Method and device for examining fluorescence image of mammalian tissue and medical treatment device

Abstract

Disclosed herein is an improved examination method for examining a fluorescence image taken of mammalian tissue treated with a fluorescent agent to be photosensitive and irradiated with excitation light, the fluorescence image (FI) comprising an array of pixels having respective fluorescence signal values, the fluorescence signal values include fluorescence signal values representing a first quantity (N1) of mammalian tissue in the fluorescence image. The improved inspection method comprises: determining (S6) a reference value (IR) such that a second number (N2) of fluorescence signal values contained in a first number of fluorescence signal values is less than or equal to said reference value and the remainder of the first number of fluorescence signal values exceeds said reference value, said second number being a predetermined fraction of said first number; determining (S7) an average fluorescence signal value ([mu] R) and a standard deviation ([sigma] R) of fluorescence signal values contained in the second number of fluorescence signal values; performing (S8) image segmentation in the fluorescence image (FI) to distinguish between a target region (TR) representing a portion identified as tumor tissue in the mammalian tissue (MT) and a reference region (RR) representing a remaining portion in the mammalian tissue (MT), including a determination for each pixel (p) and a determination for each pixel (p); determining that the pixel belongs to the target region (TR) if the signal-to-noise ratio CNR (p) of the pixel exceeds a predetermined threshold value (TCNR), otherwise determining that the pixel belongs to the reference region (RR), where the signal-to-noise ratio CNR (p) of the pixel is as defined in (I): (I) # imgabs0 # where FI (p) is the fluorescence signal value of the pixel (p), c is a predetermined constant; at least one contour (B) of the target area (TR) is identified (S9).

Description

Background Art

[0001] The present application relates to an image processing method.

[0002] The present application relates to an image processing device.

[0003] The present application also relates to a medical treatment device comprising an examination device.

[0004] For most types of solid cancers, treatment involves radical surgical resection of all tumor tissue. However, distinguishing normal tissue from tumor tissue during surgery remains difficult. Therefore, it is not uncommon for pathological evaluation to reveal positive tumor margins two to five days after surgery. According to the literature, the rate of positive tumor margins (TPM) varies from 10% to 35% depending on the tumor type. See Orosco, RK, et al., "Positive surgical margins in the 10 most common solid cancers," Sci. Rep. 8, 56-86 (2018). https: / / doi.org / 10.1038 / s41598-018-23403-5. If tumor tissue is present at or near the margins of resected tissue, the risk of local recurrence and distant metastasis increases, implying a decrease in survival. Therefore, positive tumor margins require additional treatment, such as reoperation, radiation therapy, and / or systemic therapy. Unfortunately, this is associated with increased morbidity and a greater psychological burden for the patient. Therefore, being able to correctly visualize tumor tissue during surgery is crucial. However, current optical technology and the visual and tactile information obtained by the physician are insufficient to fully determine tumor boundaries. Therefore, new technologies that can visualize tumors in real time are being investigated with the aim of reducing the number of positive tumor margins, thereby reducing additional treatment and morbidity.

[0005] Fluorescence molecular imaging (FMI) is an imaging technique gaining increasing attention due to its ability to visualize tumors in real time, both within the patient's body (i.e., in vivo) and immediately after resection (i.e., ex vivo). To achieve this, the tissue to be examined must be treated with a fluorescent agent, either by administering a fluorescent agent (FA) to the patient or by impregnating the tissue with it. The FA can be a non-targeted fluorescent dye, such as indocyanine green (ICG), or a targeted fluorescent dye used for imaging tumor tissue and infection, as well as for tracking drug therapy. Although FMI studies in the near-infrared (NIR) spectral range (700-900 nm) have shown promising results, several challenges remain. For example, light scattering and absorption by biological components such as water and blood can attenuate the excitation light, resulting in reduced sensitivity and contrast in fluorescence images. These factors can lead to the appearance of false-positive TPMs on fluorescence images, even when no TPMs are actually present in the patient. Consequently, the resulting fluorescence images (FI) are not always directly useful for guiding surgeons or other medical professionals or for use in medical treatment devices.

[0006] W. Heeman et al., "A Guideline for Clinicians Performing Clinical Studies with Fluorescence Imaging," J. Nucl. Med, 63 (2022) 640, describe a method in which a CNR ratio is determined after a pathologist examines a tissue sample. The CNR is the contrast-to-noise ratio calculated for the entire target region determined by the pathologist. Summary of the Invention

[0007] According to a first object, an improved examination method is provided for examining fluorescent images of mammalian tissue to facilitate treatment performed by a specialist or surgical facility.

[0008] According to a second object, an improved examination device is provided for examining fluorescent images of mammalian tissue to facilitate treatment performed by a specialist or surgical facility.

[0009] According to a third object, an improved medical treatment device is provided, comprising an improved examination device.

[0010] Typically, a fluorescence image to be inspected is obtained through preliminary steps as described below, the fluorescence image comprising an array of pixels having respective fluorescence signal values. The fluorescence signal values ​​comprise fluorescence signal values ​​representing a first quantity of mammalian tissue in the fluorescence image. The improved inspection method includes subsequent fluorescence image processing steps.

[0011] In a preliminary step, after mammalian tissue (e.g., human tissue) is treated with a fluorescent agent and made photosensitized, the tissue is irradiated with excitation light, and a fluorescent image is captured from the irradiated and photosensitized tissue. Fluorescent agents are used to display different types of tissue, such as tumor tissue and tumor-free tissue. In one example, the fluorescent agent is a targeted fluorescent tracer, such as vancomycin-IRDye800CW or hexvix, which is used to image tumor tissue and / or infection and to track drug treatment. In another example, the fluorescent agent is a non-targeted fluorescent dye, such as indocyanine green (ICG), which is used to image tissue perfusion. The excitation light used to irradiate tissue in vivo or in vitro is typically in the infrared range. The fluorescent agent can be administered to the patient or used to impregnate the tissue.

[0012] A subsequent step of the inspection method includes performing image segmentation to distinguish a target region and a reference region in the fluorescence image, wherein the target region represents a portion identified as tumor tissue in the mammalian tissue and the reference region represents a portion identified as healthy tissue in the mammalian tissue.

[0013] Before performing this image segmentation, a reference value is determined such that a second number of fluorescence signal values ​​included in the first number of fluorescence signal values ​​is less than or equal to the reference value, and a remaining portion of the first number of fluorescence signal values ​​exceeds the reference value. The second number is a predetermined fraction of the first number. In one example, a histogram of the fluorescence signal values ​​included in the first number of fluorescence signal values ​​is obtained, and the reference value is a value of a predetermined k-th q-quantile of the histogram. The predetermined k-th q-quantile should be based on an expected tumor tissue area ratio r of the tumor tissue area to the total tissue area in the fluorescence image (FI). e The ratio k / q should not exceed the tumor tissue area ratio. At the same time, the ratio k / q should not be too small. For example:

[0014]

[0015] This improved method is useful, for example, for in vitro examination of mammalian tissue removed by medical professionals for the purpose of removing tumor tissue. Typically, the mammalian tissue removed by medical professionals also includes a significant portion of normal tissue. This is because tissue that is easily identifiable as a tumor is surrounded by tissue that appears normal at first glance but may later develop into a tumor. Furthermore, it may not be practical to precisely follow the boundaries of the tumor tissue during the excision.

[0016] Therefore, in these cases the ratio k / q is usually chosen as:

[0017]

[0018] Calculate the average fluorescence signal value F of the fluorescence signal values ​​included in the second number of fluorescence signal values Band standard deviation S. Since the ratio k / q is chosen not to be too small, i.e. at least 0.1*r e , and preferably 0.2, so that there is enough data to reliably estimate the mean fluorescence signal value and standard deviation of the fluorescence signal of normal tissue. Since the ratio k / q is not too large, that is, at most 0.95*r e , and preferably 0.9, thus avoiding the estimation of the statistical data from being affected by the fluorescence signal data of the tumor tissue.

[0019] The average fluorescence signal value μ R and standard deviation σ R , image segmentation is then performed to distinguish between a target region (TR) and a reference region (RR) in the fluorescence image (FI). The target region (TR) is the region in the fluorescence image (FI) that corresponds to the portion of the mammalian tissue (MT) identified as tumor tissue in the segmentation. The reference region (RR) is the region in the fluorescence image (FI) that corresponds to the remaining portion of the mammalian tissue. In this segmentation operation, a signal-to-noise ratio CNR(p) is determined for each pixel (p). If the signal-to-noise ratio CNR(p) of a pixel exceeds a predetermined threshold, it is classified as part of the target region (TR), otherwise it is classified as part of the reference region (RR). The signal-to-noise ratio CNR(p) of a pixel is defined as:

[0020]

[0021] Where FI(p) is the fluorescence signal value of pixel (p), and c is a predetermined constant.

[0022] Thus, in contrast to the method known from Heeman, the CNR ratio is calculated on a pixel-by-pixel basis without a priori knowledge provided by a pathologist.

[0023] Best results are obtained in an embodiment where the predetermined threshold is 1 and the predetermined constant is 2.

[0024] However, variations are possible depending on further requirements. For example, if a safety margin needs to be considered when determining the target region, a predetermined threshold value less than 1 and / or a predetermined constant less than 2 may be selected. In this case, the estimated target region (TR) will be larger, so that it not only indicates the portion of the mammalian tissue (MT) identified as tumor tissue, but also indicates the portion of the mammalian tissue surrounding the portion identified as tumor tissue and at risk of developing into tumor tissue.

[0025] The method also includes identifying an outline of the target area. This outline can be displayed on a display, for example, superimposed on the fluorescence image, to aid the medical professional in performing an intervention on the selected tissue portion corresponding to the target area in the fluorescence image, such as by administering therapeutic radiation to the selected tissue portion, administering a therapeutic agent to the selected tissue portion represented by the target area, selectively activating a therapeutic agent in the selected tissue portion, or resecting the selected tissue portion. In addition to superimposing the outline on the fluorescence image, the outline can also be superimposed on a natural image of the tissue, i.e., an image that appears to have been taken under ambient light conditions. Thus, the medical professional can monitor the tissue during the medical intervention as if he / she were directly observing the tissue under ambient light conditions, rather than monitoring the fluorescence response of the image. In another example, the outline is projected onto the tissue.

[0026] In some embodiments of the method, the fluorescence image also captures the background, and the method provides preliminary image segmentation to distinguish between a foreground region representing mammalian tissue and a background region representing background in the fluorescence image. In one example, the background is formed by a carrier surface on which mammalian tissue (e.g., a completely excised tissue or a slice thereof) is placed for in vitro examination. In another example, when the fluorescence image is captured in vivo, the background is placed in the camera's field of view as a reference. In the preliminary image segmentation, for each pixel, a determination is made as to whether its fluorescence signal value (FSV) significantly exceeds the average background fluorescence signal value. If the fluorescence signal value (FSV) of the pixel is determined to significantly exceed the average background fluorescence signal value, the pixel is classified as part of the foreground region; otherwise, it is determined to be part of the background region. For example, if the difference between the fluorescence signal value (FSV) of the pixel and the background fluorescence signal value exceeds a predetermined multiple of the standard deviation of the background fluorescence signal value, the fluorescence signal value (FSV) of the pixel is determined to significantly exceed the average background fluorescence signal value. The predetermined multiple is, for example, selected within the range of 1-5, for example, approximately 2.

[0027] In one example, the mean fluorescence signal value and standard deviation are estimated in a calibration step, wherein a calibration fluorescence image is taken of only the background before taking the fluorescence image, and the mean fluorescence signal value and standard deviation of the fluorescence signal values ​​in the calibration fluorescence image are determined.

[0028] A rough but useful estimate of the mean fluorescence signal value can be obtained by:

[0029]

[0030] Where min and max are the minimum and maximum fluorescence values ​​of all fluorescence values ​​in the fluorescence image, respectively.

[0031] In another example, a mean fluorescence signal value and a standard deviation of fluorescence signal values ​​are determined (S2A) from a portion (BP) of a fluorescence image (FI) designated as representing background. For example, an operator can select a rectangular region in the fluorescence image (FI) that represents a portion of the background represented in the fluorescence image (FI). A complete preliminary image segmentation can then be performed based on the mean fluorescence signal value and the standard deviation of the fluorescence signal values ​​in the region.

[0032] Initial image segmentation can be followed by a dilation operation, where the foreground region is expanded by one or more pixels to mitigate the risk of edge effects. Other corrections can also be applied, such as removing regions identified in the initial image segmentation that are smaller than a threshold area. This can include small, isolated regions that were initially identified as foreground or background. Typically, the largest initially identified foreground region is selected for further processing, with any smaller initially identified foreground regions considered part of the background.

[0033] If a preliminary image segmentation is applied, a subsequent image segmentation to determine one or more target regions is applied to the image portions determined to be foreground regions. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] These and other aspects of the present invention are disclosed in more detail in the accompanying drawings, in which:

[0035] Figure 1 The steps of an improved inspection method for inspecting fluorescent images of mammalian tissue are schematically shown;

[0036] Figure 2 Optional steps to improve the inspection method are shown;

[0037] Figure 3A and Figure 3B Further optional steps to improve the inspection method are shown;

[0038] Figure 4A and Figure 4B The segmentation steps of the improved inspection method are shown;

[0039] Figure 5A 、 Figure 5B and Figure 5C The method is shown applied to fluorescence images obtained from sample tissue;

[0040] Figure 6A 、 Figure 6B and Figure 6C The method is shown applied to fluorescence images obtained from further sample tissue;

[0041] Figure 7 The steps of another improved inspection method for inspecting fluorescent images of mammalian tissue are schematically shown;

[0042] Figures 8 to 10 Shown Figure 7 The embodiment is applied to the fluorescence image of the sample tissue;

[0043] Figures 11 to 14 The method of claim 1 shows the use of heuristic information Figure 7 Application of embodiments;

[0044] Figure 15 Schematically illustrates an inspection device according to an embodiment of the present invention;

[0045] Figure 16 Schematically illustrates a medical treatment device according to one embodiment of the present invention;

[0046] Figure 17A 、 Figure 17B The tray with the sample imaged with visible light and its fluorescence image are shown respectively;

[0047] 18A to 18F Shows the different quantiles Figure 17B Segmentation of images in ;

[0048] Figure 19A 、 Figure 19B and Figure 19C A further method for detecting the location of the boundary between normal tissue and affected tissue is shown. DETAILED DESCRIPTION

[0049] Figure 1 Steps S5-S9 of an improved inspection method for inspecting a fluorescence image obtained from mammalian tissue are schematically shown.The fluorescence image is obtained by the following preparatory steps S1-S4.

[0050] In the preparation step S1, the mammalian tissue is treated with a fluorescent agent to make it photosensitized. Fluorescent agents are used to display different types of tissues, such as tumor tissue and tumor-free tissue. In one example, the fluorescent agent is a targeted fluorescent tracer, such as vancomycin-IRDye800CW or hexamidopropyl benzyl ammonium chloride (also referred to as hexvix), which is used to image tumor tissue and / or infection and track drug treatment. In another example, the fluorescent agent is a non-targeted fluorescent dye, such as indocyanine green (ICG), which is used to image tissue perfusion. The excitation light used to irradiate tissue in vivo or in vitro is typically in the infrared range. The fluorescent agent can be administered to the patient or used to impregnate the tissue.

[0051] In preparatory step S3, mammalian tissue treated with a fluorescent agent to render it photosensitized is illuminated with excitation light, and in preparatory step S4, a fluorescence image of the mammalian tissue is captured. The fluorescence image includes an array of pixels having respective fluorescence signal values. The fluorescence signal values ​​include a first number of fluorescence signal values ​​representing the mammalian tissue in the fluorescence image. If only mammalian tissue is captured in the image, the first number is the number of pixels in the fluorescence image (FI), but may be a smaller number if background is also present in the image.

[0052] In step S6 of the improved inspection method, a reference value is determined such that a second number of fluorescence signal values ​​included in the first number of fluorescence signal values ​​is less than or equal to the reference value and a remaining portion of the first number of fluorescence signal values ​​exceeds the reference value. The second number is a predetermined fraction of the first number.

[0053] In step S7, the average fluorescence signal value μ of the fluorescence signal values ​​included in the second number of fluorescence signal values ​​is determined. R and standard deviation σ R .

[0054] In step S8, image segmentation is performed to distinguish the target region TR from the reference region RR representing the remaining portion of the mammalian tissue MT in the fluorescence image FI. This step is performed pixel by pixel. That is, for each pixel (p), it is determined whether the signal-to-noise ratio CNR(p) of the pixel exceeds a predetermined threshold (T CNR ), then it belongs to the target region TR, otherwise it is determined that the pixel belongs to the reference region RR. The pixel signal-to-noise ratio CNR(p) is defined as:

[0055]

[0056] Where FI(p) is the fluorescence signal value of pixel (p), and c is a predetermined constant. The optimal value is when the threshold is 1 and the constant c is 2.

[0057] In step S9, at least one contour B of the target region TR is identified. In one example, the at least one contour includes a primary contour B indicating the boundary between the target region and a reference region. In another example, the at least one contour includes a secondary contour extending a certain distance outside the boundary between the target region and the reference region, thereby extending the target region by a safety zone to mitigate the risk of seemingly normal tissue near the target region later developing into tumor tissue. In a specific implementation of the secondary contour B', it extends a certain distance outside the boundary to avoid intersection with specific anatomical structures. In one example, the primary contour and the secondary contour are identified simultaneously.

[0058] Reference Figure 2An embodiment of the improved method is described. In this example, the method is used for in vitro examination of mammalian tissue, wherein the mammalian tissue is placed on a background. In this example, the mean fluorescence signal value and the standard deviation of the fluorescence signal values ​​are determined from a portion BP of a fluorescence image FI designated as representing the background. Through a user interface, a human operator can easily indicate a rectangular or square portion BP in the image that does not represent mammalian tissue. Statistical properties of the image data in this portion are then determined and used to perform Figure 1 The preliminary segmentation indicated as step S5 in FIG. 1 is performed based on the statistical properties of the background region estimated from the designated portion BP, thereby segmenting the fluorescence image (FI) into a foreground region and a background region. Typically, the estimated statistical properties include the mean fluorescence signal value of the background and the standard deviation of the fluorescence signal values. These statistical properties can be effectively estimated from the minimum and maximum fluorescence values ​​identified in the designated portion BP as follows:

[0059]

[0060] In preliminary image segmentation S5, for each pixel, a determination is made as to whether its fluorescence signal value (FSV) significantly exceeds the average background fluorescence signal value. If the fluorescence signal value (FSV) of the pixel is determined to significantly exceed the average background fluorescence signal value, the pixel is classified as part of the foreground region; otherwise, the pixel is determined to be part of the background region. For example, if the difference between the fluorescence signal value (FSV) of the pixel and the background fluorescence signal value exceeds a predetermined multiple of the standard deviation of the background fluorescence signal value, the fluorescence signal value (FSV) of the pixel is determined to significantly exceed the average background fluorescence signal value. The predetermined multiple is, for example, selected within the range of 1-5, such as approximately 2.

[0061] In one example, the mean fluorescence signal value and standard deviation are estimated in a calibration step, wherein a calibration fluorescence image is taken of only the background before taking the fluorescence image, and the mean fluorescence signal value and standard deviation of the fluorescence signal values ​​in the calibration fluorescence image are determined.

[0062] A rough but useful estimate of the mean fluorescence signal value can be obtained by:

[0063]

[0064] Where min and max are the minimum and maximum fluorescence values ​​of all fluorescence values ​​in the fluorescence image, respectively.

[0065] exist Figure 2 In the example shown, the following statistical properties of the background are estimated:

[0066] μ est ≈5.18;σ est ≈1.59

[0067] Based on this estimation, pixels with a fluorescence value of at least 8.36 were identified as foreground FG, i.e., representing mammalian tissue, and pixels with lower fluorescence values ​​were identified as background BG.

[0068] Figure 3A Schematically shows how a fluorescence image (FI) is segmented into a background region BG and a foreground region FG with a boundary B. Figure 3B A further correction of the boundary to the correction boundary B' is shown. The further correction is achieved by dilating the foreground FG by 1 pixel. The fluorescence signal values ​​in the fluorescence image (FI) include a first number N1 of fluorescence signal values ​​in the foreground. These represent mammalian tissue in the fluorescence image.

[0069] Figure 4A The reference value IR is determined such that the second number N2 of fluorescence signal values ​​included in the first number N1 of fluorescence signal values ​​is less than or equal to the reference value IR, and the remaining portion of the first number of fluorescence signal values ​​exceeds the reference value IR. The second number N2 is a predetermined fraction of the first number N1. In this example, the predetermined fraction is 0.5. This means that the reference value IR is the median of the fluorescence signal values ​​included in the first number N1 of fluorescence signal values. In further analysis, it is determined that the fluorescence signal values ​​included in the second number N2 of fluorescence signal values ​​have the following statistical properties:

[0070] μ R ≈37.54;σ R ≈12.91

[0071] where μ RR and σ RR are respectively the estimated mean value and the estimated standard deviation of the fluorescence signal values ​​contained in the second number N2 of fluorescence signal values.

[0072] Based on these statistical characteristics, image segmentation is performed to distinguish the target region TR from the reference region RR in the fluorescence image, where the reference region represents the remaining portion of the mammalian tissue other than the portion represented by the target region. Figure 4B In the case where the signal-to-noise ratio CNR(p) exceeds a predetermined threshold (T CNR ) are identified as Figure 4B The remaining pixels represented by white are identified as part of the reference region RR. The signal-to-noise ratio CNR(p) of a pixel is defined as:

[0073]

[0074] Where FI(p) is the fluorescence signal value of pixel (p), c is a predetermined constant. In this example, the value of c is 2, and the predetermined threshold (T CNR ) has a value of 1.

[0075] Figure 5A 、 Figure 5B 、 Figure 5C and Figure 6A 、 Figure 6B 、 Figure 6C Various examples of application of the above-described method are shown. Figure 5A 、 Figure 5B and Figure 5C The example shown is a tissue section of penile squamous cell carcinoma obtained from tissue resected from the first patient. Figure 6A 、 Figure 6B and Figure 6C The example shown is a penile squamous cell carcinoma tissue section obtained from a second patient. After the tissue was treated with vancomycin-IRDye800CW to render it fluorescent, a fluorescence image was acquired using a PEARL imaging device, capturing wavelengths in the 800 nm range. Reference B' represents the corrected boundary of the mammalian tissue relative to the background. References C, C1, and C2 represent the outline of the target region in the tissue estimated by this method. Reference GT represents the actual tumor tissue outline, as determined by a pathologist.

[0076] The above method can be used in conjunction with the scanline-based method. Figure 7 An example of a scanning trajectory based approach is schematically shown.

[0077] The inspection method shown therein comprises step S10, wherein at least one fluorescence signal value vector in the fluorescence image along a scanning trajectory is obtained. For practical purposes, the scanning trajectory is typically a line aligned with the main axis of the fluorescence image.

[0078] In step S11, for a threshold value ts, one or more candidate scanning trajectory segments (usually scanning line segments) whose fluorescence signal value exceeds the threshold value ts are determined. This is repeated for multiple threshold values. In step S12, it is verified whether the procedure has been performed for all threshold values ​​in the multiple threshold values.

[0079] The remainder of the scan trajectory is considered as a reference segment and the statistical properties of the fluorescence signal values ​​of the pixels forming part of it are derived. Typically the mean value μ of these fluorescence signal values ​​is determined. rs and standard deviation σ rs Since the values ​​of these statistical properties depend on the choice of threshold ts, they can be written as μ rs (ts) and σ rs (ts).

[0080] After determining these statistical characteristics, determine which pixels (p) on the scan trajectory meet the requirement CNR(p)≥tr, where:

[0081]

[0082] Wherein c is a predetermined constant selected from the range of 1.5 to 3, for example, about 2, and tr is a threshold value, for example, a value of 1.

[0083] In step S13 , it is determined which threshold value ts among the multiple threshold values ​​has the best matching set of candidate scan trajectory segments with the one or more scan trajectory segment sets obtained using the above CNR requirement. In one example, the matching degree of the identified scan trajectory segments is determined by the above specified F measure.

[0084] Alternatively, the optimal value t can be determined by opt :

[0085]

[0086] exist Figure 7 In the embodiment of the present invention, the procedure of steps S10-S13 is repeated for a plurality of mutually different scanning trajectories in a set of scanning trajectories. To this end, it is verified in step S14 whether steps S10-S13 have been performed for all scanning trajectories in the set of scanning trajectories. Therefore, for each scanning trajectory in the set of scanning trajectories, the value t is determined. opt And based on the value t opt As mentioned above, the scanning trajectory is preferably a line aligned with the main axis of the fluorescence image, such as the set of all horizontal lines, or a subset of horizontal lines.

[0087] In the embodiment shown, the procedure of steps S10-S14 is also repeated for multiple groups of scanning trajectories. To this end, in step S15 it is verified whether steps S10-S14 have been performed for a group of scanning trajectories, for example a group of horizontal scanning lines and a group of vertical scanning lines.

[0088] Figure 7 The method can be used in various ways with Figure 1 In one example, Figure 1 The method provides the expected optimal threshold t opt An indication of the scope you are in. For example, if you select:

[0089] μ r ≤t opt ≤μ r +4*σ r

[0090] or

[0091] μ r +σ r ≤t opt ≤μ r +3*σ r

[0092] In another example, Figure 1 The method provides an extension range for limiting the extension range of the scanning trajectory. For example, the extension range may indicate that the scanning trajectory passes through Figure 1 The method is to take the line segments of the indicated contour and extend them a predetermined distance on both sides of the intersection. Alternatively, Figure 1 The method may indicate an extension range of the scanning trajectory that passes through the target area and extends a predetermined distance on both sides of the target area.

[0093] In a further example, Figure 1 The method provides an indication of the direction of the target area contour to be identified. Figure 1 Using this information, the analysis can be optimally performed by performing a scan trajectory based analysis using scan lines transverse to the main direction. Figure 1 method.

[0094] A suitable definition of the principal axis is the line segment that minimizes the average distance metric between the line segment and the contour points. For example, the distance metric is the Euclidean distance metric, but other options can also be used. The principal axis can be found using the OpenCV tool "fitLine".

[0095] For practical purposes, when executing Figure 7 The fluorescence image (FI) is rotated before the method. In this case, the rotation aligns the main axis of the contour with the coordinate axis of the fluorescence image and the scan line is along the other coordinate axis of the fluorescence image (FI).

[0096] In the above Figure 7 In the method, the threshold ts is changed to determine the optimal threshold t opt , the corresponding candidate scanning trajectory segment set of the scanning trajectory under this threshold best matches one or more scanning trajectory segment sets obtained using the above CNR requirements. rs and σ rs is estimated from scan trajectory segments that have not been identified as candidate scan trajectory segments. In another approach, it is assumed that the scan trajectory has an intersection at an assumed position, and the statistical properties μrs and σrs are estimated from the scan trajectory segment with the lowest mean fluorescence value on the side of the assumed intersection. Based on the CNR metric with these estimated statistical properties, the scan trajectory is segmented into a reference scan trajectory segment and a target scan trajectory segment, and it is determined whether the assumed intersection coincides with the transition from the reference scan trajectory segment to the target scan trajectory segment or vice versa. If this is the case, the assumed intersection is a candidate contour point. This method is described in more detail in European patent application 23154545.0 filed on February 1, 2023 by the same applicant. Figure 1 The method can provide the location of the contour as heuristic information to limit Figure 7The search range for the optimal intersection position in alternative methods.

[0097] Figure 7 The method is through Figure 8 Provide explanation.

[0098] Figure 8 The upper portion of FIG shows a fluorescence image (FI) obtained from a mammalian tissue sample. The lower portion shows the intensity values ​​of the fluorescence image (FI) along a scan line L. In the example shown, each intensity value along the scan line is obtained by Gaussian-weighted averaging of pixels in a strip W with the same x-coordinate.

[0099] Figure 9 It is shown how, in step S11 , a corresponding set of candidate scanning trajectory segments is determined for each of a plurality of threshold values, wherein the fluorescence signal value exceeds the threshold value. Figure 9 The upper part of shows a first example, where a threshold of 50 results in a segmentation with two candidate target segments T50a and T50b and two reference segments R50a and R50b. Figure 9 The lower portion of shows the segmentation for a threshold value 58 with one candidate target segment T58 and one reference segment R58.

[0100] For each threshold, the statistical property μ is estimated from the fluorescence signal values ​​in one or more reference segments of the scan trajectory. rs and σ rs , and determining an alternative segmentation of the scan trajectory according to the CNR criterion based on these statistical characteristics. Identifying a threshold t that makes the threshold-based segmentation best match the CNR-based segmentation using the statistical characteristics of one or more reference segments opt .

[0101] like Figure 10 As shown, in Figure 7 In step S13 of the method, it is found that for this scan line, the optimal threshold is 62. Therefore, Figure 10 As shown in the upper part, the scan line is divided into a reference segment R62 and a target segment T62. Figure 10 As shown in the lower portion, contour points Ts and Te are identified, which represent the start and end points of the target area T62 when traversing the scan line in the positive x direction.

[0102] As another example, Figure 11 Shows how to Figure 1 The method is applied to identify the first contour C1 and the second contour C2 (see Figure 11 Lower) to Figure 11 Fluorescence images shown in the upper part.

[0103] Figure 12 The upper part shows how the main axis AX1 of the target area with the contour C1 is determined. Figure 12The lower part shows how to rotate the fluorescence image (FI) so that the main axis AX1 is aligned with one of the main axes of the fluorescence image (FI) (here, the y-axis). It should be noted that for contours with more complex shapes, multiple lines can be used for fitting.

[0104] The fluorescence image (FI) rotated in this way can be optimally scanned by scanning lines along the x-axis. In addition, the position of the contour C1 can provide heuristic information to reduce Figure 7 For example, the heuristic information indicates a spatial search range or indicates an intensity search range.

[0105] Figure 13 The example shown above indicates the use of Figure 1 The heuristic information of the method and compare the image with Figure 11 The first contour C1 indicated in the alignment is passed Figure 7 The threshold-based version of the method identifies the start (+) and end (·) points of the target segment. Figure 13 The lower part indicates the use Figure 1 The heuristic information of the method and compare the image with Figure 11 The second contour C2 indicated in the figure is aligned by Figure 7 The threshold-based version of the method identifies the start (+) and end (·) points of the target segment.

[0106] Figure 14 The top part shows the points mapped to each contour identified in the original image.

[0107] Figure 14 The lower part shows how isolated points are removed in subsequent steps. A point is considered isolated if the local point density is below a threshold, that is, if the ratio of the number of points in a predetermined area centered on the point to the area size is less than a predetermined value. Typically, the area size is selected in the range of 100 to 1000 pixels, and the minimum number of points required to be present in the area is selected in the range of 5 to 50. The best results are achieved if the number of points is within the range of 1 / 30 to 1 / 20 of the area size. In this example, the area radius is 40 pixels, which means the area contains approximately 500 pixels, and the minimum number of points is 20.

[0108] In an additional or alternative processing step, point clusters whose maximum pixel intensity is less than a reference value are removed. For example, the reference value is the global intensity mean plus a multiple of the standard deviation, which can be used as a modifiable parameter. In a further additional or alternative processing step, point clusters are removed if their area (to be calculated) is less than a certain fraction of the total image area.

[0109] Figure 15An inspection device 1 for inspecting mammalian tissue is schematically shown, which is configured to obtain a fluorescence image FI of mammalian tissue MT treated with a fluorescent agent to make it photosensitivity and irradiated with excitation light. The fluorescence image FI includes an array of pixels having respective fluorescence signal values. The fluorescence signal values ​​include fluorescence signal values ​​representing a first number N1 of mammalian tissue in the fluorescence image. If there is background in the image, the first number is less than the total number of pixels in the fluorescence image (FI). Otherwise, the first number N1 may be equal to the total number of pixels. In the example shown, the inspection device 1 obtains the fluorescence image (FI) from an external input 1i. Alternatively, the inspection device may include a camera for capturing the image. The inspection device may also include an appropriate excitation light source.

[0110] Inspection apparatus 1 includes a reference value determination module 11 configured to determine a reference value IR such that a second number N2 of fluorescence signal values ​​included in a first number of fluorescence signal values ​​is less than or equal to the reference value IR, and the remaining portion of the first number of fluorescence signal values ​​exceeds the reference value IR. In one example, the second number is a predetermined fraction of the first number, e.g., N2 / N1 = 0.5, in which case the reference value is the median. A statistical characteristic evaluation module 12 determines a mean fluorescence signal value (μR) and a standard deviation (σR) of the fluorescence signal values ​​included in the second number N2 of fluorescence pixel values.

[0111] The segmentation module 13 performs image segmentation to distinguish a target region (TR) from a reference region (RR) in the fluorescence image (FI), wherein the target region (TR) represents the portion of the mammalian tissue (MT) identified as tumor tissue and the reference region (RR) represents the remaining portion of the mammalian tissue (MT). The segmentation module operates in a pixel-by-pixel manner, i.e., for each pixel (p), it is determined whether the signal-to-noise ratio (CNR(p)) of the pixel exceeds a predetermined threshold (T CNR ), then it belongs to the target region (TR), otherwise it is determined that the pixel belongs to the reference region (RR), where the pixel signal-to-noise ratio CNR(p) is defined as:

[0112]

[0113] Where FI(p) is the fluorescence signal value of pixel (p), c is a predetermined constant. The optimal value is T CNR =1 and c=2.

[0114] The segmentation module 13 is further configured to identify a contour C of the target region. Contour C is a primary contour indicating the boundary between the target region and the reference region. The segmentation module 13 is further configured to identify a secondary contour C' that extends a certain distance outside the boundary between the target region and the reference region. In the illustrated example, the segmentation module 13 is configured to generate the secondary contour C' so that it extends a certain distance outside the boundary in a manner that avoids intersecting specific anatomical structures.

[0115] Figure 16 A medical treatment device 100 is shown, which comprises, in addition to the elements of an inspection device, an excitation light source 7 for irradiating mammalian tissue, a camera 6 for obtaining a fluorescence image FI of the mammalian tissue, and a treatment device 5 for ablating or irradiating a tumor according to a constructed contour or activating a treatment substance within a range specified by the constructed contour.

[0116] The present invention allows for more accurate identification of the contours of affected tissue, such as tumor-affected or infected tissue. Precisely knowing the location of these contours is crucial for treatment. This is important, for example, during surgical removal of affected tissue to ensure that no affected tissue remains after surgery, while also ensuring that no more healthy tissue is removed than necessary. Similarly, precisely knowing the location of these contours is crucial for applying photodynamic therapy. Consequently, the therapeutic substance is locally activated within the area of ​​tissue indicated as affected. The therapeutic substance is inactivated outside of the indicated area, minimizing damage to healthy tissue.

[0117] In one example, the therapeutic substance can be activated to act as a chemotherapeutic agent in an area indicated as containing a tumor. In another example, the therapeutic substance can be activated to have antimicrobial activity in an area indicated as containing infected tissue.

[0118] Therefore, the photodynamic therapy comprises at least the following steps: obtaining in vivo a fluorescence image of mammalian tissue treated with a fluorescent agent to be photosensitive and irradiated with excitation light. The obtained fluorescence image comprises a pixel array having respective fluorescence signal values.

[0119] Fluorescent agents are used to visualize affected tissue, such as tumors or infected tissue. In one example, the fluorescent agent is a targeted fluorescent tracer, such as Vancomycin-IRDye800CW or Hexvix, used to image tumor tissue and / or infection and to track drug treatment. In another example, the fluorescent agent is a non-targeted fluorescent dye, such as indocyanine green (ICG), used to image tissue perfusion.

[0120] Exemplary agents for imaging infected tissue are vancomycin-IRDye800CW and 1D9-IRDye800CW. 1D9 is a monoclonal antibody against Staphylococcus aureus (including MRSA). More examples are listed in the references in the appendix.

[0121] An example of the latter application is in Figure 17A 、 Figure 17B and Figures 18A-18F Shown in.

[0122] Figure 17A Shown is an image of a tray containing five Foam Samples IV, prepared as specified in the table below.

[0123] Figure 17B Shown are fluorescence images obtained from the same tray samples at a wavelength of 800 nm using a Pearl imaging system.

[0124] sample Tracer (Y / N) Coccal culture (Y / N) I N N II N Y III Y Y: IV Y N V Y Y

[0125] The first sample I was a sterile foam in which no tracer was yet present.

[0126] The second sample, Foam II, had been immersed in a solution containing a culture of S. aureus but without a tracer.

[0127] The third sample, Foam III, had been immersed in a solution containing a culture of Staphylococcus aureus and using 10 μL of the tracer tIRDye800CW solution.

[0128] The fourth sample, Foam IV, had been immersed in a sterile solution containing the tracer tIRDye800CW.

[0129] The fifth sample, Foam V, has been immersed in a solution containing a Staphylococcus aureus culture and 20 μL of the tracer tIRDye800CW solution.

[0130] Figures 18A-18F Shows the Q value according to different quantile settings Figure 17B The fluorescence image is segmented, and Q ranges from Figure 18A Q = 0.50 to Figure 18F Q = 0.99. Therefore, what is shown is the fraction of pixels whose fluorescence intensity is higher than the fraction Q of unselected pixels (1-Q).

[0131] Based on the reference values ​​used to determine the mean fluorescence signal value μR and the standard deviation σR in step S7, a fraction Q of pixels have fluorescence intensities less than the reference values, and a fraction 1-Q has fluorescence intensities greater than or equal to the reference value. Optimal results have been found using Q = 0.70.

[0132] As part of the current research, the inventors hypothesized that the fluorescent radiation observed at the boundary of affected tissue (e.g., tumor tissue or infected tissue) is stray radiation, i.e., fluorescent radiation from the affected tissue is scattered by healthy tissue near the boundary. The inventors further hypothesized that, therefore, the intensity of the fluorescent radiation is expected to decrease exponentially in the direction away from the boundary, i.e.,

[0133] I(d)=I0e -α·d

[0134] where I0 is the intensity measured at the boundary of the affected tissue, I(d) is the intensity at a location d from the boundary in the healthy tissue, and α is a constant. Based on these observations, Figure 19A -C describes the following methods.

[0135] Figure 19A The measured fluorescence values ​​I(x) are shown for the positions along the scan path in the fluorescence image. In this example, the scan path is a scan line in the x-direction of the image, but the scan line can have any other direction. The scan path can also be a curved path. Figure 19A As shown, the scan range extends from 0 to about 210 pixels. Since the value of function I(x) at position B is relatively high, it is assumed that the pixel at coordinate x=130 represents a portion of the affected tissue. Now, how to determine the boundary of the affected tissue at the position to the left of B is described.

[0136] Figure 19B The logarithm of the measured fluorescence value I(x) is shown as log(I(x)). Based on the above observations, it is expected that the value of log(I(x)) decreases as follows in the direction away from the boundary:

[0137] log(I(d))=-α·d·log(I(x0))

[0138] Therefore, the function log(I(x)) is expected to have a linear part in healthy tissue areas near the affected tissue. To identify these areas, the linearity of log(I(x)) as a function of x is determined by fitting the curve to a linear function in a sliding window. For example, the sliding window length is 30 pixels and the linear function is fitted using the least squares method. Different lengths can also be used, but the length should not be too small in order to have a sufficient signal-to-noise ratio. Preferably, the length is at least 10 pixels. In order to have a sufficiently high resolution, the length should not be too large. However, this depends on the resolution of the image (mm / pixel). For the current case, where the resolution is 85 μm, the length of the sliding window is preferably not more than 70 pixels. However, if the resolution is higher by a certain factor, the maximum length of the window can also be larger by this factor. The sliding window is symmetrical to avoid deviations in the linearity function. However, it is possible to consider using an asymmetric window and compensating for deviations.

[0139] exist Figure 19C In Figure 1, the correlation of the function log(I(x)) within the sliding window is shown as the linearity (Linearity(x)). A value of 1 represents the extreme case where the function log(I(x)) within the sliding window is completely linear, and a value of 0 represents the maximum deviation from the linearity of the function log(I(x)) within the sliding window. It should be noted that in the region of constant intensity I(x), the function Linearity(x) will also be close to 1. For practical purposes, if the function Linear(x) is at least 0.95 (e.g. Figure 19CThe function log(I(x)) is assumed to be linear, as shown by the dashed line in the figure. In the x-axis direction, the function drops below this threshold at position A. This is considered to be the location where the scan line intersects the boundary of the affected tissue.

[0140] This is the transition from the region where log(I(x)) increases linearly to the region where it exhibits a more constant value. Point A' indicates the location along the scan line where a more constant value has been achieved. Similarly, the location of affected tissue adjacent to healthy tissue to the right of point B can be determined.

[0141] The method can be repeated using different scan lines. For example, in this case, the scan lines extend along the x-direction, and the boundary position can be determined for corresponding scan lines with different y-coordinates to estimate the contour of the affected tissue within the healthy tissue. The method can also be applied alternatively or additionally to scan lines with different orientations.

[0142] Reference Figures 19A-19C The described method can be used independently, but can also be used as a preprocessing step for further analysis. For example, based on the contours thus identified, it is possible to estimate how many pixels in the image represent affected tissue and how many represent healthy tissue. For example, if the image of the tissue includes an area of ​​size N0 pixels representing affected tissue and an area of ​​size N2 pixels containing healthy tissue, the predetermined fraction used to determine the reference value IR is N2 / (N0+N2). This method is also applicable to the preprocessing step of the method described in international patent application PCT / NL2024 / 050047 filed by the same applicant. This preprocessing step can be used to determine a provisionally assigned location.

[0143] Or, refer to Figures 19A-19C The method can be used as a post-processing step to correct the segmentation results obtained by another method.

[0144] In short, refer to Figures 19A-19C The illustrated inspection method includes the following steps.

[0145] Assume that a fluorescence image of mammalian tissue treated with a fluorescent agent to be photosensitive and irradiated with excitation light has been obtained, and the fluorescence image (FI) includes an array of pixels having respective fluorescence signal values.

[0146] Fluorescence images can be obtained in vitro to verify complete removal of affected tissue during surgery. Alternatively, fluorescence images can be obtained in vivo to assist surgeons during surgery.

[0147] Fluorescent agents are used to visualize affected tissue, such as tumors or infected tissue. Exemplary agents for imaging infected tissue are vancomycin-IRDye800CW and 1D9-IRDye800CW. 1D9 is a monoclonal antibody against Staphylococcus aureus (including MRSA). Further examples are listed in the references in the appendix.

[0148] At least one sequence of fluorescence values ​​I(x) is obtained for a pixel at position x along the scan path. The logarithmic value of each value in the sequence of fluorescence values ​​I(x) is determined. Alternatively, one may calculate the logarithmic values ​​of all fluorescence values ​​I(x) in the image and then obtain a sequence of logarithmic values ​​log(I(x)) for the pixel at position x along the scan path. However, it is generally preferred to first obtain the sequence of fluorescence values ​​and then apply the logarithmic function, so that the logarithmic function does not need to be applied to pixels that do not participate in the calculation.

[0149] The degree to which the sequence of logarithmic values ​​log(I(x)) locally approximates a linear function is then determined.

[0150] The transition from healthy tissue to affected tissue boundary is estimated where the linearity begins to degrade significantly (eg, falls below a threshold value, such as 0.95).

[0151] The scanning path can be a scanning line in any direction, but can also be a curved path. However, a scanning line or a scanning path with negligible curvature is preferred to avoid the curvature of the path affecting the linearity of the sequence of values ​​in the boundary region.

[0152] Appendix: Further References

[0153] Empowering antimicrobial photodynamic therapy of Staphylococcus aureus infections with potassium iodide;

[0154] Bispo M, Suhani S, van Dijl JM, J Photochem Photobiol B, December 2021, 225: 112334, doi: 10.1016 / j.jphotobiol.2021.112334.

[0155] Comparison of two fluorescent probes in preclinical non-invasive imaging and image-guided debridement surgery of Staphylococcal biofilm implant infections;

[0156] Park HY,Zoller SD,Hegde V,Sheppard W,Burke Z,Blumstein G,Hamad C,Sprague M,Hoang J,Smith R,Romero Pastrana F,Czupryna J,Miller LS, M, Bispo M, van Oosten M, van Dijl JM, Francis KP, Bernthal NM, SciRep, January 15, 2021, 11(1):1622, doi:10.1038 / s41598-020-78362-7.

[0157] Fighting Staphylococcus aureus infections with light and photoimmunoconjugates;

[0158] Bispo M,Anaya-Sanchez A,Suhani S,Raineri EJM, M, Heuker M, Szymański W, Romero Pastrana F, Buist G, Horswill AR, Francis KP, van Dam GM, van Oosten M, van Dijl JM, "JCI Insight (JCI Vision)", November 19, 2020, 5(22):e139512, doi: 10.1172 / jci.insight.139512.

[0159] A Facile and Reproducible Synthesis of Near-Infrared Fluorescent Conjugates with Small Targeting Molecules for Microbial Infection Imaging;

[0160] Reeβing F,Bispo M, M, van Oosten M, Feringa BL, van Dijl JM, Szymański W, ACS Omega, 26 August 2020, 5(35): 22071–22080, doi:10.1021 / acsomega.0c02094.

[0161] Novel in vivo mouse model of shoulder implant infection;

[0162] Sheppard WL, Mosich GM, Smith RA, Hamad CD, Park HY, Zoller SD, Trikha R, McCoy TK, Borthwell R, Hoang J, Truong N, Cevallos N, Clarkson S, Hori KR, van DijlJM, Francis KP, Petrigliano FA, Bernthal NM, "J Shoulder Elbow Surg (Journal of Shoulder and Elbow Surgery)", July 2020, 29(7):1412-1424, doi:10.1016 / j.jse.2019.10.032.

[0163] Multimodal imaging guides surgical management in a preclinical spinal implant infection model;

[0164] Zoller SD,Park HY,Olafsen T,Zamilpa C,Burke ZD,Blumstein G,SheppardWL,Hamad CD,Hori KR,Tseng JC,Czupryna J,McMannus C,Lee JT,Bispo M,RomeroPastrana F,Raineri EJ,Miller JF,Miller LS,van Dijl JM,Francis KP,Bernthal NM, "JCI Insight (JCI Vision)", February 7, 2019, 4(3):e124813, doi:10.1172 / jci.insight.124813.

[0165] Noninvasive optical and nuclear imaging of Staphylococcus-specific infection with a human monoclonal antibody-based probe;

[0166] Romero Pastrana F,Thompson JM,Heuker M,Hoekstra H,Dillen CA,OrtinesRV,Ashbaugh AG,Pickett JE,Linssen MD,Bernthal NM,Francis KP,Buist G,vanOosten M,van Dam GM,Thorek DLJ,Miller LS,van Dijl JM, "Virulent", January 1, 2018, 9(1):262-272, doi:10.1080 / 21505594.2017.1403004.

[0167] Real-time in vivo imaging of invasive-and biomaterial-associated bacterial infections using fluorescently labelled vancomycin;

[0168] van Oosten M, T, Gazendam JA, Ohlsen K, Tsompanidou E, de Goffau MC, Harmsen HJ, Crane LM, Lim E, Francis KP, Cheung L, Olive M, Ntziachristos V, vanDijl JM, van Dam GM, *Nat Commun* (Nature Communications), 2013, 4: 2584, doi: 10.1038 / ncomms3584.

Claims

1. A method for inspecting a fluorescence image captured of mammalian tissue treated with a fluorescent agent to render it photosensitivity and irradiated with excitation light, the fluorescence image (FI) comprising an array of pixels having respective fluorescence signal values, the fluorescence signal values ​​including fluorescence signal values ​​representing a first number (N1) of mammalian tissue in the fluorescence image; the method comprising: determining (S6) a reference value (IR) such that a second number (N2) of fluorescence signal values ​​included in the first number of fluorescence signal values ​​is less than or equal to the reference value and a remaining portion of the first number of fluorescence signal values ​​exceeds the reference value, the second number being a predetermined fraction of the first number; Determine (S7) an average fluorescence signal value (μ) of the fluorescence signal values ​​included in the second number of fluorescence signal values. R ) and standard deviation (σ R ); performing (S8) image segmentation to distinguish a target region (TR) from a reference region (RR) in the fluorescence image (FI), wherein the target region (TR) represents a portion of the mammalian tissue (MT) identified as a tumor tissue or an infected tissue, and the reference region (RR) represents a remaining portion of the mammalian tissue (MT), including determining for each pixel (p) whether a signal-to-noise ratio CNR(p) of the pixel exceeds a predetermined threshold (T CNR ), then it belongs to the target region (TR), otherwise it is determined that the pixel belongs to the reference region (RR), where the signal-to-noise ratio CNR(p) of the pixel is defined as: Where FI(p) is the fluorescence signal value of pixel (p), and c is a predetermined constant; At least one contour (B) of the target region (TR) is identified (S9). 2 . The inspection method according to claim 1 , wherein the predetermined threshold is 1 and the predetermined constant is 2.

3. The inspection method according to claim 1 or 2, wherein the fluorescent image (FI) also captures the background (S4), and wherein the method further comprises: Performing (S5) preliminary image segmentation to distinguish in the fluorescence image (FI) a foreground region (FR) representing the mammalian tissue (MT) and a background region (BR) representing the background, including determining for each pixel that if the fluorescence signal value (FSV) significantly exceeds the background mean fluorescence signal value taking into account the standard deviation of the fluorescence signal value (FSV) in the background, then it belongs to the foreground region (FR), otherwise determining that the pixel belongs to the background region (BR).

4. Inspection method according to claim 3, wherein the mean fluorescence signal value and the standard deviation of the fluorescence signal values ​​are determined (S2A) from a portion (BP) of the fluorescence image (FI) designated as representing background.

5. The examination method according to any one of the preceding claims, wherein the fluorescent image of the mammalian tissue is taken in vivo.

6. The inspection method according to any one of claims 1 to 4, further comprising wherein the fluorescent image of the mammalian tissue is taken in vitro when the mammalian tissue is placed on the background.

7. The inspection method according to any one of claims 2 to 6, wherein performing (S5) preliminary image segmentation further comprises a dilation operation. 8 . The inspection method according to claim 1 , wherein the predetermined k-th quantile is the k-th percentile, wherein k is in the range of 25 to 75. The inspection method according to claim 8 , wherein the percentile is a median.

10. The inspection method according to any one of the preceding claims, further comprising an evaluation based on a scanning trajectory, comprising: obtaining (S10) at least one fluorescence signal value vector in the fluorescence image (FI) along a scanning trajectory; For each threshold value among the plurality of threshold values, determining (S11) a corresponding set of candidate scanning trajectory segments having fluorescence signal values ​​exceeding the threshold value, and determining statistical characteristics of scanning trajectory segments that are not candidate scanning trajectory segments; It is determined ( S13 ) which threshold value among a plurality of threshold values ​​has a corresponding set of candidate scanning trajectory segments that best matches the fluorescence image (FI) segmentation along the scanning trajectory obtained by the signal-to-noise ratio-based segmentation using the determined statistical characteristics.

11. The inspection method according to claim 10, wherein the procedure (S14) is repeated for a plurality of mutually different scanning trajectories in a set of scanning trajectories.

12. The inspection method according to claim 11, wherein the procedure (S15) is repeated for multiple sets of scanning trajectories.

13. The inspection method according to any one of claims 10 to 12, further comprising removing (S16) isolated points.

14. The inspection method according to any one of the preceding claims 1 to 13, further comprising determining at least one principal axis for at least one contour; obtaining at least one fluorescence signal value vector in the fluorescence image (FI) along a scan line orthogonal to the main axis; Based on information of the contour at an intersection with the scan line, selectively identifying one or more segments of the scan line wherein a signal-to-noise ratio exceeds a predetermined level, the scan line segments each including a respective first endpoint indicating a transition (NT) from normal tissue to tumor tissue or infected tissue and a respective second endpoint indicating a transition (TN) from the tumor tissue or infected tissue to normal tissue. 15 . The inspection method according to claim 14 , wherein the principal axis is defined as a line segment that minimizes an average distance between the line segment and points of the contour.

16. The inspection method according to claim 14 or 15, wherein before the step of obtaining (S22) at least one fluorescence signal value vector, the fluorescence image (FI) is rotated (S21) so that the main axis of the contour coincides with the coordinate axis of the fluorescence image (FI), and wherein the scanning line is along the direction of another coordinate axis of the fluorescence image (FI).

17. The inspection method according to claim 14, 15 or 16, wherein the information on which the selective execution (S23) is based includes the intensity of the fluorescence image (FI) at the intersection points of the contours.

18. The inspection method according to claim 14, 15 or 16, wherein the information on which the selective execution (S23) is based includes the coordinates of the intersection point.

19. The inspection method according to any one of claims 14 to 18, further comprising combining (S24) and indicating the coordinates of the start points and end points of the plurality of contours in a single image.

20. The inspection method according to claim 19, wherein the combining (S24) comprises performing a counter-rotation according to a major axis of the profile.

21. The inspection method according to any one of claims 10 to 16, further comprising Determine the corresponding endpoint density in the corresponding region at each endpoint, and Remove those endpoints whose corresponding density is less than a predetermined threshold.

22. Inspection method according to any of the preceding claims, wherein the at least one contour comprises a main contour (C) indicating a boundary between a target area and a reference area.

23. Inspection method according to any of the preceding claims, wherein the at least one contour comprises a secondary contour (C') extending a distance outside the border of the target area and the reference area.

24. The inspection method according to claim 23, wherein the secondary contour (C') extends a distance outside the border in such a way as to avoid intersection with specific anatomical structures.

25. The inspection method according to claim 1, further comprising the following steps: obtaining at least one sequence of respective logarithmic fluorescence signal values ​​for consecutive pixels arranged along a scanning trajectory within the fluorescence image, wherein the scanning trajectory begins at a location identified as part of a reference region in the fluorescence image; calculating an indicator of local linearity of the log fluorescence signal values ​​at positions in the sequence; A position in the sequence where the indicator indicates that the logarithmic fluorescence signal value as a function of position in the sequence is no longer linear is identified, the identified position being a candidate boundary position for a boundary between normal tissue and tumor tissue or infected tissue.

26. An examination device (1) for examining mammalian tissue, the device being configured to: obtaining a fluorescence image (FI) of mammalian tissue (MT) treated with a fluorescent agent to render it photosensitivity and illuminated by excitation light, the fluorescence image comprising an array of pixels having respective fluorescence signal values, the fluorescence signal values ​​comprising a first number of fluorescence signal values ​​representing the mammalian tissue in the fluorescence image; Determine the average fluorescence signal value (μ) of the fluorescence signal values ​​contained in the second number of fluorescence pixel values. R ) and standard deviation (σ R ); performing image segmentation to distinguish a target region (TR) from a reference region (RR) in the fluorescence image (FI), wherein the target region (TR) represents a portion of the mammalian tissue (MT) identified as a tumor tissue or an infected tissue, and the reference region (RR) represents a remaining portion of the mammalian tissue (MT), wherein the apparatus is configured to determine, for each pixel (p), whether the pixel is positive or negative if a signal-to-noise ratio CNR(p) of the pixel exceeds a predetermined threshold (T CNR ), then it belongs to the target region (TR), otherwise it is determined that the pixel belongs to the reference region (RR), where the signal-to-noise ratio CNR(p) of the pixel is defined as: Where FI(p) is the fluorescence signal value of pixel (p), and c is a predetermined constant; Identify the outline of the target region (TR). 27 . The inspection apparatus according to claim 26 , wherein the predetermined threshold is 1 and the predetermined constant is 2.

28. An examination device according to claim 26 or 27, configured to obtain a fluorescence image (FI) representing background in addition to mammalian tissue (MT), the device being configured to: A preliminary image segmentation is performed to distinguish in the fluorescence image (FI) a foreground region (FR) representing the mammalian tissue (MT) and a background region (BR) representing the background, wherein the device is configured to determine, for each pixel, whether it belongs to the foreground region (FR) if the fluorescence signal value (FSV) significantly exceeds the background mean fluorescence signal value taking into account the standard deviation of the fluorescence signal values ​​(FSV) in the background, otherwise determine that the pixel belongs to the background region (BR). 29 . The inspection apparatus according to claim 28 , further configured to obtain a calibration fluorescence image containing only background before obtaining the fluorescence image, and to determine a mean fluorescence signal value and a standard deviation from the calibration fluorescence image.

30. An inspection apparatus according to claim 26, configured to determine (S2A) a mean fluorescence signal value and a standard deviation of fluorescence signal values ​​from a portion (BP) of the fluorescence image (FI) designated as representing background.

31. The examination apparatus according to any one of claims 26 to 30, configured to obtain fluorescent images of mammalian tissue taken in vivo.

32. An examination apparatus according to any one of claims 26 to 31, configured to obtain a fluorescent image taken in vitro of mammalian tissue placed on a background.

33. The inspection device according to any one of claims 26 to 32, configured to perform (S5) a dilation operation after preliminary image segmentation.

34. The inspection apparatus of any one of claims 26-33, wherein the predetermined k-th quantile is a k-th percentile, wherein k is in the range of 25 to 75.

35. The inspection apparatus of claim 34, wherein the percentile is a median.

36. The inspection device according to any one of claims 26 to 35, further configured to perform a scan trajectory-based evaluation comprising: obtaining (S10) at least one fluorescence signal value vector in the fluorescence image (FI) along a scanning trajectory; For each threshold value among the plurality of threshold values, determining (S11) a corresponding set of candidate scanning trajectory segments whose fluorescence signal values ​​exceed the threshold value; It is determined (S13) which threshold value among a plurality of threshold values ​​has a corresponding set of candidate scanning trajectory segments that best matches a second image segmentation of the fluorescence image (FI) along the scanning trajectory.

37. An inspection apparatus according to claim 36, configured to repeat the procedure for a plurality of mutually different scanning trajectories in a set of scanning trajectories.

38. An inspection apparatus according to claim 37, configured to repeat the procedure described therein for a plurality of sets of scanning trajectories.

39. The inspection device according to any one of claims 26 to 38, further comprising removing (S16) isolated points.

40. The inspection device according to any one of claims 26 to 39, further configured to: determining ( S20 ) at least one principal axis of at least one contour; obtaining (S22) at least one fluorescence signal value vector in the fluorescence image (FI) along a scan line orthogonal to the main axis; Based on information of the contour at the intersection with the scan line, selectively performing (S23) identifying one or more segments of the scan line, wherein the signal-to-noise ratio exceeds a predetermined level, the scan line segments each including a corresponding first endpoint indicating a transition (NT) from normal tissue to tumor tissue or infected tissue and a corresponding second endpoint indicating a transition (TN) from tumor tissue or infected tissue to normal tissue.

41. The inspection apparatus of claim 40, wherein the principal axis is defined as a line segment that minimizes a distance metric between the line segment and a point of the contour.

42. The inspection device according to claim 40 or 41 is configured to rotate the fluorescence image (FI) before obtaining (S22) the at least one fluorescence signal value vector, wherein the rotation aligns the main axis of the contour with the coordinate axis of the fluorescence image (FI), and wherein the scan line is along another coordinate axis direction of the fluorescence image (FI).

43. The inspection apparatus according to claim 40, 41 or 42, wherein the information on which the selective execution (S23) is based comprises the intensity of the fluorescence image (FI) at the intersection points of the contours.

44. An inspection apparatus according to claim 40, 41 or 42, wherein the information on which the selective execution (S23) is based comprises the coordinates of the intersection point.

45. An inspection device according to any one of claims 40 to 44, configured to combine (S24) and indicate the coordinates of points of a plurality of contours in a single image.

46. ​​The inspection apparatus according to claim 45, configured to perform a reverse rotation according to a major axis of the profile before performing the combining (S24).

47. The inspection device according to any one of claims 26 to 46, further configured to selectively remove isolated candidate points.

48. The examination device according to any one of claims 26 to 43, further comprising a camera (6) for obtaining a fluorescence image (FI) of mammalian tissue (MT).

49. The examination apparatus according to claim 48, further comprising an excitation light source (7) for illuminating the mammalian tissue.

50. An inspection device according to any one of claims 26 to 49, wherein the at least one contour comprises a main contour (C) indicating a boundary between a target region and a reference region.

51. An inspection device according to any one of claims 26 to 50, wherein the at least one contour comprises a secondary contour (C') extending a distance outside a boundary of the target area and the reference area.

52. An examination device according to claim 51, wherein the secondary contour (C') extends a distance outside the border in such a way as to avoid intersections with specific anatomical structures.

53. The inspection device according to claim 26, further configured to: obtaining at least one sequence of respective logarithmic fluorescence signal values ​​for consecutive pixels arranged along a scanning trajectory within the fluorescence image, wherein the scanning trajectory begins at a location identified as part of a reference region in the fluorescence image; calculating an indicator of local linearity of the log fluorescence signal values ​​at positions in the sequence; A position in the sequence where the indicator indicates that the logarithmic fluorescence signal value as a function of position in the sequence is no longer linear is identified, the identified position being a candidate boundary position for a boundary between normal tissue and tumor tissue or infected tissue.

54. A medical treatment device (100) comprising an examination device and a treatment device (5) according to claim 49, the treatment device being used to remove or irradiate a tumor according to a constructed contour, or to activate a treatment substance within a range specified by the constructed contour.

Images