Method and system for visualizing weak fluorescence in surgery, software program

Abstract

The present disclosure relates to methods and systems, software programs for visualizing weak fluorescence in surgery. The method comprises the following steps: emitting excitation light (12) from an excitation light source (6) to an operating area (10) containing a weak fluorescent light source (16) and emitting white light from a white light source to the operating area (10); capturing one or more fluorescence images and one or more white light images of the operating area (10) at the wavelength range of fluorescence emitted by the weak fluorescence source (16); performing image processing on the one or more fluorescence images and creating one or more false color fluorescence images from the one or more fluorescence images wherein the contrast between the fluorescence emitted by the weak fluorescence source (16) and the background is increased; and creating one or more composite images (50) by superimposing the one or more false color fluorescence images on the one or more white light images.

Description

Technical Field

[0001] The present disclosure relates to a method, system, and software program for visualizing weak fluorescence in surgery, particularly open surgery. Background Art

[0002] Fluorescence-guided surgery using fluorescent dyes is known in the art. A fluorescent dye, such as indocyanine green (ICG), is injected into the tissue to be examined or operated on, and once the fluorescent dye has penetrated the tissue, a fluorescent excitation light is shone on the dye-soaked tissue, causing the fluorescent dye to emit fluorescent emission light (also referred to as fluorescence).

[0003] The basic principle of induced fluorescence is that the molecules that make up the fluorescent dye have an excitation wavelength spectrum. Under the excitation wavelength spectrum, these molecules are able to capture the excitation light, absorb the energy of the captured excitation light, and transform into an excited state. The excited energy state of the fluorescent molecule is unstable. After a short time, the excess energy is released from the molecule in the form of de-excitation light (also called fluorescence), allowing the fluorescent molecule to return to its basic energy state. Fluorescence has slightly less energy than the excitation light, which means that the excitation spectrum of the fluorescent dye is slightly shifted to a smaller wavelength relative to its fluorescence spectrum.

[0004] Therefore, the excitation light used to stimulate fluorescence in the fluorochrome has a wavelength that is slightly shifted toward a smaller wavelength than the fluorescence to be observed. This is useful because the amount of excitation light reflected from fluorochrome-impregnated tissue can be several orders of magnitude more intense than the resulting fluorescence itself, and in the presence of all the reflected excitation light, it would be impossible to detect fluorescence. This slight wavelength shift is exploited by using a wavelength filter configured to pass light at the fluorescence wavelength, thereby preventing reflected excitation light from entering the camera.

[0005] Fluorescent dyes are not the only source of fluorescence in body tissues. Other sources include fluorescent probes used to label certain tissue types (e.g., cancerous tissue) and autofluorescent tissues (e.g., the parathyroid glands). The parathyroid glands contain naturally occurring fluorescent molecules that emit fluorescence in the near-infrared spectrum when excited by excitation light of the appropriate wavelength. However, the fluorescence signal emitted by autofluorescence is several orders of magnitude weaker than that of the aforementioned fluorescent dyes, i.e., orders of magnitude more easily masked by excitation light or other sources (such as ambient light with components in the spectral region of autofluorescence).

[0006] The autofluorescence of the parathyroid gland is excited in the wavelength range of about 785 nm and emits fluorescence at about 820 to 830 nm. This overlaps with, for example, ICG, which emits in the range of 800 to 830 nm and is excited around 780 nm.

[0007] There is also autofluorescence of colon tissue and colon cancer in the UV and visible spectra, as described, for example, in Li BH, Xie SS, “Autofluorescence excitation-emission matrices for diagnosis of colonic cancer,” World Journal of Gastroenterology, July 7, 2005, 11(25):3931-4 or Bhaskar Banerjee MD et al., “Tryptophan autofluorescence imaging of neoplasms of the human colon,” J Biomed Opt 17(1)016003 (Feb 1, 2012).

[0008] The same applies to fluorescent probes used to label tissue types such as cancerous tissue. Such probes and labels produce a weak fluorescence signal similar in size to the autofluorescence of the parathyroid gland, i.e., several orders of magnitude smaller than that of the fluorescent dyes commonly used in fluorescence-guided surgery.

[0009] For these reasons, fluorescence-guided surgery based on weak fluorescence sources, such as the autofluorescence of the parathyroid gland or fluorescent probes that image fluorescence at similarly weak intensities as the autofluorescence, is currently unavailable. Summary of the Invention

[0010] The object of the present invention is to provide a means for performing fluorescence-guided surgery (especially open surgery) using a weak fluorescence source.

[0011] This object is achieved by a method for visualizing weak fluorescence during surgery, particularly open surgery, comprising the following steps:

[0012] - emitting excitation light from an excitation light source to an operating area containing a weak fluorescence source, and emitting white light from a white light source to the operating area, in particular, the weak fluorescence source is parathyroid tissue exhibiting autofluorescence and / or a fluorescent probe exhibiting fluorescence with an intensity similar to the autofluorescence;

[0013] - capturing one or more fluorescence images of the operating area at the wavelength range of the fluorescence emitted by the weak fluorescence source, and one or more white light images;

[0014] - performing image processing on the one or more fluorescence images and creating one or more false colour fluorescence images from the one or more fluorescence images, wherein the contrast between the fluorescence emitted by the weak fluorescence source and the background is increased; and

[0015] - creating one or more composite images by superimposing the one or more false-color fluorescence images on the one or more white-light images.

[0016] This method is applicable to both still images and image sequences, such as video images. Increasing the contrast between the fluorescence emitted by a weak fluorescent source and the background helps to make the weak fluorescent signal visible against the background, thereby masking the background.

[0017] Preferably, the fluorescence image is captured using an image sensor with a high gain setting. Due to the high gain setting, the sensor will saturate in portions of the image with medium to high light intensities, while darker areas will be enhanced and dominate the image's dynamic range. However, weak fluorescence signals will occupy the lower portion of the dynamic spectrum.

[0018] In an embodiment, the contrast between the fluorescent light emitted by the weak fluorescent source and the background is increased by applying inverse gamma correction. Gamma correction is a nonlinear transformation of brightness values ​​in video or still image systems. This nonlinear transformation darkens relatively dark areas in the image, while relatively bright areas of the image maintain their relatively bright intensity.

[0019] However, in the present case, the contrast of the dark parts of the fluorescence image needs to be enhanced, so it is useful to use the inverse exponential of the logarithmic function, hence the name "inverse gamma correction." Inverse gamma correction can also be called decoded gamma or gamma expansion.

[0020] Depending on the setup and the nature of the fluorescence, in embodiments, white light images and fluorescence images are captured simultaneously or alternately. Simultaneous capture of white light and fluorescence images is feasible if the fluorescence signal is outside the visible light spectrum captured in the white light image and an image capture device is available that includes an image sensor that is sensitive in white light on the one hand and in the region of the fluorescence signal on the other hand.

[0021] On the other hand, if the fluorescence has a wavelength that falls within the white light spectrum, it is not possible to capture both a white light image and a fluorescence image simultaneously because the white light illumination will drown out the fluorescence signal. In this case, it is possible to alternate between white light illumination and fluorescence illumination, and to use pairs of a white light image and an immediately subsequent fluorescence image, or pairs of a fluorescence image and an immediately subsequent white light image, to provide a composite image.

[0022] Particularly if the sequence of white-light images does not experience rapid motion, or if such motion is identified and the positions of the intermittent fluorescence images are adjusted to conform to the motion so as to be superimposed on the correct light point in each of the white-light images, it is also possible to occasionally insert fluorescence image captures between the otherwise uninterrupted sequence of white-light images and use these captured fluorescence images to superimpose on each of the white-light images in the sequence. This intermittent capture of fluorescence images has the added benefit of less flicker in the illumination of the operating area. Motion can be counteracted by the methods described with respect to realignment of the region of interest.

[0023] In an embodiment, the image processing performed on the one or more fluorescence images includes at least one of masking specular reflections, noise suppression, and normalization. Specular reflections will appear as saturated regions of the image in the white light image, the fluorescence image, or both. Such saturated regions can be excluded from further image processing of the fluorescence image and masked as background, for example, by setting the brightness value of affected pixels of the fluorescence image to zero (i.e., black).

[0024] Noise suppression is a known technique in image processing and is beneficial in the context of the present invention because the fluorescence signal from a weak fluorescent source is very weak. This weak signal has a low signal-to-noise ratio, making it difficult to select the weak fluorescent signal from the noise depending on the environment. Known noise suppression techniques are to filter adjacent pixels and take the average of the brightness of adjacent pixels or pixel clusters. Known removal algorithms reduce or remove the visibility of noise by smoothing the entire image, leaving areas near contrast boundaries. These methods may blur fine, low-contrast details, but in the context of the present invention (where the goal is to identify structures and tissues that exhibit weak fluorescence), this compromise is acceptable in order to receive a clear signal to first identify the structure or tissue.

[0025] Normalization performed during image processing of one or more fluorescence images is used to create a false-color image representing the fluorescence image. Normalization focuses the dynamic range of the fluorescence image on a portion of interest, which in the case of weak fluorescence sources is a low-intensity portion of the fluorescence image. In an embodiment, normalization includes percentile normalization, specifically by cutting out the darkest 1% to 2% and the brightest 1% to 2% of pixels in the fluorescence image. This removes saturated regions and background regions containing no signal but only noise. Alternatively or in addition, normalization can include selecting a bottom portion of the luminance spectrum and spreading the luminance information contained in the bottom portion across the full luminance spectrum, wherein the bottom portion of the luminance spectrum specifically includes 0% to 20%, specifically 0% to 10%, and particularly 0% to 5%, of the luminance spectrum of the input fluorescence image. The range of the selected region is adapted to the observed signal distribution and the actual gain setting in the specific case of the surgical procedure.

[0026] The threshold (below which weak fluorescence signals are detected and above which they should be discarded) can also be adjusted automatically or manually by the surgeon or the surgeon's assistant to achieve optimal visibility of weak fluorescence signals in the composite image. Automatic adjustment of the threshold can be achieved by automatic analysis of the signal amplitude distribution. For example, the normalization algorithm can start with a threshold of 30% of the maximum value of the brightness of the original fluorescence image and select the isolated point in the fluorescence image with the highest intensity below this threshold, and select a new threshold representing a brightness value just above this relative brightest point. By selecting isolated points, the algorithm is prevented from simply reaching the edge of the saturated region that was masked out using the initial 30% threshold. However, if there are other isolated points that entered saturation using the lower threshold, the threshold is increased again until all isolated points whose centers have not entered saturation are shown, i.e., the brightness of the corresponding point's center is greater than the threshold.

[0027] In an embodiment, the image processing comprises creating a virtual region of interest in the white light image and / or the fluorescence image, wherein the fluorescence image outside the region of interest is cut away.

[0028] By creating a virtual region of interest and cutting out the fluorescence image outside the region of interest, fluorescence will be visible only within the region of interest, while outside the region of interest, the composite image will consist only of the white light image. This makes it easier for surgeons to identify weak fluorescence sources because fluorescence from outside the region of interest does not interfere with fluorescence from weak fluorescence sources. Specifically, the intensity of the false color fluorescence image is normalized according to the intensity distribution within the region of interest. Typically, fluorescence from weak fluorescence sources can be easily masked by the excitation light source or other sources (such as ambient light with components in the same spectral region). By limiting the visibility of light in this spectral region to the region of interest, these sources are greatly suppressed, resulting in much greater contrast for weak fluorescence sources.

[0029] Furthermore, the brightness of fluorescence from weak fluorescent sources within the region of interest can be used to adjust the gain of the imaging sensor, which might otherwise be set too low due to specular reflections that may be present outside the region of interest. Such bright spots outside the region of interest are ignored, thereby focusing control of the sensor gain and / or image processing on the region of interest containing the weak fluorescent signal being sought.

[0030] The image can be a still image or a video. By using video for both the white light image and the fluorescence image, a video feed is provided in which regions of interest containing parathyroid tissue or fluorescent probes are highlighted by providing fluorescence information over the white light video feed.

[0031] Preferably, the region of interest is created by using manual control elements to control the positioning and / or shape of the region of interest. The manual control elements can be buttons, switches and / or touch screen displays. With the help of these manual control elements, the surgeon can select the region of interest. For example, the surgeon can view a white light image of the operating area and select the region of interest based on the white light image. The surgeon can alternatively view a composite image consisting of a superposition of a false color fluorescence image on the white light image and select the region of interest based on the composite image, or the surgeon can select the region of interest based solely on the fluorescence image. The region of interest can have any shape and form, such as a circular, oval, rectangular shape or a shape that conforms to the outline of an organ or tissue visible in the white light image.

[0032] The region of interest is preferably automatically repositioned after the movement so that the region of interest is realigned relative to the weak fluorescent light source in one or more composite images. The movement can be movement of the patient or movement of the image capture device used to capture the images. Such movement can occur unexpectedly during surgery, for example due to slight shifting of the patient or shaking of the image capture device. However, such movement can change the position of the region of interest relative to the weak fluorescent light source. This is undesirable because it reduces the visibility of the fluorescence in the composite images and can hinder the surgeon during surgery. To counteract this movement, the region of interest is automatically realigned with the weak fluorescent light source. This way, the movement does not interfere with the surgery.

[0033] According to another embodiment, the region of interest is automatically repositioned using an algorithm that is trained to recognize the shape of an organ containing or consisting of a weak fluorescent source, wherein the region of interest is realigned relative to the organ. The algorithm may be an artificial neural network that has been trained using white light image data of organs (e.g., parathyroid glands) to reliably detect these organs in new white light images. The region of interest is then realigned relative to the organ to offset some of the motion.

[0034] According to another embodiment, a region of interest (ROI) is automatically repositioned by performing image analysis on a white light image to identify an image region dominated by the red wavelength spectrum. The region of interest is then realigned relative to or set to the image region dominated by the red wavelength spectrum. During surgery, the patient is typically covered with a blanket or sheet, with only a small opening surrounding the surgical area. An incision made during open surgery is typically visible in a white light image as a red area compared to the surrounding skin and the blanket or sheet. This allows the incision to be identified through image analysis of the white light image, thereby identifying the incision as a region dominated by the red wavelength spectrum. The region of interest can then be realigned so that it is centered over the region dominated by the red wavelength spectrum, or it can be set equal to the region dominated by the red wavelength spectrum. This is an easy way to compensate for movement during surgery and mask any light from the relevant wavelength spectrum outside the incision.

[0035] According to an embodiment, high-intensity information in a white-light image is used to detect specular reflections, wherein regions containing specular reflections are removed from the region of interest. These regions also reflect fluorescence, particularly near-infrared light, which results in unwanted high-intensity spots in the fluorescence image. Therefore, these regions are removed from the region of interest, thereby removing the fluorescence information in these regions from the composite image.

[0036] The object is also achieved by a system for visualizing weak fluorescence during surgery, in particular during open surgery, comprising:

[0037] - a control device comprising an image processing unit;

[0038] - a light source device configured to generate excitation light and white light; and

[0039] - an image capture device configured to capture a fluorescence image and a white light image,

[0040] Therein, the control device is configured to control the operation of the light source device and the image capture device and to perform the method steps of the method according to any of the previously described embodiments.

[0041] The image capture device may be a camera. The image capture device may include an image capture unit for white light and a separate image capture unit for fluorescence. The control device may be a controller, in particular a computer. The control device is configured to control the light source device to emit excitation light and white light toward the operating area, and to control the image capture device to capture one or more fluorescence images and one or more white light images. The image processing unit performs image processing, in particular, creates a region of interest and / or creates a false-color fluorescence image and / or creates a composite image.

[0042] In particular, the light source device includes an excitation light source and a white light source. The white light source is preferably a light source that does not emit within the spectral range of the fluorescence signal of a weak fluorescence source, such as an LED or laser source. Specifically, the system includes an illumination and image capture device configured to emit white light and excitation light and capture white light images and fluorescence images. Thus, light emission and image capture are combined in a single device.

[0043] The object is furthermore achieved by a software program which is configured to perform the method steps of the method according to any of the previously described embodiments when run on a control device of a system according to any of the previously described embodiments.

[0044] The system and software program for visualizing weak fluorescence during surgery embody the same advantages, features, and characteristics as the method described above. Features mentioned with respect to the method are expressly applicable to the system and vice versa, and to the software program that controls and receives input from the system components and the image processing portion of the method.

[0045] Other features of the invention will become apparent from the description of the embodiments according to the invention, the claims and the included drawings.Embodiments according to the invention may satisfy individual features or a combination of several features. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The invention is described below based on exemplary embodiments without limiting the overall scope of the invention, with explicit reference to the accompanying drawings which disclose all details according to the invention which are not explained in greater detail in the text. The drawings show:

[0047] Figure 1 is a schematic, simplified representation of a system for visualizing weak fluorescence during surgery;

[0048] Figure 2 is a schematic, simplified representation of another system for visualizing weak fluorescence during surgery;

[0049] Figure 3 is a schematic, simplified representation of a composite image of the operating area consisting of a superposition of a white light image and a fluorescence image; and

[0050] Figure 4 is a schematically simplified representation of the method steps of a method for visualizing weak fluorescence during surgery.

[0051] Reference Signs List

[0052] 1 system

[0053] 3 Control device

[0054] 4 image processing units

[0055] 5. Light source device

[0056] 6 Excitation light source

[0057] 7 White light source

[0058] 8 Image capture device

[0059] 10 operating areas

[0060] 12 Excitation light

[0061] 14 Fluorescence

[0062] 20 Areas of Interest

[0063] 30 autofluorescent tissues

[0064] 31 Autofluorescent Organs

[0065] 32 fluorescence signals

[0066] 40 organizations

[0067] 42 surgical tools

[0068] 50 composite images

[0069] 60 optical cable

[0070] 61 data cable

[0071] 62 control elements

[0072] 63 displays

[0073] 64 lighting and image capture devices

[0074] 101-104 Method Steps

[0075] In the drawings, elements of the same or similar type or respectively corresponding parts are provided with the same reference numerals to avoid the need to reintroduce items. DETAILED DESCRIPTION

[0076] Figure 1A schematically simplified representation of a system 1 for visualizing weak fluorescence during surgery is shown. The system 1 includes a control device 3 having an image processing unit 4, which is implemented as image processing software on a general-purpose computer or a dedicated image processor configured to process images. The control device 3 is connected to a light source device 5 and an image capture device 8 and controls the light source device 5 and the image capture device 8. The light source device 5 includes an excitation light source 6 and a white light source 7 to illuminate an operating area 10. The excitation light 12 emitted by the light source device 5 stimulates autofluorescent tissue 30 in the operating area 10 or a fluorescent probe that exhibits fluorescence with an intensity similar to that of autofluorescence. Autofluorescent tissue 30 (e.g., parathyroid tissue) will emit fluorescence 14 in a specific wavelength range. This wavelength range will typically be in the near-infrared spectrum and will typically not overlap with the white illumination light in order to separate the white light from the fluorescence 14.

[0077] The image capture device 8 is configured to capture images of this wavelength range as well as white light images. Typically, the image capture device 8 includes an image capture unit for white light (e.g., an RGB unit) and an image capture unit for fluorescence 14. The images captured by the image capture device 8 are transmitted to the image processing unit 4 of the control device 3. The image processing unit 4 performs image processing to enhance the visibility of specific features of interest in the operating area 10.

[0078] Figure 2 Another embodiment of a system 1 for visualizing weak fluorescence during open surgery is shown. In this embodiment, an image capture device 8 and an illumination device are combined into a single illumination and image capture device 64. This illumination and image capture device 64 is connected to a light source device 5 via an optical cable 60 to illuminate the parathyroid operative region 10. The operative region 10 includes a weak fluorescence source 16, such as an autofluorescent parathyroid gland. The image capture device 8 captures white light and fluorescence images of the operative region 10 and transmits the image information to a control device 3, such as an industrial computer, via a data cable 61. The images can be displayed on a monitor 63. The surgeon can use manual control elements 62 (e.g., a keyboard and mouse) to select a region of interest 20 in the image and / or perform image processing. Image processing can include increasing the contrast between the signal from the weak fluorescence source 16 and the background, applying inverse gamma correction, masking specular reflections, applying noise suppression, and / or applying normalization.

[0079] Figure 3 A schematically simplified representation of a composite image 50 created by the image processing unit 4 is shown. The composite image 50 can be displayed on a display 63 and consists of a superposition of a false-color fluorescence image on a white light image. The operating area 10 shown in the composite image 50 includes tissue 40 and several surgical tools 42 used during open surgery. Figure 3Medium fluorescence is indicated by a dotted line. In the center of the image, an autofluorescent organ 31 (e.g., the parathyroid gland) is shown, wherein a specific portion emits a specific fluorescence signal 32 with a higher intensity than the fluorescence of other autofluorescent tissues 30. Furthermore, the intensity of the fluorescence signal 32 is much smaller than the intensity of the excitation light or other light sources (such as ambient light having a component in the spectral region of autofluorescence).

[0080] To enhance the contrast in the composite image 50, the surgeon may create or select a region of interest 20 in the image. The region of interest 20 is Figure 3 50. The region of interest 20 is shown as a dotted circle, however, it may have a different form and / or size. Outside the region of interest, only the white light image is shown in the composite image 50, while the fluorescence image is cut out. The surgeon can select and place the region of interest 20 using a manual control element 62 or directly select the region of interest 20 on a display 63 showing the composite image 50.

[0081] If the patient or image capture device 8 moves during surgery, the region of interest 20 may no longer be aligned with the organ 31. To counteract such movement, the system 1 is configured to track the movement and realign the region of interest 20 accordingly. To track this movement, the image processing unit 4 can perform shape detection on the high-intensity information in the white light image to detect specular reflections. Specular reflections typically cause intensity peaks in the white light image, which are used to identify certain areas in the image and help correct the positioning of the region of interest 20. Specular reflections can also be detected by the image capture unit for fluorescence 14. In addition, the image processing unit 4 can utilize an algorithm trained to recognize the shape of the organ 31 (e.g., the parathyroid gland) to correct the positioning of the region of interest 20. Furthermore, the image processing unit 4 can locate the region of interest 20 based on color in the white light image. Specifically, because tissue in open surgery typically appears red in white light images, the region of interest 20 is positioned in an area dominated by the red wavelength range.

[0082] Figure 4A schematic, simplified representation of the method steps for visualizing weak fluorescence during surgery is shown. In step 101, excitation light 12 is emitted from an excitation light source 6 onto an operating area 10 containing a weak fluorescence source, such as parathyroid tissue exhibiting autofluorescence. At step 102, one or more images of the operating area 10 are captured at a wavelength range that covers the wavelength range of the weak fluorescence source, as well as one or more white-light images. In step 103, image processing is performed on the captured images. Image processing may include creating a virtual region of interest 20 in the white-light image and / or the fluorescence image. Any image information in the fluorescence image outside of the region of interest 20 is removed. Image processing may also include increasing the contrast between the weak fluorescence source 16 and the background, applying inverse gamma correction, masking specular reflections, applying noise suppression, and / or applying normalization. In step 104, a false-color fluorescence image, visible in the visible spectrum, is created from the fluorescence image. Then, in step 105, a composite image 50 is created by superimposing the false-color fluorescence image on the white-light image. In this composite image 50 , false-colored fluorescence is visible only within the region of interest 20 .

[0083] All named features, including those taken from the drawings alone and disclosed in combination with other features, are considered to be essential features of the invention, whether considered alone or in combination. Embodiments according to the invention can be achieved by individual features or by a combination of several features. Features combined with the word "particularly" or "especially" are to be considered as preferred embodiments.

Claims

1. A method for visualizing weak fluorescence during surgery, particularly open surgery, comprising the following steps: - emitting excitation light (12) from an excitation light source (6) to an operating area (10) containing a weak fluorescence source (16), and emitting white light from a white light source to the operating area (10), wherein the weak fluorescence source (16) is in particular parathyroid tissue exhibiting autofluorescence and / or a fluorescent probe exhibiting fluorescence having an intensity similar to the autofluorescence; - capturing one or more fluorescence images and one or more white light images of the operating area (10) at the wavelength range of the fluorescence emitted by the weak fluorescence source (16); - performing image processing on the one or more fluorescence images and creating one or more false-color fluorescence images from the one or more fluorescence images, wherein the contrast between the fluorescence emitted by the weak fluorescence source (16) and the background is increased; as well as - creating one or more composite images (50) by superimposing the one or more false-color fluorescence images on the one or more white-light images.

2. The method according to claim 1, wherein The fluorescence images were captured using an image sensor at a high gain setting.

3. The method according to claim 1 or 2, wherein: The contrast between the fluorescent light (14) emitted by the weak fluorescent light source (16) and the background is increased by applying an inverse gamma correction.

4. The method according to any one of claims 1 to 3, wherein: The white light image and the fluorescence image are captured simultaneously or alternately.

5. The method according to any one of claims 1 to 4, wherein: The image processing performed on the one or more fluorescence images includes at least one of masking specular reflections, noise suppression, and normalization.

6. The method according to claim 5, wherein: The normalization includes percentile correction, in particular cutting off the lowest 1% to 2% of the pixels of the fluorescence image and the highest 1% to 2% of the pixels of the fluorescence image and / or selecting the bottom part of the brightness spectrum and spreading the brightness information contained in the bottom part over the full brightness spectrum, wherein, in particular, the bottom part of the brightness spectrum includes 0% to 20%, in particular, 0% to 10%, in particular, 0% to 5% of the brightness spectrum of the input fluorescence image.

7. The method according to any one of claims 1 to 6, wherein: The image processing includes creating a virtual region of interest (20) in the white light image and / or the fluorescence image, wherein the fluorescence image outside the region of interest (20) is cut away.

8. The method according to claim 7, wherein: The region of interest (20) is created by using a manual control element (62) to control the location and / or shape of the region of interest (20).

9. The method according to claim 7 or 8, wherein The region of interest (20) is automatically repositioned after the movement so that the region of interest (20) is realigned relative to the weak fluorescent light source (16) in the composite image (50).

10. The method according to claim 9, wherein: The region of interest (20) is automatically repositioned by an algorithm that is trained to recognize the shape of an organ (31) that includes or consists of the weak fluorescent source (16), wherein the region of interest (20) is realigned relative to the organ (31).

11. The method according to claim 9 or 10, wherein: The region of interest (20) is automatically repositioned by performing image analysis on the white light image to identify an image region dominated by a red wavelength spectrum, wherein the region of interest (20) is realigned or positioned relative to the image region dominated by a red wavelength spectrum as the image region dominated by a red wavelength spectrum.

12. The method according to any one of claims 7 to 11, wherein: High intensity information in the white light image is used to detect specular reflections, wherein regions including the specular reflections are deleted from the region of interest (20).

13. A system for visualizing weak fluorescence during surgery, particularly during open surgery, the system comprising: - a control device comprising an image processing unit; - a light source device configured to generate excitation light and white light; as well as - an image capture device configured to capture a fluorescence image and a white light image, Wherein the control device is configured to control the operation of the light source device and the image capture device and to perform the method steps of the method according to any one of claims 1 to 12.

14. A software program configured to perform the method steps of the method according to any one of claims 1 to 12 when run on the control device of the system according to claim 13.

Images