System, device and method for turbidity analysis

Abstract

To solve the problem in which a view of a biological structure provided by an imager is limited when an imaging environment is cloudy or turbid.SOLUTION: An endoscopic system includes: an endoscopic imager configured to capture images of a target site within a living body; and a processor configured to determine one or more image metrics for each multiple image frames captured over a time span, analyze changes in the image metrics over the time span, and determine a turbidity metric for the target site based on the analyzed changes in the image metrics.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] [Claim of priority] This disclosure claims priority to U.S. Provisional Patent Application No. 62 / 904,882, filed Sep. 24, 2019, the disclosure of which is incorporated herein by reference.

[0002] The present disclosure relates to endoscopic procedures, and more particularly to systems, devices and methods for performing turbidity analysis of an endoscopic imaging environment. [Background technology]

[0003] Endoscopic imagers may be used during various medical interventions. A cloudy or turbid imaging environment limits the view of anatomy provided by the imager. Cloudiness may be caused by blood, urine, or other particles. In some endoscopic procedures (e.g., ureteroscopic procedures), the cloudy imaging environment may be managed by a fluid management system that circulates fluid within the imaging cavity. Summary of the Invention [Means for solving the problem]

[0004] The present disclosure relates to an endoscopic system that includes an endoscopic imager configured to capture image frames of a target site within a living body, and a processor configured to determine one or more image metrics for each image frame of a plurality of image frames captured over a period of time, analyze changes in the image metrics over the period of time, and determine a turbidity metric of the target site based on the analyzed changes in the image metrics.

[0005] In one embodiment, the image metrics are image entropy metrics that include a red entropy metric and a cyan entropy metric.

[0006] In an embodiment, the processor is further configured to estimate blood content in the current image frame and modify the current image frame to reduce the visual effect of the blood content to enhance the remainder of the current image frame.

[0007] In some embodiments, the processor is further configured to identify and classify particles within the current image frame.

[0008] In an embodiment, the endoscopic system further includes a display configured to annotate the current image frame with the turbidity metric.

[0009] In an embodiment, the display is further configured to bracket identified particles and annotate the current image frame with the particle classification.

[0010] In one embodiment, the identified particles are kidney stones and the classification relates to the size of the kidney stones.

[0011] The present disclosure also relates to an endoscopic system including an endoscopic imager configured to capture image frames of a target site within a living body, a fluid delivery mechanism that supplies irrigation fluid to the target site to clarify the field of view of the endoscopic imager, and a processor configured to determine a turbidity metric of at least one image from the imager, determine a fluid delivery adjustment of the irrigation fluid based on the turbidity metric, and control the fluid delivery mechanism to adjust the fluid delivery supplied by the fluid delivery mechanism based on the determined fluid delivery adjustment.

[0012] In an embodiment, the processor is further configured to determine a type of intervention activity based on feature detection identifying the interventional instrument.

[0013] In an embodiment, the processor is further configured to determine a phase of the intervention activity and adjust the fluid delivery provided by the fluid delivery mechanism based on the phase of the intervention activity.

[0014] In an embodiment, the processor is further configured to identify and classify particles within the current image frame, and adjust the fluid delivery provided by the fluid delivery mechanism based on the particle classification.

[0015] In certain embodiments, the particles are kidney stones and the particle classification relates to the size of the kidney stones.

[0016] In an embodiment, the processor is further configured to determine a blood metric of the at least one image and adjust the fluid delivery provided by the fluid delivery mechanism based on the blood metric.

[0017] In an embodiment, the processor is further configured to determine an image entropy metric for the at least one image.

[0018] In one embodiment, the turbidity metric and the blood metric are determined based in part on the image entropy metric.

[0019] The present disclosure also relates to a method that includes determining one or more image metrics for each image frame of a plurality of image frames of a target site in a living body captured over a period of time, analyzing changes in the image metrics over the period of time, and determining a turbidity metric of the target site based on the changes in the analyzed image metrics.

[0020] In one embodiment, the image metrics are image entropy metrics that include a red entropy metric and a cyan entropy metric.

[0021] In an embodiment, the method further includes estimating blood content in the current image frame and modifying the current image frame to reduce the visual effect of the blood content to enhance the remainder of the current image frame.

[0022] In some embodiments, the method further includes identifying and classifying particles in the current image frame.

[0023] In an embodiment, the method further includes annotating the current image frame with a turbidity metric. [Brief description of the drawings]

[0024] [Figure 1] FIG. 1 illustrates a system for performing an endoscopic procedure, according to various exemplary embodiments of the present disclosure. [Diagram 2] FIG. 1 illustrates a method for managing cleaning fluid flow in a closed loop feedback system according to various exemplary embodiments of the present invention. [Diagram 3] 1 illustrates a method for enhancing visibility through a liquid medium clouded by blood according to a first exemplary embodiment. [Figure 4] 11 illustrates a method for enhancing visibility through a liquid medium clouded by blood, according to a second exemplary embodiment. FIG. [Diagram 5] 5 is a flow chart illustrating an embodiment combining the methods of FIGS. 2 to 4. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] The present disclosure may be further understood with reference to the following description and the accompanying drawings, in which like elements are referred to by the same reference numerals. Exemplary embodiments describe algorithmic improvements for managing haze within an endoscopic imaging environment. A typical urological procedure utilizes an imaging device (e.g., a ureteroscope or other endoscopic imager), a fluid delivery mechanism (to clear the view and / or widen or distend the body cavity), and a treatment mechanism (e.g., a Boston Scientific LithoVue™ device, and / or a laser or RF energy source, etc.).

[0026] The improvements described herein include, for example, methods for analyzing endoscopic images to determine turbidity and providing this information to a fluid management system to manage clarity within the field of view. Other improvements include algorithmically subtracting blood features from endoscopic images and amplifying image features obscured by blood, particularly for urological procedures. Some common urological procedures include kidney stone management (e.g., lithotripsy), BPH (i.e., benign prostatic hyperplasia) treatment (e.g., GreenLight™ laser surgery), prostatectomy, bladder tumor resection, uterine fibroid management, diagnostics, etc., but one of ordinary skill in the art will appreciate that image-improving devices and techniques can be used in a variety of procedures (i.e., including non-urological procedures) where turbidity is an issue.

[0027] 1 illustrates a system 100 for performing an endoscopic procedure, according to various exemplary embodiments of the present disclosure. The system 100 includes an endoscope 102 having an imager 104 for acquiring image frames of an anatomical site within a living body during an endoscopic procedure, and a fluid delivery mechanism 106 for providing a fluid (e.g., saline) to the anatomy to remove blood and debris that may obscure the view of the imager 104. The fluid delivery mechanism 106 also simultaneously provides suction to remove fluid from the anatomy. In this manner, the anatomy is continually refreshed with a substantially transparent fluid to produce clearer images.

[0028] The system 100 may further include a therapeutic device 108 selected depending on the nature of the endoscopic procedure. The therapeutic device 108 may operate through the endoscope 102 or may be external to the endoscope 102. For example, the therapeutic device 108 may be a laser or shock wave generator, for example to break up kidney stones, or a resectoscope for removing prostate tissue. If the urological procedure is intended to be diagnostic, i.e. to examine the anatomy rather than to treat a symptom, a therapeutic device may not be used. The exemplary embodiments are described with respect to urological imaging, but are not limited thereto. Some embodiments are equally applicable to a wide range of procedures, such as endoscopic procedures, for example in the digestive system, including endoscopic procedures that do not include a fluid delivery mechanism.

[0029] The system 100 includes a computer 110 that processes image frames provided by the imager 104 and provides processed images on a display 112. The computer 110 and display 112 are provided in an integrated station, such as an endoscope console in this embodiment. The endoscope console may also implement other functionality for performing urological procedures, including, for example, actuators that control the flow rate, pressure, or delivery method of fluid delivered through the fluid delivery mechanism 106. The illustrative embodiment describes an algorithmic process that alters and enhances the displayed image, generally continuously or in any other desired manner.

[0030] In one embodiment, image metrics determined from captured image frames are used to estimate the degree to which a video sequence is hidden or occluded by a murky imaging or imaging environment. The image metrics include image entropy metrics. Image entropy is generally defined as a measure of the amount of information or quantity of information in an image and can be approximated by evaluating the frequency of intensity values ​​in the image. High entropy can reflect, for example, a large amount of anatomical detail associated with a clear image, or a swirling cloud of particles obscuring the anatomy, associated with an unclear image. Changes in the entropy measurement over time can help distinguish, for example, between high entropy clear images and high entropy unclear images. For example, entropy measurements over time for a series of images distinguish these two types of images through the variability of entropy in the unclear images.

[0031] Low entropy, on the other hand, reflects a loss of contrast and / or blurring of details. Low entropy may result from a very simple scene (e.g., a plain white wall). The amount of image entropy can be measured using image entropy metrics, including the total entropy in the image, or the ratio of red entropy to cyan entropy in the image. In some embodiments, the ratio of red entropy to cyan entropy can emphasize the contribution of blood (represented by red entropy) over other fluids (represented by cyan entropy), emphasizing the blood contribution. In some embodiments, the image entropy metric can further include the entropy variation at a particular time frequency band in the sequence of images. As mentioned above, the entropy variation can be measured over time to distinguish between clear and unclear images. Specifically, abrupt changes in high entropy measurements of an observed image reflect the chaotic entropy associated with a swirling cloud of particles, and therefore signify that the observed image is an unclear image.

[0032] To characterize the changing scene between images, optical flow analysis is utilized. In general, optical flow analysis identifies nearby regions in a given image that correspond to nearby regions in a previous image and identifies the difference in their positions. One such algorithm that estimates such misregistration is the Farneback algorithm. Such analysis provides information about stationary and moving objects in the field of view, as well as systematic field of view motion, i.e., camera panning, rotation, advancement and retreat.

[0033] Machine learning systems can also be employed to characterize and classify video segments according to occlusion levels or scene content. For example, machine learning methods such as neural networks, convolutional neural networks, optimizers, linear regression and / or logistic regression classifiers can be used to discover novel image analysis and / or combinations of the entropy metrics described above to characterize and classify video segments. Machine learning methods can generate scalar estimates of turbidity, as well as blood and particle-field probabilities.

[0034] In another embodiment, other metrics can be used to classify the video segments, such as spatial and time frequency decomposition. Spatial frequency analysis can directly reveal image clarity, effectively measuring the sharpness of the image. Images with particulate and hazy vision may exhibit relatively high spatial frequencies, which are similar to those of images with high anatomical details. Distinguishing between images with high spatial frequencies and images with hazy vision and high anatomical details can be done using time frequency analysis. As mentioned above, time frequency analysis means tracking the changes in a metric over a period of time.

[0035] As described in more detail below, the machine learning system can optimize the nonlinear combiners and various assessments, including particle / stone assessment, blood assessment, and clarity assessment. Various metrics can be derived through linear combinations of pixel values, intensity histogram values, nonlinear functions of these (e.g., logarithms), and time sequences of these values. The machine learning system can use algorithms to find alternative combinations of the above values ​​that do not directly correspond to spatial frequency or entropy but effectively correspond to the observer's impression of turbidity.

[0036] Machine learning systems operate on the feature space they present (i.e., the system's values ​​determine the classification). In one embodiment, a video machine learning system has video pixels input directly into the system. In a typical machine learning model, image metrics (e.g., spatial frequency spectrum) are calculated from these values. The feature space can be augmented by providing system analyses that are useful but not suitable for computation in machine learning systems. For example, entropy, which would be useful for this computation, is not easily computed in the architecture of many common video processing neural networks. As a result, machine learning systems can perform better when provided with other analyses in addition to the raw pixels. Providing intensity histogram data, including logarithms, allows the neural network to incorporate entropy and entropic measures, which can quickly converge to better results. The system can identify mathematical variations of the metrics as superior. Thus, performance can be dramatically improved by carefully augmenting the feature space.

[0037] Certain metrics directly derivable from image data do not by themselves do a suitable job of estimating turbidity in a clinical environment where the image background is constantly changing. Exemplary embodiments describe methods to inform and tune neural networks, for example, entropy-related metrics that estimate turbidity levels from image data.

[0038] The metrics and analysis described above can be output to a display for annotating the displayed image frames and / or input into appropriately designed machine learning algorithms that perform multiple evaluations that can inform further analysis.

[0039] In a first example, a particle or stone assessment is performed. The particle / stone assessment determines the characteristics of particles suspended in the fluid. Feature detection can be used to determine the nature of the particle. For example, a kidney stone can be identified as such and metrics such as the size of the stone, as well as a depth estimate (i.e., distance from the imager) can be determined. The movement of the stone can also be estimated based in part on the optical flow analysis described above. Other particles can include, for example, proteins, tissue, protrusions, blood clots, etc. The particle / stone assessment identifies particles, segments them within the image frame, and sizes them.

[0040] This particle identification and classification can be used to annotate the current image frame. For example, particles can be bracketed and associated metrics displayed to inform the surgeon during the procedure, allowing for quicker decision making. Particle / stone analysis can also be used to manage the delivery of irrigation fluid by the fluid delivery mechanism 106. For example, a cloudy field of view typically suggests the need to increase fluid flow, while identification of a mobile stone can suggest attenuating fluid flow to keep the stone location within the imager's field of view. Particle / stone assessment can also directly or indirectly influence the use of the treatment device 108 depending on the nature of the treatment. For example, in lithotripsy, classification of the stone as, for example, too large to be retrieved, too large to pass naturally, or small enough to pass naturally can automatically (or under the supervision of the physician) drive or influence the operation of the system during the remainder of the intervention.

[0041] In a second example, a clarity assessment is performed. The clarity assessment determines the total turbidity measure in the image. As described above, a machine learning process utilizing, for example, image entropy metrics as input can generate a new turbidity measure or measurement. The turbidity measure can be, for example, a scalar estimate of turbidity, or some other metric developed by a machine learning algorithm. The clarity assessment can be used to annotate the currently displayed image or to manage the supply of cleaning fluid.

[0042] In a third example, a blood assessment is performed. The blood assessment estimates the amount of blood or total blood content measurement in the cavity, e.g., in parts per million (PPM), although other metrics can be used. The blood assessment can be used to annotate the currently displayed image or to manage the supply of irrigation fluid. The blood assessment can also inform, e.g., clot analysis, laser settings for BPH treatment, etc.

[0043] In addition to the particle, clarity and blood assessments mentioned above, a fourth assessment can also be used to manage in vivo the fluid circulation in the imaged cavity. The fourth assessment relates to the type of intervention performed during the endoscopic procedure. For example, the intervention can be kidney stone management, prostatectomy, uterine fibroid management, etc. A feature detector can be applied to the current image frame to determine the type of intervention activity in progress. For example, during laser prostatectomy, a laser device is present within the FOV of the imager, which can be identified as a laser device by the feature detector.

[0044] In another example, a stone retrieval basket device can be identified, such as by a feature detector. Once identified, the phase of the intervention is evaluated. For example, the system can measure the degree to which the stone has been fragmented or otherwise reduced in size, and / or determine when the stone has been received in the basket for retrieval, etc. Thus, the intervention evaluation can be used in combination with at least a particle / stone evaluation, etc., during a laser lithotripsy procedure to identify phases of the intervention that may affect the desired flow characteristics (e.g., no stone in field of view, stone in field of view, laser irradiation). The type of intervention may affect the optimal flow rate of the irrigation system fluid. For example, a laser prostatectomy may require a higher flow rate than, for example, a kidney stone procedure, due to heating of the tissue. Heating of the tissue may cause a complementary heating of the fluid in the cavity. A more rapid inflow and outflow of fluid can act to maintain a more stable temperature in the cavity and prevent damage to healthy tissue.

[0045] 2 illustrates a method 200 for managing the delivery of irrigation fluid with a closed-loop feedback system, according to various exemplary embodiments of the present invention. Method 200 may employ some or all of the metrics described above. Some of the described metrics may be derived from a single image frame, while others may be derived from a sequence of image frames. Thus, closed-loop adjustments of fluid delivery may be performed based on multiple image frames. However, given the fast frame rates of imagers typically used in the industry, from the perspective of the surgeon during surgery, the fluid delivery adjustments are applied quickly enough to appear substantially continuous.

[0046] At 205, imager 104 captures a sequence of image frames. As noted above, the total number of images required to perform a given calculation may vary, and therefore the number of images in the sequence may also vary.

[0047] At 210, some or all of the above metrics are determined from one or more image frames in the sequence. For example, the image entropy metrics may be derived from each image as they are captured. In another example, pixel metrics are determined directly from the images.

[0048] At 215, a machine learning combiner algorithm / function is used to process the various metrics and analyses to characterize and classify the image or video segment according to occlusion level and scene content. As discussed above, the machine learning algorithm can combine the metrics in various ways and self-adjust the weighting or use of the metrics according to the analyses discovered. Note that the metrics / analyses discussed above are all input to the machine learning combiner algorithm, except for treatment feature detection and intervention phase assessment, whereas treatment feature / phase detection / assessment are determined directly from the image frames and do not inform particle assessment, blood assessment or clarity assessment.

[0049] At 220, the flow management algorithm processes the particle assessment, blood assessment, clarity assessment, and intervention phase assessment to determine whether a fluid delivery adjustment of the irrigation fluid is necessary, and if so, the adjustment amount.

[0050] At 225, the adjustment value is fed back to a processor to adjust the flow provided by the fluid delivery mechanism.

[0051] Method steps 205-225 are performed substantially continuously to provide a closed loop feedback system for managing fluid flow during an endoscopic intervention.

[0052] The metrics mentioned above, especially those related to blood assessment and clarity assessment, can be used to algorithmically provide image enhancement to compensate for bloody fields of view in endoscopic interventions. Blood clouding can be managed by algorithmically subtracting blood features from the image and amplifying features obscured by blood. Methods 300 and 400 described below can be applied alone or in combination.

[0053] 3 illustrates a method 300 for enhancing visibility through a liquid medium clouded by blood, according to a first exemplary embodiment. Method 300 includes some steps of method 200 for managing a supply of irrigation fluid (or steps similar to those of method 200) and may be used alone or in combination therewith.

[0054] At 305, the imager 104 captures a series of images according to step 205 of the method 200. At 310, each image is divided into regions of a fixed size. Each region may be, for example, 20x20 pixels.

[0055] At 315, similar to steps 210-215 of method 200, several metrics as described above are calculated and processed for each region using the machine learning combiner algorithm described above. In this manner, each region is evaluated for blood volume or blood content to generate a "blood map" that characterizes each region according to the presence of blood. Specifically, the metrics to provide to the machine learning combiner algorithm for blood evaluation are determined and processed to classify each region. It is not necessary to perform particle / stone evaluation or clarity evaluation or to determine any interventional device characteristics for method 300. However, in another embodiment, a clarity evaluation can be performed, for example, when particles other than blood cause a cloudy field of vision.

[0056] At 320, the blood map values ​​are multiplied in a combiner with the optical flow adjusted local frame-to-frame variations of each metric related to blood detection. In this manner, individual image sub-regions can be utilized to identify areas obscured by blood and selectively or proportionally apply image enhancements to mitigate the loss of image detail caused by blood.

[0057] At 325, the convolution kernel or color transform calculated to complement each metric is scaled and applied to the region using the values ​​obtained from 320, resulting in a sigmoidal reduction at the region boundaries. The entropy-related kernel uses either a low-pass filter kernel (for positive entropy correlation) or a high-pass filter kernel (for negative entropy correlation) applied to the associated color channel.

[0058] At 330, an image frame is reconstructed from the spatial frequency representation of each region, consisting of phase and altered amplitude, to de-emphasize the foreground containing blood and to emphasize the background of interest. The visual effect of blood can be reduced to different degrees across the image depending on the scaling of the convolution kernel. This enhancement is applied to the magnitude components of the spatial frequency, which can then be recombined with the phase components and transformed back into a conventional image.

[0059] FIG. 4 illustrates a method 400 for enhancing visibility through a liquid medium clouded by blood, according to a second exemplary embodiment.

[0060] At 405, a deformable nonlinear transformation is used to correlate the current image frame with the super-resolution image map. A super-resolution image map can be generally defined as an image of improved resolution generated from the fusion of a lower resolution image. The current image can be fused with the super-resolution image map to increase the resolution of the current image.

[0061] At 410, based on the blood detection related metrics described above, a "blood mask" is created, which reflects the spatial distribution of estimated blood turbidity in a given image frame.

[0062] At 415, the current image frame is fused with the transformed super-resolution map to generate an enhanced frame that is a weighted average of the current frame and the super-resolution map. When the blood mask value is low, the pixel is encouraged toward a pixel in the original image frame, and when the blood mask value is high, the pixel is encouraged toward a pixel in the super-resolution map.

[0063] 5 is a flowchart 500 illustrating a combined embodiment that includes elements of methods 200, 300, and 400. Those skilled in the art will appreciate that the various metrics and analyses described above may be used alone or in combination in various ways.

[0064] At 505, images are extracted from the endoscopic imaging feed. As noted above, some metrics / analyses, such as optical flow analysis, require many images to determine. However, flowchart 500 can assume that enough previous images have been captured to perform each calculation.

[0065] At 510, the image is separated into red and cyan components to calculate various metrics. For example, at 515, red and cyan entropy values ​​are calculated. Metrics such as the red to cyan entropy ratio can be calculated or the red to cyan values ​​can be used directly. At 520, spatial frequency analysis is performed on the red and cyan spectral components. At 525, the red / cyan entropy values ​​and spatial frequency analysis are provided to a time frequency analysis. Those skilled in the art will appreciate that time frequency analysis requires a series of photographs to analyze the differences between the photographs. For example, the entropy variation is measured over a period of time. At 530, a color balance is performed to compare the red and cyan components. At 535, a pixel change rate analysis is performed to compare pixel values ​​between images. At 540, an optical flow analysis is performed to correlate successive images to derive information such as the movement of objects or tissues over a period of time.

[0066] At 545, the machine learning system combines the above metrics / analyses in various ways to generate particle information (Particle / Stone Assessment 550), blood volume or content information (Blood Assessment 555), and turbidity information (Clarity Assessment 560). An intervention assessment is also performed to identify intervention features in the image (Treatment Device Feature Detection 565) and assess the phase of the intervention (Treatment Phase Assessment 570). The particle, blood, clarity, and intervention analyses are used to control the fluid flow 575 of the irrigation system. As described above, fluid flow management 575 can include adjusting flow rate, pressure, or fluid delivery method.

[0067] At 580, the current image may be enhanced by removing or reducing blood content of the image according to methods 300 or 400. At 585, the current image may be annotated according to method 200. The current image may be further annotated with blood or turbidity metrics derived in the blood / clarity assessment. At 590, the enhanced / annotated image is displayed. As further images are captured, the steps / metrics / analysis described above are performed or determined.

[0068] In another embodiment, a computer readable medium includes instructions that, when executed by a computer, cause the computer to perform the various image processing steps / analyses described above, such as determining turbidity metrics.

[0069] Those skilled in the art will appreciate that changes may be made to the embodiments described above without departing from the inventive concept. It will further be understood that structural features and methods associated with one embodiment may be incorporated in other embodiments. It is understood, therefore, that the invention is not limited to the specific embodiments disclosed, but rather that modifications are intended to be included within the scope of the invention as defined by the appended claims.

Claims

1. 1. An endoscopy system comprising: an endoscopic imager configured to capture image frames of a target site within a living body; a processor; The processor, determining one or more image metrics for each image frame of a plurality of image frames captured over a period of time; Analyzing changes in the image metrics over the time period; determining a turbidity metric of the target site based on the changes in the analyzed image metrics; The endoscope system is configured as follows.

2. The endoscopic system of claim 1 , wherein the image metrics are image entropy metrics including a red entropy metric and a cyan entropy metric.

3. The processor, Estimating the blood content in the current image frame; The endoscopic system of claim 2 , further configured to modify the current image frame to reduce the visual effect of the blood content and enhance a remainder of the current image frame.

4. The endoscopic system of claim 1 , wherein the processor is further configured to identify and classify particles within a current image frame.

5. The endoscopic system of claim 4 , further comprising a display configured to annotate the current image frame with the turbidity metric.

6. The display comprises: The identified particles are enclosed in brackets, annotating the current image frame with a particle classification; The endoscopic system of claim 5 , further configured as follows:

7. The endoscopic system of claim 6 , wherein the identified particle is a kidney stone and the classification relates to a size of the kidney stone.

8. 1. An endoscopy system comprising: an endoscopic imager configured to capture image frames of a target site within a living body; a fluid delivery mechanism for providing irrigation fluid to the target site to clarify the field of view of the endoscopic imager; a processor; The processor, determining a turbidity metric of at least one image from the endoscopic imager; determining a fluid delivery adjustment of the cleaning fluid based on the turbidity metric; controlling the fluid delivery mechanism to adjust the fluid delivery provided by the fluid delivery mechanism based on the determined fluid delivery adjustment. The endoscope system is configured as follows.

9. The endoscopic system of claim 8 , wherein the processor is further configured to determine a type of interventional activity based on feature detection identifying an interventional instrument.

10. The processor, determining a phase of said intervention activity; The endoscopic system of claim 9 , further configured to adjust fluid delivery provided by the fluid delivery mechanism based on a phase of the interventional activity.

11. The processor, Identify and classify particles in the current image frame; The endoscopic system of any one of claims 8 to 10, further configured to adjust fluid delivery provided by the fluid delivery mechanism based on particle classification.

12. The endoscopic system of claim 11 , wherein the particles are kidney stones and the particle classification relates to a size of the kidney stones.

13. The processor, determining a blood metric of the at least one image; The endoscopic system of any one of claims 8 to 11, further configured to adjust fluid delivery provided by the fluid delivery mechanism based on the blood metric.

14. The endoscopic system of claim 13 , wherein the processor is further configured to determine an image entropy metric of the at least one image.

15. The endoscopic system of claim 14 , wherein the turbidity metric and the blood metric are determined based in part on the image entropy metric.

Images