thereby improving the imaging stability, tissue state discrimination reliability and biopsy guidance reliability under fiber bending or environmental disturbance conditions. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 A schematic diagram of a cavity tissue biopsy guidance device based on fiber optic imaging and state assessment; Figure 2 This is a schematic diagram of the biopsy forceps in a cavity tissue biopsy guiding device; Figure 3 This is a schematic diagram of the forceps head assembly of a biopsy forceps; Figure 4 This is a flowchart illustrating a method for image processing and discrimination guidance of cavity tissue based on fiber optic imaging. Figure 5 This is a flowchart illustrating the usage of a cavity tissue biopsy guidance device based on fiber optic imaging and state assessment. Figure 6 A schematic diagram showing the results of the tissue status assessment and sampling instructions; In the diagram: 1. Light source module, 2. First lens, 3. First polarization beam splitter, 4. First objective lens, 5. Modulation module, 6. First beam splitter, 7. Second objective lens, 8. Fiber optic probe, 9. Biopsy forceps, 10. Calibration module, 11. Single-mode fiber, 12. Second beam splitter, 13. Imaging module, 14. Dynamic calibration module, 15. Computer. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of this invention.

[0026] like Figure 1 As shown, this embodiment provides a cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination. The core components of the device include a light source module 1, a modulation module 5, a fiber optic probe 8, a biopsy forceps 9, a calibration module 10, an imaging module 13, a dynamic calibration module 14, and a computer 15.

[0027] Among them, the light source module (1) is used to output a single longitudinal mode continuous laser as the basic light field for transmission matrix measurement and scanning imaging; the light source band is selected according to the scattering intensity, absorption intensity, penetration depth, autofluorescence response or excitation spectrum of exogenous fluorescent markers of the target cavity tissue at different wavelengths, so as to improve the sensitivity of the target tissue abnormality identification; the output end of the light source module 1 is connected to the first lens 2, the linear polarizer and the first polarization beam splitter 3 in sequence. The first polarization beam splitter 3 is used to divide the output laser into an object beam and a reference beam. The reference beam is input to the calibration module 10 through the first objective lens 4 and the single-mode fiber 11, and the object beam is input to the modulation module 5.

[0028] The modulation module 5 is used to perform amplitude modulation, phase modulation, and polarization modulation on the object light beam; the emitted modulated beam is coupled into the fiber optic probe 8 after passing through the first beam splitter 6 and the second objective lens 7 in sequence. The modulation module 5 provided in this embodiment includes a digital micromirror array device, a pinhole filter device, and a polarization modulation device, used to perform amplitude modulation, phase modulation, and polarization modulation on the input object light beam; the modulation pattern is generated by any of the following methods: off-axis holography, phase conjugation, compressed sensing, and deep learning.

[0029] The fiber optic probe 8 is used to transmit the modulated beam output from the modulation module 5 to the far end of the fiber optic cable, generating a focused scanning point at the far end to illuminate the sample, and transmitting the reflected signal and / or fluorescence signal of the sample in reverse to the imaging module 13 and the dynamic calibration module 14. The fiber optic probe 8 includes a first end face and a second end face. The first end face is aligned with the focal point of the second objective lens 7 and is used to receive the modulated beam focused by the second objective lens 7. The fiber optic probe 8 can extend into the clamping channel of the biopsy forceps 9, and the second end face extends out from the far end clamping area. The laser signal and / or fluorescence signal collected in reverse by the second end face of the fiber optic probe 8 are sequentially transmitted through the second objective lens 7, the first beam splitter 6, and the second beam splitter 12, and then input to the imaging module 13 and the dynamic calibration module 14, respectively. The fiber optic probe 8 provided in this embodiment is a single multimode fiber, whose light transmission band covers the laser band of the light source module 1, and its second end face is provided with a surface functional layer with high transmittance and low reflectance.

[0030] The biopsy forceps 9 are used to mechanically hold and sample tissue samples, and provide a through-type guide channel for the fiber optic probe 8. The biopsy forceps 9 are not independent actuators, but rather a sampling instrument carrier inserted through the endoscope's working channel. Their advancement, retraction, turning, and attitude adjustment are primarily controlled by the white light endoscope's mechanical operating system and the operator. The fiber optic probe 8 can be inserted from the proximal forceps channel entrance of the biopsy forceps 9 and extended from the distal forceps jaw region to perform microscopic imaging of the target tissue near the jaws.

[0031] The calibration module 10 is used to measure the transmission matrix of the fiber optic probe 8. The calibration module 10 includes a single-mode fiber 11 for receiving a reference beam and a camera for acquiring interference spots. The calibration module 10 is mounted on a one-dimensional displacement stage and is aligned with the second end face of the fiber optic probe 8 when measuring the transmission matrix. After the measurement is completed, it can be moved away from the fiber optic probe 8.

[0032] The imaging module 13 includes a filtering device and a signal acquisition device, used to separate, detect, and digitize the sample signal collected in reverse from the second end face of the fiber optic probe 8, and input the corresponding signal into the computer 15 to generate a tissue image through a signal processing program. In this embodiment, the filtering device may be a long-pass filter and / or a dichroic mirror; the signal acquisition device may include a light signal collection device, a signal conversion device, and a digitization device, wherein the light signal collection device may be a photomultiplier tube, a single-photon avalanche diode, a semiconductor photodetector, or a camera-type detector, the signal conversion device may be a current-to-voltage converter or a charge amplifier, and the digitization device may be an analog-to-digital converter or a high-speed digital oscilloscope.

[0033] The dynamic calibration module 14 is used to acquire the reflected signal of the fiber optic probe 8 in real time during the imaging process and to dynamically calibrate the transmission matrix based on the reflected signal. Since the fiber optic probe 8 may bend, twist, be subjected to force, or have changes in attitude during insertion and operation, the pre-measured transmission matrix may deviate from the current actual transmission state; therefore, the reflected signal can be used to characterize the current fiber optic transmission state, and the transmission matrix can be corrected as a whole, locally, or through model compensation to improve imaging stability and real-time performance.

[0034] For example, the transmission modulation relationship after dynamic calibration can be expressed as: in, For the pre-measured transfer matrix, These are the feature parameters extracted from the reflected signal. The compensation term is generated based on the characteristics of the reflected signal. This represents the transmission modulation relationship after dynamic calibration.

[0035] In one specific implementation, the system completes the initial transmission matrix. After measurement, record the corresponding reference reflection pattern. During the imaging process, the dynamic calibration module 14 acquires the current reflection pattern in real time. The system extracts characteristic parameters such as reflected spot intensity, centroid position, correlation coefficient, and / or phase correlation peak position to obtain the characteristic changes of the current reflection pattern relative to the reference reflection pattern. Based on these characteristic changes, it estimates the amplitude change, phase drift, and / or linear phase tilt caused by changes in fiber state and generates compensation terms. or compensation matrix The pre-measured transfer matrix is ​​then corrected.

[0036] For example, the transmission modulation relationship can be modified in the following way: in, This is a compensation matrix generated based on the characteristic parameters of the reflected signal. The compensation term... or compensation matrix It can be obtained through lookup tables, regression models, neural network models, or correlation matching methods. Therefore, it is possible to compensate for fiber optic transmission state drift in real time using reflected signals without re-measuring the entire transmission matrix.

[0037] Computer 15 is communicatively connected to calibration module 10, imaging module 13, and dynamic calibration module 14. Its built-in information processing program is used to perform transmission matrix calculation, image reconstruction, real-time tissue status determination, and sampling prompt information generation. Since the fiber optic probe 8 extends from the jaws of the biopsy forceps 9, a spatial relationship is formed between the fiber optic probe imaging area and the biopsy area, which can be used for sampling guidance. This spatial relationship can be formed by the assembly position relationship between the fiber optic probe 8 and the biopsy forceps 9, or it can be obtained or corrected through pre-calibration, structural constraints, operational calibration, and their combinations. Based on the spatial relationship, computer 15 can convert the tissue status determination result into corresponding sampling prompt information to assist the operator in advancing, retracting, adjusting orientation, maintaining the current position, or performing sampling.

[0038] For example, the spatial relationship between the imaging area of ​​the fiber optic probe and the biopsy area can be represented as: in, This indicates the target location within the imaging area of ​​the fiber optic probe. This indicates the fiber optic probe extension length, biopsy forceps posture, jaw opening / closing status, and / or assembly structure parameters. Indicates the location of the corresponding biopsy area. This represents the spatial correlation mapping relationship between the imaging area of ​​the fiber optic probe and the biopsy area. In some embodiments, the spatial correlation relationship can be characterized by scalar offset, two-dimensional position offset, or piecewise function relationship, and can be pre-calibrated or corrected in real time for different extension lengths, different jaw opening and closing states, and different instrument postures to improve the consistency and accuracy of biopsy guidance. The above relationship is only an illustrative expression of the principle of spatial correlation guidance and does not constitute a limitation on the specific form of the mapping model.

[0039] In some embodiments, the spatial relationship between the fiber optic probe imaging area and the biopsy area can be directly formed by the assembly position relationship between the biopsy forceps 9 and the fiber optic probe 8. For example, when the fiber optic probe 8 extends from a fixed position in the jaw region, its distal imaging area is typically located in front of the actual clamping area of ​​the jaws, adjacent to the biopsy area, or partially overlapping with the biopsy area. In other embodiments, the spatial relationship can also be obtained or corrected through pre-experimental calibration, structural constraints, operational calibration, and combinations thereof. For example, under in vitro sample or standard target conditions, the relative positional relationship between the fiber optic probe imaging area and the actual biopsy area under different fiber extension states and different instrument postures can be recorded and stored in the computer 15 as a basis for subsequently generating sampling prompt information.

[0040] like Figure 2 As shown, this embodiment provides a biopsy forceps 9 in a cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination, including an operating handle assembly 9-1, an insertion cannula assembly 9-2, and a forceps head assembly 9-3, which are connected in sequence.

[0041] The insertion sleeve assembly 9-2 is a flexible plastic-coated spring sleeve with an axially extending hollow channel. Its proximal end is connected to the operating handle assembly 9-1, and its distal end is connected to the jaw assembly 9-3. The operating handle assembly 9-1 includes a pull ring 9-7, a push-pull rod 9-8, and a jaw passage inlet 9-9. The push-pull rod 9-8 is used to drive the jaw assembly to open and close. The jaw assembly 9-3 includes a first jaw flap 9-4 and a second jaw flap 9-5. The first jaw flap 9-4 and the second jaw flap 9-5 are rotatably connected around the same axis to realize the opening and closing movement. Its distal end has a through jaw passage outlet 9-6. A schematic diagram of the jaw assembly 9-3 is shown below. Figure 3 As shown, the forceps inlet 9-9, the hollow channel inside the push-pull rod, the hollow channel connecting the forceps head assembly 9-3, and the forceps outlet 9-6 are sequentially connected to form a through-path forceps structure for accommodating the fiber optic probe 8. The fiber optic probe 8 can be inserted through the forceps inlet 9-9 and extended from the forceps outlet 9-6. The overall size of the biopsy forceps 9 is preferably matched with the working channel of a conventional endoscope, enabling it to be inserted and sampled in the esophagus, stomach, intestines, respiratory tract, and other curved cavities in conjunction with the endoscope.

[0042] like Figure 4 As shown, this embodiment provides a method for image processing and discrimination guidance of cavity tissue based on fiber optic imaging, including the following steps: S1: Image Acquisition and Feature Construction: The fiber optic probe imaging images acquired in real time by the biopsy guidance device are preprocessed, including noise reduction and intensity normalization. Based on this, spatial domain features and frequency domain features are extracted, and feature vectors for tissue state discrimination are constructed. Spatial domain features may include contrast, uniformity, texture variation features, etc., while frequency domain features may include average amplitude spectrum, high-frequency energy, and spectral frequency statistics.

[0043] S2: Tissue State Discrimination: Based on a pre-trained tissue state discrimination model, discriminant analysis is performed on the feature vector to output the state category of the target tissue and its corresponding probability value. The tissue state category may include one or more of the following: normal tissue, inflammatory or benign lesion tissue, intraepithelial neoplasia tissue, and cancerous tissue.

[0044] In one embodiment, the tissue state discrimination model includes a feature input layer, a feature fusion layer, a classification discrimination layer, and a probability output layer. The feature input layer receives spatial domain features, frequency domain features, and / or fluorescence intensity distribution features; the feature fusion layer performs feature normalization and fusion; the classification discrimination layer employs one of support vector machines, random forests, multilayer perceptrons, or lightweight convolutional neural networks; and the probability output layer outputs the probability values ​​for each tissue state category. The model can be trained using cavity tissue images annotated with pathological results, and during real-time discrimination, it generates tissue state discrimination results and sampling prompts based on the probabilities of each category.

[0045] S3: Spatial Correlation Guidance: Based on the spatial correlation between the imaging area of ​​the fiber optic probe and the biopsy area, the tissue state discrimination result is converted into corresponding sampling prompt information. The spatial correlation can be formed by the structural positional relationship of the fiber optic probe 8 extending from the jaws of the biopsy forceps 9, or corrected through pre-calibration, structural limitation, and operational calibration. The sampling prompt information may include one or more of the following: continue forward, appropriate retraction, adjust orientation, maintain current position, perform sampling, or repositioning.

[0046] S4: Sampling guidance control: Based on the tissue state category and its corresponding probability value, compare it with a preset threshold; when the preset sampling conditions are met, output sampling prompt information; when the preset sampling conditions are not met, output prompts to continue observation, re-imaging, or repositioning.

[0047] In some embodiments, the generation of the sampling prompt information may also be combined with image quality indicators and / or tissue status discrimination confidence indicators. When the image quality indicators and / or tissue status discrimination confidence indicators do not meet preset conditions, prompts such as re-imaging, repositioning, or continued observation are output first, rather than directly outputting prompts to perform sampling, in order to improve the reliability of biopsy guidance and reduce the impact of a single misjudgment on the sampling decision.

[0048] like Figure 5 As shown, this embodiment provides a method for using a cavity tissue biopsy guidance device based on fiber optic imaging and state determination, including the following steps: S1: Start the cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination, including turning on the light source module 1, modulation module 5, imaging module 13 and computer 15, and start the information processing program in computer 15; set the corresponding system parameters according to the laser band of the single longitudinal mode continuous laser, the type of fiber optic probe 8 and the size of the image to be generated.

[0049] S2: Adjust the pinhole filter and polarization modulation device in the modulation module 5 so that the modulated beam is coupled into the fiber optic probe 8 with a predetermined coupling efficiency.

[0050] S3: Insert the fiber optic probe 8 into the jaw inlet of the biopsy forceps 9, and make the second end face of the fiber optic probe 8 protrude from the jaw area at the distal end of the biopsy forceps 9.

[0051] S4: Fix the insertion sleeve assembly 9-2 and the clamp head assembly 9-3 of the biopsy forceps 9 to the clamping device, and align the calibration module 10 with the second end face of the fiber optic probe 8 using a one-dimensional displacement stage; under the control of the computer 15, measure the transmission matrix of the fiber optic probe 8 and establish a dynamic calibration relationship based on the reflected signal; after completing the transmission matrix measurement and establishing the dynamic calibration relationship, control the one-dimensional displacement stage to move the calibration module 10 away from the fiber optic probe 8.

[0052] S5: Insert the biopsy forceps 9 through the endoscope working channel and enter the cavity to be tested along with the endoscope; perform macroscopic observation of the tissue in the cavity under white light endoscope mode to initially locate the suspected lesion area; then extend the fiber optic probe 8 from the forceps area and align it with the sample to be tested near the suspected area.

[0053] S6: Under the control of computer 15, a series of modulation patterns are generated in modulation module 5 to modulate the incident object light beam, so that the modulated beam forms multiple focal points in front of the second end face of the fiber optic probe 8, and performs scanning imaging according to a predetermined scanning method, which can be raster scanning or serpentine scanning.

[0054] S7: During fiber optic imaging, the imaging module 13 and computer 15 generate images in real time, and the tissue state at the current imaging position is determined in real time by the tissue state discrimination method in computer 15, outputting the probability distribution results of normal tissue, inflammatory or benign lesion tissue, intraepithelial neoplasia tissue and cancerous tissue; at the same time, the dynamic calibration module 14 collects the reflection signal of the fiber optic probe 8 in real time, and performs dynamic calibration on the transmission matrix based on the reflection signal to maintain imaging stability and tissue state discrimination reliability.

[0055] S8: Based on the tissue condition determination results and the spatial relationship between the fiber optic probe imaging area and the biopsy area, generate corresponding sampling prompt information; the operator adjusts the instrument position through the endoscopic mechanical operating system according to the sampling prompt information and in combination with the white light endoscope field of view, so that the biopsy area is close to or corresponds to the determined target area.

[0056] S9: When the tissue is determined to be intraepithelial neoplasia and / or cancerous tissue, and its corresponding probability value is higher than the preset threshold, a sampling prompt message is output; the operator controls the push-pull rod 9-8 of the biopsy forceps 9 to move relative to the pull ring 9-7, driving the first clamp flap 9-4 and the second clamp flap 9-5 to open and close, so as to perform biopsy sampling near the target area; when the determination result does not meet the preset sampling conditions, the observation state is maintained or a repositioning prompt is given.

[0057] In some embodiments, in steps S8 or S9 above, the output of a sampling prompt message can be constrained by combining image quality indicators and / or tissue state discrimination confidence indicators. When the image quality indicators and / or tissue state discrimination confidence indicators do not meet the preset conditions, prompts to continue observation, re-imaging, or repositioning are output first, instead of directly performing sampling.

[0058] like Figure 6 As shown, in this embodiment, the fiber optic probe 8 extends from the distal jaw region of the biopsy forceps 9 and is aligned with the tissue to be tested. The computer 15 performs real-time discrimination of the tissue state within the current field of view based on the fiber optic imaging results and outputs the corresponding tissue state discrimination results and sampling prompts on the display interface. When the imaging results show abnormal tissue features in the target area, the corresponding instrument adjustment instructions and sampling suggestions can be generated by further combining the spatial correlation between the imaging area of ​​the fiber optic probe 8 and the biopsy area to assist the operator in adjusting the position of the biopsy forceps 9 and performing sampling when the preset conditions are met, thereby realizing the linkage of imaging, discrimination and sampling guidance.

[0059] The key to this invention lies in establishing the spatial correlation between the imaging area of ​​the fiber optic probe and the biopsy area, as well as the dynamic feedback relationship between the reflected signal and the transmission modulation relationship, thereby realizing closed-loop guidance for cavity tissue biopsy scenarios.

[0060] The specific embodiments described above illustrate the technical solution and beneficial effects of the present invention in detail. It should be understood that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A cavity tissue biopsy guidance device based on fiber optic imaging and state determination, characterized in that, It includes a light source module (1), a modulation module (5), an optical fiber probe (8), a biopsy forceps (9), a calibration module (10), an imaging module (13), a dynamic calibration module (14), and a computer (15). The modulation module (5) is used to modulate the incident beam provided by the light source module (1) and couple the modulated beam into the fiber optic probe (8). The fiber optic probe (8) includes a first end face and a second end face. The first end face is used to receive the modulated light beam. The fiber optic probe (8) can be inserted from the clamping entrance of the biopsy forceps (9), and the second end face of the fiber optic probe (8) extends from the clamping jaw area at the distal end of the biopsy forceps (9). The sample signal and the reflected signal collected by the second end face are respectively input to the imaging module (13) and the dynamic calibration module (14). The calibration module (10) is used to measure the transmission matrix of the fiber optic probe (8); The imaging module (13) is used to receive the sample signal and generate a tissue image; The dynamic calibration module (14) is used to acquire the current reflection pattern and generate a compensation term or compensation matrix based on the changes in intensity, centroid position, correlation coefficient and / or phase correlation peak position between the current reflection pattern and the reference reflection pattern, so as to dynamically calibrate the pre-measured transmission matrix. The computer (15) is connected to the calibration module (10), the imaging module (13) and the dynamic calibration module (14) respectively, and is used to realize transmission matrix calculation, image reconstruction, dynamic calibration of transmission matrix and real-time tissue status discrimination; it is also used to generate sampling prompt information and sampling guidance control information for spatial correlation guidance based on the spatial correlation relationship between the fiber optic probe imaging area and the biopsy area, combined with the tissue status discrimination result, so as to assist the operator in adjusting the instrument position and performing biopsy sampling near the target area.

2. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The light source module (1) outputs a single longitudinal mode continuous laser, the wavelength of which is selected according to the scattering intensity, absorption intensity, penetration depth, autofluorescence response or excitation spectrum of the target cavity tissue at different wavelengths, so as to enhance the image contrast between normal and abnormal tissues.

3. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The modulation module (5) includes a digital micromirror array device, a pinhole filter device and a polarization modulation device, used to perform amplitude modulation, phase modulation and polarization modulation on the input beam; The modulation pattern is generated using any one of the following methods: off-axis holography, phase conjugation, compressed sensing, or deep learning.

4. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The fiber optic probe (8) is a single multimode fiber, whose light transmission band covers the laser band of the light source module (1), and its second end face is provided with a surface functional layer with high transmittance and low reflectance.

5. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The biopsy forceps (9) includes an operating handle assembly, an insertion cannula assembly and a forceps head assembly connected in sequence. The fiber optic probe (8) can be inserted from the forceps channel entrance of the operating handle assembly and extended from the jaw area of ​​the forceps head assembly. The insertion sleeve assembly is a flexible sleeve with a hollow channel extending along the axial direction; The clamp head assembly includes a first clamping flap and a second clamping flap, which are rotatably connected around the same axis to achieve opening and closing movements, and have a through clamping channel outlet at their distal end. The operating handle assembly includes a pull ring, a push-pull rod, and a clamping channel inlet; the push-pull rod is used to drive the clamping head assembly to open and close. The clamp entrance, the hollow channel inside the push-pull rod, the hollow channel of the clamp head assembly connection structure, and the clamp exit are connected in sequence to form a through clamp structure for accommodating the optical fiber probe (8).

6. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The spatial relationship between the imaging area of ​​the fiber optic probe and the biopsy area is formed by the assembly position relationship between the fiber optic probe (8) and the biopsy forceps (9), and / or obtained through pre-calibration, structural limitation, operation calibration and their combination.

7. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The computer performs real-time organizational status determination, including: The real-time acquired tissue images are preprocessed, and spatial domain features and frequency domain features are extracted to construct a feature vector for tissue state discrimination. Based on the pre-trained tissue state discrimination model, the feature vector is subjected to discrimination analysis, and the state category of the target tissue and its corresponding probability value are output as the tissue state discrimination result. The tissue state discrimination model includes a feature input layer, a feature fusion layer, a classification discrimination layer, and a probability output layer. The classification discrimination layer adopts one of the following: support vector machine, random forest, multilayer perceptron, or lightweight convolutional neural network. The tissue state categories include one or more of the following: normal tissue, inflammatory or benign lesion tissue, intraepithelial neoplasia tissue, and cancerous tissue.

8. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, The sampling prompt information and sampling guidance control information for spatial correlation guidance are generated based on the spatial correlation between the imaging area of ​​the fiber optic probe and the biopsy area, combined with the tissue state discrimination result. These include: Based on the spatial correlation and the organization status discrimination result, corresponding sampling prompt information is generated. The sampling prompt information may include one or more of the following: continue forward delivery, appropriate retraction, adjust orientation, maintain current position, perform sampling, or repositioning. The tissue state discrimination result is compared with a preset discrimination threshold to generate sampling guidance control information, specifically including: when the preset guidance conditions are met, outputting sampling suggestions, key observation or positioning adjustment prompts; when the preset guidance conditions are not met, outputting prompts to continue observation, re-imaging or repositioning.

9. The cavity tissue biopsy guidance device based on fiber optic imaging and state discrimination according to claim 1, characterized in that, When generating sampling prompt information, the computer (15) combines image quality indicators and / or tissue status discrimination confidence indicators for constraints. When the image quality indicators and / or tissue status discrimination confidence indicators do not meet the preset conditions, it outputs a re-imaging prompt, a repositioning prompt, or a continued observation prompt.

10. A method of using the cavity tissue biopsy guidance device based on fiber optic imaging and state determination as described in any one of claims 1-9, characterized in that, Includes the following steps: S1. Start the biopsy guidance device and set the system parameters; S2. Adjust the modulation module (5) to couple the modulated beam into the fiber optic probe (8). S3. Insert the fiber optic probe (8) into the jaw inlet of the biopsy forceps (9) and make the second end face of the fiber optic probe (8) protrude from the jaw region at the distal end of the biopsy forceps (9); S4. Measure the transmission matrix of the fiber optic probe (8) and establish a dynamic calibration relationship based on the reflected signal; S5. Insert the biopsy forceps (9) into the cavity to be tested through the endoscope working channel, and perform macroscopic positioning of the suspected area under white light endoscope; S6. Under the control of the computer (15), a modulation pattern is generated in the modulation module (5) so that the modulation beam forms multiple focal points in front of its second end face through the fiber optic probe (8) and performs scanning imaging according to a predetermined scanning method. The scanning method can be grating scanning or snake scanning. S7. During the fiber optic imaging process, the imaging module (13) and the computer (15) generate images in real time and make real-time judgments on the tissue status at the current imaging position; at the same time, the dynamic calibration module (14) collects the reflection signal of the fiber optic probe (8) in real time and performs dynamic calibration on the transmission matrix based on the reflection signal. S8. Based on the tissue condition determination results and the spatial relationship between the fiber optic probe imaging area and the biopsy area, generate corresponding sampling prompt information; S9. When the tissue is determined to be intraepithelial neoplasia and / or cancerous tissue, and its corresponding probability value is higher than the preset threshold, a sampling prompt message is output to assist the operator in controlling the biopsy forceps (9) to perform biopsy sampling near the target area; when the determination result does not meet the preset sampling conditions, the observation state is maintained or a prompt for repositioning is given.

Images