Combined narrowband and fluorescence imaging method and system for bronchoscope focus detection

Abstract

The invention relates to the technical field of medical image processing, in particular to a combined narrowband and fluorescence imaging method and system for bronchoscope focus detection. According to the method provided by the invention, based on deep learning, features with discrimination capability in a narrow-band imaging image and an auto-fluorescence imaging image are selected and extracted, fusion of different imaging technology advantages of focus detection is realized through decision fusion, and the speed and accuracy of focus identification are improved; meanwhile, the method provided by the invention can better adapt to different scenes and lesion features, has better generalization ability, improves the specificity and sensitivity of focus detection, and is beneficial to improving the diagnosis accuracy and efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a combined narrow-band and fluorescence imaging method and system for bronchoscopic lesion detection. Background Art

[0002] Lung cancer is one of the most common types of cancer worldwide and is also the leading cause of death from cancer. The survival time of lung cancer patients is highly correlated with the stage of the disease and the size of the tumor at the time of diagnosis, and early diagnosis is the most important step in the diagnosis and treatment of lung cancer. Current lung cancer screening mainly relies on low-dose chest CT and chest X-ray. The characteristics of chest CT and chest X-ray examinations determine that current chest imaging technology is more sensitive to peripheral lung cancer, and has poor sensitivity to central bronchial lung cancer with squamous cell carcinoma and small cell lung cancer as the main pathological types. Tobacco exposure causes central lung cancer with squamous cell carcinoma as the main pathological type. In order to improve the early detection rate of central lung cancer, the current clinical diagnosis of lung cancer usually uses new bronchoscopic technologies such as narrow-band imaging (NBI) and autofluorescence bronchial imaging (AFI).

[0003] Narrow band imaging and autofluorescence imaging are improvements based on white light endoscopic imaging (WLB) to better help doctors detect early lung cancer symptoms. These two technologies are currently widely used in the diagnosis of precancerous lesions and early central lung cancer (ECLC) under bronchoscopy. The results of most clinical trials show that the use of AFI and NBI has potential benefits for the diagnosis of lung cancer, especially for the detection of precancerous lesions.

[0004] NBI narrow-band imaging technology uses a filter to filter out the broadband spectrum of red, blue, and green emitted by ordinary endoscopic light sources, and selects narrow-band light waves of 415nm and 540nm. The 415nm blue light has a short wavelength and penetrates the superficial mucosa. The absorption peak of hemoglobin to light waves is also located around 415nm, which is conducive to displaying the surface blood vessels on the surface of the bronchial mucosa. The 540nm green light has strong penetrating power and has a good display effect on the blood vessels in the submucosal layer. It can display the fine structure of the mucosal surface and the submucosal blood vessels, which is clearer than the traditional white light mode endoscope and has a stronger three-dimensional sense. Through high-contrast observation of the mucosal surface, it can provide important subtle graphics for the diagnosis of vascular lesions. However, NBI focuses on detecting vascular structures, which may limit other important information in the normal field of view. In addition, the images produced by NBI have special colors, so it has high requirements for endoscopic imaging quality. The characteristics of the lesions are not obvious, and high-definition endoscopes with magnification functions are often used. In this way, the mucosal microvascular and glandular duct morphologies can be displayed more clearly and the nature of the lesions can be judged more accurately. This requires doctors to spend extra time and energy to observe the lesions, and for doctors who use NBI for diagnosis, they need certain training and practice to correctly interpret NBI images.

[0005] AFI autofluorescence bronchoscope technology is a new type of bronchoscope developed by combining cell autofluorescence and computer image analysis technology. Through the fluorescence changes of the tracheal and bronchial mucosa, the presence of abnormal lesions can be judged, and the mucosa of the abnormal fluorescence area can be brushed or biopsied according to the display of the fluorescence image, which increases the detection rate of early lung cancer, especially precancerous lesions or carcinoma in situ and local mucosal invasive cancer. However, due to its low specificity, it is easy to confuse normal bronchial mucosal tissue with abnormal tissue when observing under the microscope, making it difficult to identify the area where pathological tissue is actually needed.

[0006] The combined use of these two imaging techniques can improve sensitivity and specificity. The results of most clinical trials show that the use of AFI and NBI has potential benefits for the diagnosis of lung cancer. However, AFI and NBI use different bronchoscopes, which require the use of two different bronchoscopes on the same patient, increasing the examination time and requiring at least two in-and-out glottis procedures. This is the main reason why there are currently few applications and literature on the combined use of AFI and NBI in the diagnosis of lung cancer. Summary of the invention

[0007] The present invention aims to solve the problem that different imaging technologies are difficult to combine to improve detection accuracy in existing lesion detection methods.

[0008] To solve the above technical problems, in a first aspect, the present invention provides a combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection, comprising the following steps:

[0009] The white light imaging images, narrow band imaging images and autofluorescence imaging images of the lesion area under bronchoscopy were collected respectively, and the images were preprocessed by using the preset image enhancement methods respectively;

[0010] Using the white light imaging image as an input of a preset adversarial generative network model, and outputting a narrowband imaging generated image and an autofluorescence imaging generated image through the preset adversarial generative network model, wherein the preset adversarial generative network model includes a generator network and a discriminator network;

[0011] A lesion detection model is obtained based on the training of the narrow-band imaging generated image and the autofluorescence imaging generated image, and the lesion detection model is used to perform lesion detection on the narrow-band imaging image and the autofluorescence imaging image to obtain narrow-band imaging detection results and autofluorescence image detection results, respectively. The narrow-band imaging detection results and the autofluorescence image detection results are fused by a preset weighted fusion method to obtain a lesion detection result.

[0012] Furthermore, the preset image enhancement method includes at least one of image quality enhancement and image color correction.

[0013] Furthermore, in the preset adversarial generative network model, the probability that the generator network generates z types of images is defined as G(z), the probability that the discriminator network discriminates a real image is defined as D(x), and the process of the generator network generating an image and the discriminator network discriminating an image satisfies the value function V(G,D):

[0014]

[0015] in:

[0016]

[0017] In the above expression, is the initial data distribution, z is a random vector of Gaussian white noise, is the randomly initialized noise distribution, θ d is the network parameter of the discriminator network, θ g are the network parameters of the generator network, and m is the number of samples.

[0018] Furthermore, the lesion detection model is based on the YOLOv8 model.

[0019] Furthermore, in the step of training a lesion detection model based on the narrow-band imaging and the autofluorescence imaging, a classification loss function and a bounding box inspection function are set for the lesion detection model, and the classification loss function satisfies:

[0020]

[0021] Among them, x i is the predicted value of the current category, y i is the probability of the current category after the activation function, is the true value of the current category, and N is the number of samples;

[0022] The bounding box loss function satisfies:

[0023]

[0024] Among them, b i is the actual bounding box, is the predicted bounding box.

[0025] Furthermore, in the step of training a lesion detection model based on the narrow-band imaging generated image and the autofluorescence imaging generated image, the classification loss function and the bounding box loss function for the narrow-band imaging generated image are defined as: and The classification loss function and bounding box loss function for the image generated by the autofluorescence imaging are respectively and The joint loss function of the lesion detection model satisfies:

[0026]

[0027] Among them, α and β are the weight parameters of narrowband imaging and autofluorescence imaging, respectively, and λ is the weight parameter of weighted loss.

[0028] Furthermore, the preset weighted fusion method is specifically:

[0029] The narrowband imaging detection result and the autofluorescence image detection result are defined as D NBI , D AFI , which satisfies:

[0030] D NBI ={(B i ,C i )|i=1,2,…,n};

[0031] D AFI ={(B j ,C j )|j=1,2,…,m};

[0032] Among them, B i and B j represent the bounding boxes of the lesion area in the narrow-band imaging detection result and the autofluorescence image detection result, respectively, i and C jrepresents the confidence C of the narrowband imaging detection result and the autofluorescence image detection result NBI , C AFI ;

[0033] The lesion detection result is defined as one that satisfies:

[0034] D fused ={(B i ,C fused,i )|i=1,2,…,n};

[0035] C fused =α·C NBI +β·C AFI ;

[0036] Among them, α and β are weight parameters, and α+β=1.

[0037] In a second aspect, the present invention further provides a combined narrow band and fluorescence imaging system for bronchoscopic lesion detection, comprising:

[0038] An imaging module, used to respectively collect white light imaging images, narrow band imaging images and autofluorescence imaging images of the lesion area of ​​the bronchoscope, and respectively apply preset image enhancement methods to perform image preprocessing;

[0039] A processing module, used to use the white light imaging image as an input of a preset adversarial generative network model, and output a narrowband imaging generated image and an autofluorescence imaging generated image through the preset adversarial generative network model, wherein the preset adversarial generative network model includes a generator network and a discriminator network;

[0040] A detection module is used to obtain a lesion detection model based on the image generated by narrowband imaging and the image generated by autofluorescence imaging, and use the lesion detection model to perform lesion detection on the narrowband imaging image and the autofluorescence imaging image to obtain a narrowband imaging detection result and an autofluorescence image detection result, respectively, and fuse the narrowband imaging detection result and the autofluorescence image detection result by a preset weighted fusion method to obtain a lesion detection result.

[0041] The beneficial effect achieved by the present invention is that a combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection is proposed. The method selects and extracts features with discriminative capabilities in narrow-band imaging images and autofluorescence imaging images based on deep learning, and integrates the advantages of different imaging technologies for lesion detection through decision fusion, thereby improving the speed and accuracy of lesion identification. At the same time, the method provided by the present invention can better adapt to different scenarios and lesion characteristics, has better generalization ability, improves the specificity and sensitivity of lesion detection, and is conducive to improving diagnostic accuracy and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the steps of a combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0044] Please refer to Figure 1 , Figure 1 The present invention provides a schematic flow chart of the steps of a combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection provided by an embodiment of the present invention. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection includes the following steps:

[0045] S1, respectively collecting white light imaging images, narrow band imaging images and autofluorescence imaging images of the lesion area of ​​the bronchoscopy, and applying preset image enhancement methods to perform image preprocessing;

[0046] S2, using the white light imaging image as an input of a preset generative adversarial network model (GAN), and outputting a narrowband imaging generated image and an autofluorescence imaging generated image through the preset generative adversarial network model, wherein the preset generative adversarial network model includes a generator network and a discriminator network;

[0047] S3. A lesion detection model is obtained based on the training of the narrow-band imaging generated image and the autofluorescence imaging generated image, and the lesion detection model is used to perform lesion detection on the narrow-band imaging image and the autofluorescence imaging image to obtain narrow-band imaging detection results and autofluorescence image detection results respectively, and the narrow-band imaging detection results and the autofluorescence image detection results are fused by a preset weighted fusion method to obtain a lesion detection result.

[0048] The preset image enhancement method includes at least one of image quality enhancement and image color correction.

[0049] Specifically, both narrow-band imaging and autofluorescence imaging technologies have high requirements for endoscopic imaging quality. When performing narrow-band imaging detection, high-definition endoscopes with magnification functions are often used, so that the mucosal microvascular and glandular duct morphology can be displayed more clearly, and the nature of the lesion can be judged more accurately. In the autofluorescence imaging endoscopic mode, normal tissue is light green, inflammation is dark green or light purple, tumor-like lesions are purple or brown, and tumor tissue is magenta or reddish brown.

[0050] Therefore, the imaging quality of bronchoscopes is crucial for detecting and diagnosing lesions. When acquiring images using bronchoscopes with different imaging technologies, image enhancement and color correction are required to improve image quality and ensure color accuracy.

[0051] In the embodiment of the present invention, the adversarial generative network is used to generate images with narrowband imaging and AFI imaging effects respectively, which can effectively make up for the difficulty of the two imaging technologies in simultaneously collecting images in the same bronchoscope. Furthermore, in the preset adversarial generative network model, the probability of the generator network generating z types of images is defined as G(z), and the probability of the discriminator network distinguishing the real image is D(x). The process of the generator network generating images and the discriminator network distinguishing images satisfies the value function V(G,D):

[0052]

[0053] in:

[0054]

[0055] In the above expression, is the initial data distribution, z is a random vector of Gaussian white noise, is the randomly initialized noise distribution, θ d is the network parameter of the discriminator network, θ g are the network parameters of the generator network, and m is the number of samples.

[0056] The training process is as follows: First, train the discriminator D, from P data Dataset and random noise distribution P z Take m samples from the discriminator D and update the network parameters θ in the discriminator D through the gradient ascent algorithm d , optimize the cost function D(x) of the discriminator, that is, maximize the value function V(G,D);

[0057] In the second step, when training the generator G, we get the random noise distribution P z Take m samples from the generator G and update the network parameters θ in the generator G through the gradient descent algorithm. g , optimize the cost function D(G(z)) of the generator, that is, minimize the value function V(G,D).

[0058] During the processing of the entire adversarial generative network model, the style transfer imaging parameters set by the generator network input the original white light imaging image into the generator network to generate data samples with narrowband imaging and autofluorescence imaging effects respectively; the images with narrowband imaging and autofluorescence imaging effects are then input into the discriminator for iterative adversarial training. The adversarial training process includes alternating training of the generator and the discriminator. In each training, the discriminator network distinguishes between the generated images and the real imaging images, and the generator network adjusts the parameters of the generated images so that the imaging effect of the generated images continues to approach the real images. When the number of iterations of the adversarial generative network model reaches the threshold and the generated images and the real images cannot be distinguished, the adversarial learning training is completed.

[0059] The lesion detection model is based on the YOLOv8 model, which is a detection model based on deep learning.

[0060] Each channel of the YOLOv8 model processes the corresponding image features separately. In the first few layers of the YOLOv8 model, features are extracted for narrowband imaging and autofluorescence imaging images respectively. Then, a joint loss function is applied to the image features of the two imaging technologies to balance the detection tasks of narrowband imaging and autofluorescence imaging images, which can help the model learn and optimize the features of the two imaging modes at the same time. The joint loss function is usually composed of multiple sub-loss functions, each of which corresponds to a specific task or data source.

[0061] Specifically, in the step of training a lesion detection model based on the narrow-band imaging and the autofluorescence imaging, a classification loss function and a bounding box inspection function are set for the lesion detection model, and the classification loss function satisfies:

[0062]

[0063] Among them, x i is the predicted value of the current category, y i is the probability of the current category after the activation function, is the true value of the current category, and N is the number of samples;

[0064] The bounding box loss function satisfies:

[0065]

[0066] Among them, b i is the actual bounding box, is the predicted bounding box.

[0067] In the step of training a lesion detection model based on the narrow-band imaging generated image and the autofluorescence imaging generated image, the classification loss function and the bounding box loss function for the narrow-band imaging generated image are defined as follows: and The classification loss function and bounding box loss function for the image generated by the autofluorescence imaging are respectively and The joint loss function of the lesion detection model satisfies:

[0068]

[0069] Among them, α and β are the weight parameters of narrowband imaging and autofluorescence imaging, respectively, and λ is the weight parameter of weighted loss.

[0070] In the above process, the image datasets of narrowband imaging and autofluorescence imaging generated by the adversarial generative network model in step S2 are used to provide data sources for YOLOv8 model training, and the annotation tool is used to manually annotate the lesions in each image to determine the exact location and category of the lesions. During the training process, the weight parameters α and β need to be dynamically adjusted, and the model parameters are updated according to the loss value. By combining the loss function, the model can simultaneously optimize the detection performance of narrowband imaging and autofluorescence imaging images, and improve the overall detection effect.

[0071] The preset weighted fusion method is specifically:

[0072] The narrowband imaging detection result and the autofluorescence image detection result are defined as D NBI , D AFI , which satisfies:

[0073] D NBI ={(B i ,C i )|i=1,2,…,n};

[0074] D AFI ={(B j ,C j )|j=1,2,…,m};

[0075] Among them, B i and B j represent the bounding boxes of the lesion area in the narrow-band imaging detection result and the autofluorescence image detection result, respectively, i and C j represents the confidence C of the narrowband imaging detection result and the autofluorescence image detection result NBI , C AFI ;

[0076] The lesion detection result is defined as one that satisfies:

[0077] D fused ={(B i ,c fused,i)|i=1,2,...,n};

[0078] C fused =α·C NBI +β·C AFI ;

[0079] Among them, α and β are weight parameters, and α+β=1.

[0080] The beneficial effect achieved by the present invention is that a combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection is proposed. The method selects and extracts features with discriminative capabilities in narrow-band imaging images and autofluorescence imaging images based on deep learning, and integrates the advantages of different imaging technologies for lesion detection through decision fusion, thereby improving the speed and accuracy of lesion identification. At the same time, the method provided by the present invention can better adapt to different scenarios and lesion characteristics, has better generalization ability, improves the specificity and sensitivity of lesion detection, and is conducive to improving diagnostic accuracy and efficiency.

[0081] The present invention also provides a combined narrow band and fluorescence imaging system for bronchoscopic lesion detection, comprising:

[0082] An imaging module, used to respectively collect white light imaging images, narrow band imaging images and autofluorescence imaging images of the lesion area of ​​the bronchoscope, and respectively apply preset image enhancement methods to perform image preprocessing;

[0083] A processing module, used to use the white light imaging image as an input of a preset adversarial generative network model, and output a narrowband imaging generated image and an autofluorescence imaging generated image through the preset adversarial generative network model, wherein the preset adversarial generative network model includes a generator network and a discriminator network;

[0084] A detection module is used to obtain a lesion detection model based on the image generated by narrowband imaging and the image generated by autofluorescence imaging, and use the lesion detection model to perform lesion detection on the narrowband imaging image and the autofluorescence imaging image to obtain a narrowband imaging detection result and an autofluorescence image detection result, respectively, and fuse the narrowband imaging detection result and the autofluorescence image detection result by a preset weighted fusion method to obtain a lesion detection result.

[0085] The combined narrow-band and fluorescence imaging system for bronchoscopic lesion detection can implement the same steps as the combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection described in the above embodiment when executed, and can achieve the same technical effects. To avoid repetition, they will not be described here.

[0086] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0087] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0088] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0089] The embodiments of the present invention are described above in conjunction with the accompanying drawings. What is disclosed is only the preferred embodiment of the present invention. However, the present invention is not limited to the above-mentioned specific implementation manner. The above-mentioned specific implementation manner is only illustrative rather than restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms and equivalent changes without departing from the scope of protection of the purpose of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection, characterized in that: The following steps are involved: The white light imaging images, narrow band imaging images and autofluorescence imaging images of the lesion area under bronchoscopy are collected respectively, and the images are preprocessed by using the preset image enhancement methods respectively; Using the white light imaging image as an input of a preset adversarial generative network model, and outputting a narrowband imaging generated image and an autofluorescence imaging generated image through the preset adversarial generative network model, wherein the preset adversarial generative network model includes a generator network and a discriminator network; A lesion detection model is obtained based on the training of the narrow-band imaging generated image and the autofluorescence imaging generated image, and the lesion detection model is used to perform lesion detection on the narrow-band imaging image and the autofluorescence imaging image to obtain narrow-band imaging detection results and autofluorescence image detection results, respectively. The narrow-band imaging detection results and the autofluorescence image detection results are fused by a preset weighted fusion method to obtain a lesion detection result.

2. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection according to claim 1, characterized in that: The preset image enhancement method includes at least one of image quality enhancement and image color correction.

3. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection according to claim 1, characterized in that: In the preset adversarial generative network model, the probability that the generator network generates z types of images is defined as G(z), the probability that the discriminator network identifies a real image is defined as D(x), and the process of the generator network generating images and the discriminator network identifying images satisfies the value function V(G,D): in: In the above expression, is the initial data distribution, z is a random vector of Gaussian white noise, is the randomly initialized noise distribution, θ d is the network parameter of the discriminator network, θ g are the network parameters of the generator network, and m is the number of samples.

4. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection according to claim 1, characterized in that: The lesion detection model is based on the YOLOv8 model.

5. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection according to claim 4, characterized in that: In the step of training a lesion detection model based on the narrow-band imaging generated image and the autofluorescence imaging generated image, a classification loss function and a bounding box inspection function are set for the lesion detection model, and the classification loss function satisfies: Among them, x i is the predicted value of the current category, y i is the probability of the current category after the activation function, is the true value of the current category, and N is the number of samples; The bounding box loss function satisfies: Among them, b i is the actual bounding box, is the predicted bounding box.

6. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection according to claim 5, characterized in that: In the step of training a lesion detection model based on the narrow-band imaging generated image and the autofluorescence imaging generated image, the classification loss function and the bounding box loss function for the narrow-band imaging generated image are defined as follows: and The classification loss function and bounding box loss function for the image generated by the autofluorescence imaging are respectively and The joint loss function of the lesion detection model satisfies: Among them, α and β are the weight parameters of narrowband imaging and autofluorescence imaging, respectively, and λ is the weight parameter of weighted loss.

7. The combined narrow-band and fluorescence imaging method for bronchoscopic lesion detection according to claim 5, characterized in that: The preset weighted fusion method is specifically: The narrowband imaging detection result and the autofluorescence image detection result are defined as D NBI , D AFI , which satisfies: D NBI ={(B i ,C i )∣i=1,2,…,n}; D AFI ={(B j ,C j )∣j=1,2,…,m}; Among them, B i and B j represent the bounding boxes of the lesion area in the narrow-band imaging detection result and the autofluorescence image detection result, respectively, i and C j represents the confidence C of the narrowband imaging detection result and the autofluorescence image detection result NBI , C AFI ; The lesion detection result is defined as one that satisfies: D fused ={(B i ,C fused,i )∣i=1,2,…,n}; C fused =α·C NBI +β·C AFI ; Among them, α and β are weight parameters, and α+β=1.

8. A combined narrow-band and fluorescence imaging system for bronchoscopic lesion detection, characterized in that: include: An imaging module, used to respectively collect white light imaging images, narrow band imaging images and autofluorescence imaging images of the lesion area of ​​the bronchoscope, and respectively apply preset image enhancement methods to perform image preprocessing; A processing module, used to use the white light imaging image as an input of a preset adversarial generative network model, and output a narrowband imaging generated image and an autofluorescence imaging generated image through the preset adversarial generative network model, wherein the preset adversarial generative network model includes a generator network and a discriminator network; A detection module is used to obtain a lesion detection model based on the image generated by narrowband imaging and the image generated by autofluorescence imaging, and use the lesion detection model to perform lesion detection on the narrowband imaging image and the autofluorescence imaging image to obtain a narrowband imaging detection result and an autofluorescence image detection result, respectively, and fuse the narrowband imaging detection result and the autofluorescence image detection result by a preset weighted fusion method to obtain a lesion detection result.

Images