Operation area spectrum dynamic regulation and control method based on short wave infrared imaging
By dynamically adjusting the spectrum and illuminance of the surgical area using short-wave infrared imaging technology and machine learning algorithms, the problem of traditional surgical lighting equipment being unable to meet the imaging needs of different tissues is solved, achieving clear, accurate imaging and safe lighting of the surgical area.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING SHIJITAN HOSPITAL CAPITAL MEDICAL UNIVERSITY
- Filing Date
- 2026-03-26
- Publication Date
- 2026-04-28
AI Technical Summary
Traditional surgical lighting equipment cannot dynamically adjust the spectrum and illuminance according to the surgical location and lighting requirements, making it difficult to enhance the visible differences between tissues, thus increasing surgical risks and operational difficulty.
Using a short-wave infrared imaging method, images are acquired and preprocessed by a high-resolution camera. Edge detection and machine learning algorithms are used to identify the surgical area, and the spectrum and intensity of the LED supplementary lighting device are dynamically adjusted. Combined with temperature monitoring and automatic exposure control, precise lighting adjustment of the surgical area is achieved.
It improves the visibility of different types of tissues in the surgical area, provides clear and accurate visual information, adapts to complex surgical scenarios, and has versatility and practicality.
Smart Images
Figure CN121940913A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical surgical imaging and lighting control technology, and more specifically to a method for dynamic spectral control of the surgical area based on shortwave infrared imaging. Background Technology
[0002] Traditional surgical lighting equipment typically provides only uniform visible light illumination, which is insufficient to meet the imaging needs of different tissue types and cannot effectively enhance the visible differences between tissues. This hinders surgeons from accurately identifying various tissue structures within the surgical area, potentially increasing surgical risks and operational difficulty. While short-wave infrared imaging technology offers certain advantages in tissue imaging due to its unique spectral characteristics, there is currently no system or method capable of dynamically adjusting the spectrum and illuminance based on surgical location and lighting requirements to fully realize the potential of short-wave infrared imaging in the surgical field.
[0003] Therefore, providing a method for dynamic control of the surgical area spectrum based on short-wave infrared imaging, which intelligently controls the short-wave infrared camera and LED supplementary lighting equipment according to the preset surgical location and intraoperative lighting area requirements, and realizes dynamic adjustment of the illuminance and spectrum of the surgical area, is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0004] In view of this, the present invention provides a method for dynamic spectral control of surgical area based on shortwave infrared imaging, which can intelligently control the shortwave infrared camera and LED supplementary lighting equipment to achieve dynamic adjustment of illuminance and spectrum of surgical area and enhance the visible differences between different types of tissues.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for dynamic spectral control of surgical areas based on shortwave infrared imaging, comprising: Step 1: Acquire images of the surgical area and perform preprocessing to determine the current surgical location; Step 2: Preset the size and shape of the surgical location and lighting area; Step 3: Determine the illumination parameters for the current surgical location based on image analysis methods; Step 4: Compare the size and shape of the preset surgical location and lighting area with the lighting parameters, and adjust the exposure parameters of the short-wave infrared camera and the lighting intensity of the LED supplementary lighting device in real time; Step 5: After determining the supplementary light intensity, adjust the wavelength and spectral distribution of the light emitted by the LED supplementary light device, and compare it with the preset value until the surgical area image can be clearly displayed.
[0006] Preferably, the preprocessing of the surgical area image includes: Images of the surgical area were acquired using a high-resolution short-wave infrared camera, and the images were preprocessed to remove noise and enhance contrast. The surgical region features are extracted from the preprocessed image using edge detection and blood vessel enhancement algorithms to determine the current image location. The machine learning algorithm identifies key tissues in the current image location, and based on the identification results, it delineates areas that require focused lighting and imaging, and marks them on the image.
[0007] Preferably, the preset surgical position is determined according to the type of surgery, and a three-dimensional coordinate system is established on the operating table to represent the position of the surgical area with coordinate values; the length, width and height parameters of the lighting area are input to set the size, and the shape of the lighting area is designed using computer-aided design software.
[0008] Preferably, the image analysis method includes: Edge detection: Identifying the boundaries of surgical area images using the Canny edge detection algorithm; Region segmentation: Image segmentation is used to separate the surgical region from the background of the surgical region image; the segmented region is directly used as a reference for the illumination region; Feature extraction: Extract key features of the surgical area to determine the intensity and angle of illumination.
[0009] Preferably, the training process of the machine learning algorithm includes: A large amount of shortwave infrared image data of different types of tissues was collected and labeled to clarify the boundaries and types of each tissue, forming a labeled dataset. Feature extraction is performed on shortwave infrared image data in the labeled dataset; The machine learning algorithm, which is a convolutional neural network, is trained using a labeled dataset. By adjusting the parameters of the convolutional neural network, the model can accurately identify different types of tissue.
[0010] Preferably, the preprocessed surgical area image is input into a convolutional neural network to identify the tissue in the image in real time. Based on the identification results, the areas that require focused illumination and imaging are divided and marked on the surgical area image; at the same time, multi-frame consistency verification is performed.
[0011] Preferably, brightness monitoring points are set at marked locations on the image of the surgical area, and the brightness values of the monitoring points are acquired in real time by a short-wave infrared camera. The illumination parameters of the current surgical location are adjusted according to the brightness values. The size and shape of the preset surgical location and lighting area, along with the lighting parameters, are compared to obtain the brightness analysis results; Based on the brightness analysis results, the exposure parameters of the shortwave infrared camera are adjusted using an automatic exposure control algorithm; Based on the brightness analysis results and camera exposure parameters, the intensity of the LED fill light device is controlled, and the image brightness after fill light is monitored in real time.
[0012] Preferably, based on the illumination intensity of the LED illumination device, the presence of instrument reflection interference is determined by detecting the grayscale value of the image. When surgical instrument reflection interference is detected, the illumination intensity of the LED illumination in the interference area and the exposure time of the short-wave infrared camera are automatically reduced, and the brightness calculation of the interference area is masked by an image masking algorithm.
[0013] Preferably, it also includes temperature monitoring, with an infrared temperature sensor set in the area covered by the LED supplementary light to monitor the surface temperature of the surgical area tissue in real time. If the temperature is higher than a first preset value, the intensity of the LED supplementary light is forcibly reduced until the temperature drops below a safe value; if the temperature continues to exceed the first preset value, an audible and visual alarm is triggered.
[0014] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method for dynamic spectral control of surgical areas based on short-wave infrared imaging, which fully utilizes the advantages of short-wave infrared imaging technology and expands a new dimension of surgical illumination and imaging. Through precise dynamic control of spectrum and illuminance, the visible differences of different types of tissues in the surgical area are effectively improved, providing clearer and more accurate surgical visual information. It can also be flexibly adjusted according to different surgical locations and lighting requirements, adapting to various complex surgical scenarios, and has strong versatility and practicality. Attached Figure Description
[0015] 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0016] Figure 1 The method flowchart provided by the present invention. Detailed Implementation
[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] like Figure 1As shown, this embodiment of the invention discloses a method for dynamic spectral control of the surgical region based on shortwave infrared imaging, comprising: Step 1: Acquire images of the surgical area and perform preprocessing to determine the current image location; Step 2: Preset the size and shape of the surgical location and lighting area; Step 3: Determine the illumination parameters for the current surgical location based on image analysis methods; Step 4: Compare the size and shape of the preset surgical location and lighting area with the lighting parameters, and adjust the exposure parameters of the short-wave infrared camera and the lighting intensity of the LED supplementary lighting device in real time; Step 5: After determining the supplementary light intensity, adjust the wavelength and spectral distribution of the light emitted by the LED supplementary light device, and compare it with the preset value until the surgical area image can be clearly displayed.
[0019] Specifically, the preprocessing process for the surgical area image includes: Images of the surgical area were acquired using a high-resolution short-wave infrared camera, and the images were preprocessed to remove noise and enhance contrast. The surgical region features are extracted from the preprocessed image using edge detection and blood vessel enhancement algorithms to determine the current image location. The machine learning algorithm identifies key tissues in the current image location, and based on the identification results, it delineates areas that require focused lighting and imaging, and marks them on the image.
[0020] In a specific embodiment of the present invention, a shortwave infrared camera with a resolution of 1280×1024 is used to acquire initial images of the brain tumor region before the start of surgery. Gaussian filtering is used to remove image noise, and histogram equalization is applied to enhance image contrast. The Canny edge detection algorithm is then used to extract the boundary features between the tumor and brain tissue, combined with a vascular enhancement algorithm (such as Frangi filtering) to highlight the distribution of cerebral blood vessels, thus determining the location of the tumor region in the current image. The preprocessed image is input into a trained convolutional neural network (trained from over 5000 shortwave infrared labeled images of different types of brain tumors, achieving a tissue recognition accuracy of 98%), which identifies key tissues such as tumor tissue, normal brain tissue, and cerebral blood vessels, delineates the tumor region as the key illumination and imaging area, and marks it with a red box on the image.
[0021] Specifically, the preset surgical position is determined according to the type of surgery, and a three-dimensional coordinate system is established on the operating table to represent the position of the surgical area with coordinate values; the length, width and height parameters of the lighting area are input to set the size, and the shape of the lighting area is designed using computer-aided design software.
[0022] In one specific embodiment of the present invention, based on the marked tumor region location and the surgical operation range, the illumination area is determined to cover the tumor region and a 5mm radius of surrounding normal brain tissue. Depending on the surgeon's skill level, a three-dimensional coordinate system is established on the operating table, with the center coordinates of the tumor region set to (x0, y0, z0). The preset illumination area size is a cuboid with a length of 20mm, a width of 15mm, and a height of 8mm. AutoCAD software is used to design the illumination area into an arc-shaped cuboid that conforms to the curvature of the brain surface (to avoid blind spots).
[0023] Furthermore, the center coordinates of the tumor region are set as (x0, y0, z0) as the anchor point in three-dimensional space, which clarifies the absolute position of the tumor region in the operating table coordinate system, so that the lighting system can accurately lock the core area that needs to be focused on lighting and avoid the light from deviating to non-surgical areas.
[0024] If the tumor position shifts slightly (e.g., 2mm) during surgery, the system can calculate the deviation between the actual coordinates and the preset coordinates using real-time images, and dynamically adjust the spatial angle and irradiation range of the LED supplementary lighting device accordingly to ensure that the lighting always covers the target area.
[0025] By combining the preset lighting area size parameters (20mm×15mm×8mm), the spatial boundary of the lighting can be precisely defined through the coordinate system, so that the design of the curved cuboid can strictly fit the curvature of the brain surface, eliminate lighting dead angles, and at the same time avoid excessive light diffusion from interfering with the surrounding normal brain tissue.
[0026] Specifically, the image analysis method includes: Edge detection: Identifying the boundaries of surgical area images using the Canny edge detection algorithm; Region segmentation: Image segmentation is used to separate the surgical region from the background of the surgical region image; the segmented region is directly used as a reference for the illumination region; Feature extraction: Extract key features of the surgical area to determine the intensity and angle of illumination.
[0027] Specifically, the training process of the machine learning algorithm includes: A large amount of shortwave infrared image data of different types of tissues was collected and labeled to clarify the boundaries and types of each tissue, forming a labeled dataset. Feature extraction is performed on shortwave infrared image data in the labeled dataset; The machine learning algorithm, which is a convolutional neural network, is trained using a labeled dataset. By adjusting the parameters of the convolutional neural network, the model can accurately identify different types of tissue.
[0028] Specifically, the preprocessed surgical area image is input into a convolutional neural network to identify the tissue in the image in real time. Based on the identification results, the areas that require focused illumination and imaging are divided and marked on the surgical area image. At the same time, multi-frame consistency verification is performed.
[0029] Specifically, brightness monitoring points are set at marked locations on the image of the surgical area, and the brightness values of the monitoring points are acquired in real time by a short-wave infrared camera. The illumination parameters of the current surgical position are adjusted according to the brightness values. The size and shape of the preset surgical location and lighting area, along with the lighting parameters, are compared to obtain the brightness analysis results; Based on the brightness analysis results, the exposure parameters of the short-wave infrared camera are adjusted using an automatic exposure control algorithm; Based on the brightness analysis results and camera exposure parameters, the intensity of the LED fill light device is controlled, and the image brightness after fill light is monitored in real time.
[0030] Specifically, based on the illumination intensity of the LED lighting device, the presence of instrument reflection interference is determined by detecting the grayscale value of the image. When surgical instrument reflection interference is detected, the LED illumination intensity and short-wave infrared camera exposure time in the interference area are automatically reduced, and the brightness calculation of the interference area is masked through an image masking algorithm.
[0031] In one specific embodiment of the present invention, images during the surgical procedure are acquired in real time, and the location of the tumor region and the illumination parameters in the actual images are compared with preset values. If a slight shift (e.g., a shift of 2mm) is found in the actual tumor region due to the surgical operation, the exposure time of the short-wave infrared camera is adjusted (from 10ms to 8ms) and the gain (from 1.2 to 1.0) through an automatic exposure control algorithm, while the illumination intensity of the LED supplementary lighting device is controlled (from 300cd / m² to 280cd / m²) to ensure that the tumor boundary is clearly displayed in the image.
[0032] Specifically, it also includes temperature monitoring. An infrared temperature sensor is set in the area covered by LED supplementary lighting to monitor the surface temperature of the surgical area in real time. If the temperature is higher than the first preset value, the intensity of LED supplementary lighting will be forcibly reduced until the temperature drops below the safe value. If the temperature continues to exceed the first preset value, an audible and visual alarm will be triggered.
[0033] In a specific embodiment of the present invention, based on the spectral differences between brain tumor tissue and normal brain tissue in the short-wave infrared band (tumor tissue has higher reflectivity at 1300nm wavelength, while normal brain tissue has higher reflectivity at 1500nm wavelength), the wavelength of light emitted by the LED supplementary lighting device is adjusted to be mainly concentrated in the 1300-1500nm range, with a bimodal spectral distribution (peaks at 1300nm and 1500nm). The adjusted spectral distribution is compared with preset values (45% peak at 1300nm and 55% peak at 1500nm). If there is a deviation (e.g., the peak at 1300nm is only 40%), the LED driving circuit parameters are adjusted in real time until the spectral distribution meets the preset values. At this point, the tumor region in the image is clearly contrasted with the normal brain tissue, and the boundary is clearly distinguishable.
[0034] During the surgery, when forceps were used to grasp tumor tissue, the short-wave infrared camera detected a reflection (grayscale value exceeding 240, while the preset normal grayscale value range is 0-220) at the forceps location in the image. The system automatically reduced the LED supplementary light intensity in the reflective area (from 280 cd / m² to 200 cd / m²), shortened the short-wave infrared camera exposure time from 8 ms to 6 ms, and used an image masking algorithm to shield the brightness calculation of the reflective area, preventing the reflection from affecting the overall image observation.
[0035] Three infrared temperature sensors are evenly distributed within the LED illumination area to monitor the surface temperature of the brain tissue in real time. When one sensor detects a temperature of 38.5℃ (the first preset value is 38℃) after 30 minutes of surgery, the system immediately and forcibly reduces the LED illumination intensity to 220 cd / m², while continuing to monitor the temperature. After one minute, the temperature drops to 37.8℃, and the illumination intensity returns to normal. If the temperature remains above 38℃ for more than one minute, an audible and visual alarm is triggered to alert medical staff to check the equipment.
[0036] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.
[0037] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for dynamic spectral control of surgical areas based on shortwave infrared imaging, characterized in that, include: Step 1: Acquire images of the surgical area and perform preprocessing to determine the current surgical location; Step 2: Preset the size and shape of the surgical location and lighting area; Step 3: Determine the illumination parameters for the current surgical location based on image analysis methods; Step 4: Compare the size and shape of the preset surgical location and lighting area with the lighting parameters, and adjust the exposure parameters of the short-wave infrared camera and the lighting intensity of the LED supplementary lighting device in real time; Step 5: After determining the supplementary light intensity, adjust the wavelength and spectral distribution of the light emitted by the LED supplementary light device, and compare it with the preset value until the surgical area image can be clearly displayed.
2. The method for dynamic spectral control of the surgical area based on shortwave infrared imaging according to claim 1, characterized in that, The preprocessing process for the surgical area image includes: Images of the surgical area were acquired using a high-resolution short-wave infrared camera, and the images were preprocessed to remove noise and enhance contrast. The surgical region features are extracted from the preprocessed image using edge detection and blood vessel enhancement algorithms to determine the current image location. The machine learning algorithm identifies key tissues in the current image location, and based on the identification results, it delineates areas that require focused lighting and imaging, and marks them on the image.
3. The method for dynamic spectral control of the surgical area based on shortwave infrared imaging according to claim 1, characterized in that, The preset surgical position is determined according to the type of surgery, and a three-dimensional coordinate system is established on the operating table to represent the position of the surgical area using coordinate values. Input the length, width, and height parameters of the lighting area to set its size, and use computer-aided design software to design the shape of the lighting area.
4. The method for dynamic spectral control of the surgical area based on shortwave infrared imaging according to claim 1, characterized in that, The image analysis method includes: Edge detection: Identifying the boundaries of surgical area images using the Canny edge detection algorithm; Region segmentation: Image segmentation is used to separate the surgical region from the background of the surgical region image; the segmented region is directly used as a reference for the illumination region; Feature extraction: Extract key features of the surgical area to determine the intensity and angle of illumination.
5. The method for dynamic spectral control of the surgical area based on shortwave infrared imaging according to claim 2, characterized in that, The training process of the machine learning algorithm includes: A large amount of shortwave infrared image data of different types of tissues was collected and labeled to clarify the boundaries and types of each tissue, forming a labeled dataset. Feature extraction is performed on shortwave infrared image data in the labeled dataset; The machine learning algorithm, which is a convolutional neural network, is trained using a labeled dataset. By adjusting the parameters of the convolutional neural network, the model can accurately identify different types of tissue.
6. The method for dynamic spectral control of the surgical area based on shortwave infrared imaging according to claim 5, characterized in that, The preprocessed surgical area image is input into a convolutional neural network to identify the tissue in the image in real time. Based on the identification results, the areas that need to be highlighted and imaged are divided and marked on the surgical area image. At the same time, multi-frame consistency verification is performed.
7. The method for dynamic spectral control of surgical area based on shortwave infrared imaging according to claim 6, characterized in that, Brightness monitoring points are set at marked locations on the image of the surgical area. The brightness values of the monitoring points are acquired in real time by a short-wave infrared camera, and the illumination parameters of the current surgical position are adjusted according to the brightness values. The size and shape of the preset surgical location and lighting area, along with the lighting parameters, are compared to obtain the brightness analysis results; Based on the brightness analysis results, the exposure parameters of the short-wave infrared camera are adjusted using an automatic exposure control algorithm; Based on the brightness analysis results and camera exposure parameters, the intensity of the LED fill light device is controlled, and the image brightness after fill light is monitored in real time.
8. The method for dynamic spectral control of surgical areas based on shortwave infrared imaging according to claim 7, characterized in that, Based on the illumination intensity of the LED lighting device, the presence of instrument reflection interference is determined by detecting the grayscale value of the image. When surgical instrument reflection interference is detected, the LED illumination intensity and short-wave infrared camera exposure time in the interference area are automatically reduced, and the brightness calculation of the interference area is masked by an image masking algorithm.
9. The method for dynamic spectral control of surgical areas based on shortwave infrared imaging according to claim 1, characterized in that, It also includes temperature monitoring, with an infrared temperature sensor set up in the LED supplementary lighting coverage area to monitor the surface temperature of the surgical area tissue in real time. If the temperature is higher than the first preset value, the intensity of the LED supplementary lighting will be forcibly reduced until the temperature drops below the safe value; if the temperature continues to exceed the first preset value, an audible and visual alarm will be triggered.