CN116364265B - Medical endoscope image optimization system and method - Google Patents

Abstract

The invention provides a medical endoscope image optimization system and a method, wherein the system comprises the following steps: an inspection image trajectory graph generation module for generating an inspection image trajectory graph of the endoscope when the endoscope is ready to be withdrawn from the body of the patient; the endoscope image optimization demand prediction module is used for predicting the endoscope image optimization demand of a user; the preprocessing module is used for preprocessing the examination image track graph based on the optimization requirement of the endoscope image and outputting and displaying the preprocessing result. When the endoscopic examination of the patient is finished, the method generates the examination image track map, and a doctor can intuitively determine which position in the body of the patient corresponds to the snapshot image by checking the examination image track map, so that global analysis is facilitated.

Description

Medical endoscope image optimization system and method

Technical Field

The invention relates to the technical field of image data processing, in particular to a medical endoscope image optimization system and method.

Background

When a patient performs an endoscopy, a doctor uses an endoscope to visit the body of the patient, the lens of the endoscope returns an internal image of the body of the patient, the doctor performs necessary image snapshot according to the internal image, and finally performs analysis according to the snapshot image and gives an examination conclusion.

However, when a doctor analyzes according to a snap shot image, the analysis flow is complicated: the parameters of the images (such as brightness, chromaticity, saturation, etc.) of the snap-shot images need to be adjusted, the positions (such as cecum, transverse colon, descending colon, etc.) of the snap-shot images corresponding to the positions in the body of the patient need to be determined through recall or view, disease conditions need to be assessed according to the snap-shot images (such as analyzing intestinal ulcer areas, etc.), and in addition, doctors need to conduct global analysis according to the position distribution of each snap-shot image corresponding to the position in the body of the patient and the disease conditions corresponding to each snap-shot image (such as analyzing whether the disease conditions of adjacent parts in the body of the patient have relevance, etc.).

The more cumbersome analysis procedure reduces the efficiency of the physician in making the examination conclusion, and thus a solution is needed.

Disclosure of Invention

The invention aims to provide a medical endoscope image optimizing system, when the endoscopic examination of a patient is finished, an examination image track diagram is generated, a doctor can visually determine which position in the body of the patient corresponds to the snapshot image by checking the examination image track diagram, can also visually determine the position distribution in the body of the patient corresponding to each snapshot image and the disease condition corresponding to each snapshot image, so that global analysis is convenient, in addition, the examination image track diagram is preprocessed according to the endoscope image optimizing requirement of the doctor, the doctor does not need to process the examination image track diagram by itself, the analysis efficiency of the doctor for analyzing according to the snapshot images is improved, and meanwhile, the medical endoscope image optimizing system is more humanized.

The embodiment of the invention provides a medical endoscope image optimization system, which comprises:

an inspection image trajectory graph generation module for generating an inspection image trajectory graph of the endoscope when the endoscope is ready to be withdrawn from the body of the patient;

the endoscope image optimization demand prediction module is used for predicting the endoscope image optimization demand of a user;

the preprocessing module is used for preprocessing the examination image track graph based on the optimization requirement of the endoscope image and outputting and displaying the preprocessing result.

Preferably, the examination image trajectory graph generating module generates an examination image trajectory graph of the endoscope, and performs the following operations:

acquiring a detection track, a snap-shot image and corresponding snap-shot parameters when the endoscope enters the body of a patient; the snapshot parameters include: a snapshot position and a snapshot direction;

determining the track point position corresponding to the snapshot position from the detection track;

constructing a first direction vector based on the track point position and the snapshot direction;

randomly setting the snap-shot images beside the track point positions;

acquiring the center position of an image of the snap shot image and the opposite direction of the image;

constructing a second direction vector based on the image center position and the image facing direction;

calculating a first vector included angle between the first direction vector and the second direction vector;

acquiring a first linear distance between a track point position and an image center position;

position adjustment is carried out on the snap shot image until the position adjustment is stopped when the first vector included angle is equal to a preset first included angle value and the first linear distance is equal to a preset first distance value;

an inspection image trajectory graph is generated based on the penetration trajectory.

Preferably, the endoscope image optimization demand prediction module predicts an endoscope image optimization demand of a user, and performs the following operations:

Acquiring a generated voice fragment and corresponding voice start-stop time when a user enters the body of a patient by using an endoscope;

determining the time interval position corresponding to the starting and ending time of the voice from a preset time axis;

setting the voice segment at the time interval position;

acquiring snapshot time of a snapshot image;

determining a first time position corresponding to the snapshot time from a time axis;

a second time position which accords with the time position condition within a preset time period before the first time position on the time axis;

determining a local human voice segment between a leftmost second time position and a first time position on a time axis;

carrying out semantic extraction on the local voice segments to obtain first voice semantics;

matching the first human voice semantics with second human voice semantics in a preset characterization semantic library;

when the matching is met, taking the matched first human voice semantic as a first target semantic, taking the matched second human voice semantic as a second target semantic, and acquiring a preset representation requirement item and a preset representation type corresponding to the second target semantic; the characterization types include: individual and combined characterizations;

when the characterization type of the second target semantic is single characterization, taking the characterization requirement item as an endoscope image optimization requirement;

When the representation type of the second target semantics is the combination representation, acquiring a preset combination representation semantics library corresponding to the second target semantics;

matching any third human voice semantics in the combined representation semantic library with first human voice semantics except the first target semantics;

when the two images are matched and met, taking the representation requirement item as an endoscope image optimization requirement;

wherein the time position condition includes:

and checking that the track corresponding to the second time position exists in the image track graph.

Preferably, the preprocessing module performs preprocessing on the examination image trajectory graph based on the optimization requirement of the endoscope image, and performs the following operations:

determining a preprocessing requirement corresponding to the endoscope image optimization requirement from a preset preprocessing requirement library; the pretreatment requirements include: image requirements, image processing requirements, and image analysis requirements;

determining a snap image meeting the image requirements from the checked image track diagram and taking the snap image as a target image;

only the target image is reserved in the inspection image track diagram;

determining an image processing template corresponding to the image processing requirement from a preset image processing template library;

performing image processing on the target image based on the image processing template;

Determining an image analysis template corresponding to the image analysis requirement from a preset image analysis template library;

based on the image analysis template, performing image analysis on the target image to obtain an image analysis result;

mapping the image analysis result into a preset information frame, associating the information frame with a corresponding target image, and setting the information frame beside the corresponding target image.

Preferably, the medical endoscope image optimization system further comprises:

a position adjustment module for including:

acquiring the center position of the information frame and the opposite direction of the information frame;

constructing a third direction vector based on the center position of the information frame and the opposite direction of the information frame;

acquiring a viewing position and a viewing direction of a user viewing information frame;

constructing a fourth direction vector based on the viewing position and the viewing direction;

calculating a second vector angle between the third direction vector and the fourth direction vector;

acquiring a second linear distance between the center position of the information frame and the image center position of the target image associated with the information frame;

and performing position adjustment on the information frame until the position adjustment is stopped when the second vector included angle is equal to a preset second included angle value and the second linear distance is equal to a preset second distance value.

The medical endoscope image optimization method provided by the embodiment of the invention comprises the following steps:

step S1: generating an inspection image trajectory graph of the endoscope when the endoscope is ready to be withdrawn from the body of the patient;

step S2: predicting an endoscopic image optimization requirement of a user;

step S3: based on the optimization requirement of the endoscope image, preprocessing the examination image track diagram, and outputting and displaying the preprocessing result.

Preferably, in step S1, generating an inspection image trajectory chart of the endoscope includes:

acquiring a detection track, a snap-shot image and corresponding snap-shot parameters when the endoscope enters the body of a patient; the snapshot parameters include: a snapshot position and a snapshot direction;

determining the track point position corresponding to the snapshot position from the detection track;

constructing a first direction vector based on the track point position and the snapshot direction;

randomly setting the snap-shot images beside the track point positions;

acquiring the center position of an image of the snap shot image and the opposite direction of the image;

constructing a second direction vector based on the image center position and the image facing direction;

calculating a first vector included angle between the first direction vector and the second direction vector;

acquiring a first linear distance between a track point position and an image center position;

Position adjustment is carried out on the snap shot image until the position adjustment is stopped when the first vector included angle is equal to a preset first included angle value and the first linear distance is equal to a preset first distance value;

an inspection image trajectory graph is generated based on the penetration trajectory.

Preferably, step S2: predicting an endoscopic image optimization need of a user, comprising:

acquiring a generated voice fragment and corresponding voice start-stop time when a user enters the body of a patient by using an endoscope;

determining the time interval position corresponding to the starting and ending time of the voice from a preset time axis;

setting the voice segment at the time interval position;

acquiring snapshot time of a snapshot image;

determining a first time position corresponding to the snapshot time from a time axis;

a second time position which accords with the time position condition within a preset time period before the first time position on the time axis;

determining a local human voice segment between a leftmost second time position and a first time position on a time axis;

carrying out semantic extraction on the local voice segments to obtain first voice semantics;

matching the first human voice semantics with second human voice semantics in a preset characterization semantic library;

when the matching is met, taking the matched first human voice semantic as a first target semantic, taking the matched second human voice semantic as a second target semantic, and acquiring a preset representation requirement item and a preset representation type corresponding to the second target semantic; the characterization types include: individual and combined characterizations;

When the characterization type of the second target semantic is single characterization, taking the characterization requirement item as an endoscope image optimization requirement;

when the representation type of the second target semantics is the combination representation, acquiring a preset combination representation semantics library corresponding to the second target semantics;

matching any third human voice semantics in the combined representation semantic library with first human voice semantics except the first target semantics;

when the two images are matched and met, taking the representation requirement item as an endoscope image optimization requirement;

wherein the time position condition includes:

and checking that the track corresponding to the second time position exists in the image track graph.

Preferably, step S3: based on the optimization requirement of the endoscope image, preprocessing the examination image track map comprises the following steps:

determining a preprocessing requirement corresponding to the endoscope image optimization requirement from a preset preprocessing requirement library; the pretreatment requirements include: image requirements, image processing requirements, and image analysis requirements;

determining a snap image meeting the image requirements from the checked image track diagram and taking the snap image as a target image;

only the target image is reserved in the inspection image track diagram;

determining an image processing template corresponding to the image processing requirement from a preset image processing template library;

Performing image processing on the target image based on the image processing template;

determining an image analysis template corresponding to the image analysis requirement from a preset image analysis template library;

based on the image analysis template, performing image analysis on the target image to obtain an image analysis result;

mapping the image analysis result into a preset information frame, associating the information frame with a corresponding target image, and setting the information frame beside the corresponding target image.

Preferably, the medical endoscope image optimization method further comprises:

acquiring the center position of the information frame and the opposite direction of the information frame;

constructing a third direction vector based on the center position of the information frame and the opposite direction of the information frame;

acquiring a viewing position and a viewing direction of a user viewing information frame;

constructing a fourth direction vector based on the viewing position and the viewing direction;

calculating a second vector angle between the third direction vector and the fourth direction vector;

acquiring a second linear distance between the center position of the information frame and the image center position of the target image associated with the information frame;

and performing position adjustment on the information frame until the position adjustment is stopped when the second vector included angle is equal to a preset second included angle value and the second linear distance is equal to a preset second distance value.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.

The technical scheme of the invention is further described in detail through the drawings and the embodiments.

Drawings

The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:

FIG. 1 is a schematic diagram of a medical endoscopic image optimization system in accordance with an embodiment of the present invention;

fig. 2 is a schematic diagram of a medical endoscope image optimization method according to an embodiment of the present invention.

Detailed Description

The preferred embodiments of the present invention will be described below with reference to the accompanying drawings, it being understood that the preferred embodiments described herein are for illustration and explanation of the present invention only, and are not intended to limit the present invention.

An embodiment of the present invention provides a medical endoscope image optimization system, as shown in fig. 1, including:

An inspection image trajectory graph generation module 1 for generating an inspection image trajectory graph of an endoscope when the endoscope is ready to be withdrawn from the body of a patient; wherein the end of the patient's endoscopy is indicated when the endoscope is ready to be withdrawn from the patient's body; when a patient performs the endoscopic examination in the examination image track diagram, the endoscope stretches into a detection track in the body of the patient, the detection track is recorded and generated by the endoscope, and when a doctor controls the endoscope to take a snapshot each time, the snapshot image is automatically marked at the position of a corresponding track point on the detection track according to the snapshot position;

an endoscope image optimization demand prediction module 2 for predicting an endoscope image optimization demand of a user; the user is typically a doctor; the endoscope image optimization requirements may be, for example: the brightness of the snap shot image is increased, the area of the intestinal ulcer in the snap shot image is analyzed, and only the snap shot image with the intestinal ulcer is displayed;

the preprocessing module 3 is used for preprocessing the examination image track map based on the optimization requirement of the endoscope image and outputting and displaying the preprocessing result. Such as: the endoscope image optimization requirement is to analyze the intestinal ulcer area in the snap images, then analyze the intestinal ulcer area of each snap image in the inspection image track map and mark beside the corresponding snap image.

The working principle and the beneficial effects of the technical scheme are as follows:

when the endoscopic examination of the patient is finished, the examination image track map is generated, a doctor can intuitively determine which position in the body of the patient the snapshot image corresponds to by checking the examination image track map, can intuitively determine the position distribution in the body of the patient corresponding to each snapshot image and the disease condition corresponding to each snapshot image, is convenient for global analysis, and in addition, the examination image track map is preprocessed according to the optimization requirement of the doctor's endoscopic image, so that the doctor does not need to process the examination image track map by himself, the analysis efficiency of the doctor for analyzing according to the snapshot images is improved, and meanwhile, the method is more humanized.

When the application is implemented, a doctor only needs to carry out the endoscopy on a patient according to a normal flow, if necessary, the image snapshot is carried out, and after the endoscopy is finished, the analysis is carried out by observing the output preprocessed examination image track diagram and the examination conclusion is given.

In one embodiment, the examination image trajectory graph generation module 1 generates an examination image trajectory graph of an endoscope, performing the following operations:

acquiring a detection track, a snap-shot image and corresponding snap-shot parameters when the endoscope enters the body of a patient; the snapshot parameters include: a snapshot position and a snapshot direction; wherein the detection track is three-dimensional and is placed in a three-dimensional space; the snapshot position is the position of the lens of the endoscope in the patient body when the snapshot image is taken, and the snapshot direction is the alignment direction of the lens when the snapshot is taken;

Determining the track point position corresponding to the snapshot position from the detection track;

constructing a first direction vector based on the track point position and the snapshot direction;

randomly setting the snap-shot images beside the track point positions;

acquiring the center position of an image of the snap shot image and the opposite direction of the image; the right direction of the image is the right direction of the front face of the snap-shot image;

constructing a second direction vector based on the image center position and the image facing direction;

calculating a first vector included angle between the first direction vector and the second direction vector;

acquiring a first linear distance between a track point position and an image center position;

position adjustment is carried out on the snap shot image until the position adjustment is stopped when the first vector included angle is equal to a preset first included angle value and the first linear distance is equal to a preset first distance value; the first included angle value is preset to be 180 degrees; the preset first distance value may be, for example: 1.5 cm;

an inspection image trajectory graph is generated based on the penetration trajectory. When the method is used for generating, the detection track and the three-dimensional space where the detection track is located are taken as a detection image track diagram.

The working principle and the beneficial effects of the technical scheme are as follows:

when the first vector included angle is equal to a preset first included angle value, the relative position relationship between the snap image and the track point position is consistent with the relative position relationship between the lens and the snap region when the endoscope enters the patient body to snap the snap image, and a doctor can visually and intuitively see the relative position relationship between the snap image and the track point position when observing, so that the relative position relationship between the lens and the snap region when the endoscope enters the patient body to snap the snap image is known, the snap region is accurately judged, and the rationality of generating the track map of the inspection image is improved.

When the first linear distance is equal to a preset first distance value, the visually captured image is close enough to the track point position, so that a doctor can know that the captured image is associated with the track point position when viewing, namely the captured image is captured at the position, and the rationality of generating the track map of the inspection image is further improved.

In one embodiment, the endoscopic image optimization demand prediction module 2 predicts the endoscopic image optimization demand of the user, performing the following operations:

acquiring a generated voice fragment and corresponding voice start-stop time when a user enters the body of a patient by using an endoscope;

determining the time interval position corresponding to the starting and ending time of the voice from a preset time axis; the left boundary and the right boundary of the time interval position correspond to the starting and stopping time of the voice;

setting the voice segment at the time interval position;

acquiring snapshot time of a snapshot image;

determining a first time position corresponding to the snapshot time from a time axis;

a second time position which accords with the time position condition within a preset time period before the first time position on the time axis; the preset duration is, for example: 12 seconds;

determining a local human voice segment between a leftmost second time position and a first time position on a time axis; wherein the farther left the time position on the time axis is, the earlier the time is represented;

Carrying out semantic extraction on the local voice segments to obtain first voice semantics;

matching the first human voice semantics with second human voice semantics in a preset characterization semantic library; the second human voice semantics are semantics for representing that a doctor has an endoscope image optimization requirement, for example: if the image is a little dark, the endoscope image optimization requirement of the represented doctor is that the brightness of the snap shot image is adjusted;

when the matching is met, taking the matched first human voice semantic as a first target semantic, taking the matched second human voice semantic as a second target semantic, and acquiring a preset representation requirement item and a preset representation type corresponding to the second target semantic; the characterization types include: individual and combined characterizations; the representation requirement item is an endoscope image optimization requirement of a doctor represented by second voice semantics, for example: the second voice semantic is that the image is a little dark, and the characterization requirement item is brightness of the bright snapshot image;

when the characterization type of the second target semantic is single characterization, taking the characterization requirement item as an endoscope image optimization requirement; the second target semantic representation type is divided into an independent representation and a combined representation, when the representation type is the independent representation, the second target semantic representation can be used for independently representing the endoscope image optimization requirement of a doctor, for example, the second human semantic representation is 'image is a little dark', and the endoscope image optimization requirement of the doctor is brightness of a bright snapshot image;

When the representation type of the second target semantics is the combination representation, acquiring a preset combination representation semantics library corresponding to the second target semantics; when the token type is a combination token, the second target semantic requirement and a third human voice semantic combination in the combination token semantic library are described to represent the endoscope image optimization requirement of a doctor, for example: the second voice meaning is 'ulcer is present here', the third voice meaning is 'ulcer is present there', and then the combination characterizes that the optimization requirement of the endoscope image which is present by doctors is that only the snap-shot image with the intestinal ulcer is displayed independently, and the intestinal ulcer area in the snap-shot image with the intestinal ulcer is calculated;

matching any third human voice semantics in the combined representation semantic library with first human voice semantics except the first target semantics;

when the two images are matched and met, taking the representation requirement item as an endoscope image optimization requirement; when the third human voice semantics have the first human voice semantics matched with the third human voice semantics, the second human voice semantics are successfully represented by combination, and the representation requirement item is an endoscope image optimization requirement;

wherein the time position condition includes:

and checking that the track corresponding to the second time position exists in the image track graph. When the detection track in the examination image track graph is recorded and generated by the endoscope, when the endoscope stops moving in the patient body, one track stopping point is generated, and the starting and stopping time of stopping the movement of the endoscope is given to the track stopping point, so that if the corresponding track stopping point exists in the examination image track graph at the second time position, the description time is positioned at the second time position, and the endoscope is in a static state in the patient body.

The working principle and the beneficial effects of the technical scheme are as follows:

generally, in a practical application scenario, when a doctor uses an endoscope to perform endoscopy on a patient, before each necessary snapshot is performed, an image returned by the lens of the endoscope is described orally for recording by a doctor or a nurse responsible for recording at the side, for example: the intestinal ulcer area is seen in the image returned by the endoscope lens, the intestinal ulcer area needs to be snapped, before the snapping, the "ulcer is not small, how big to see", and the following is say: when the snapshot is prepared, the brightness of the image returned by the endoscope lens is found to be darker, and the snapshot is dictated as to how dark the brightness is. Thus, this feature can be used to predict the physician's need for endoscopic image optimization.

According to the embodiment of the invention, the local voice segments are screened out, the first voice semantics are extracted, the endoscope image optimization requirement of a doctor is predicted based on the first voice semantics, and the applicability of the system is improved. During prediction, a characterization semantic library is introduced, and the endoscope image optimization requirement of a doctor is accurately and rapidly predicted.

In addition, when a doctor endoscopes a patient using an endoscope, the dictated content is not necessarily valuable for predicting the need for optimization of the endoscopic image, such as: the doctor dictates what medical history the patient has, namely inquires the medical history conditions of other doctor patients, so that the embodiment of the invention introduces the first time position and the leftmost second time position conforming to the time position condition, reasonably determines the local voice segments of the image returned by the description endoscope lens before the doctor captures the image, predicts the optimization requirement of the endoscope image, and improves the accuracy and the prediction efficiency of the prediction.

In one embodiment, the preprocessing module 3 preprocesses the examination image trajectory graph based on the endoscope image optimization requirements, performing the following operations:

determining a preprocessing requirement corresponding to the endoscope image optimization requirement from a preset preprocessing requirement library; the pretreatment requirements include: image requirements, image processing requirements, and image analysis requirements; wherein, the pretreatment requirements corresponding to different endoscope image optimization requirements are in the pretreatment requirements library; the image requirement is a requirement of a doctor on a snap shot image, such as: images containing intestinal ulcers, etc.; the image processing requirements are the processing requirements of doctors on snap shots, such as: increasing the brightness of the image, etc.; the image analysis requirements are the analysis requirements of doctors on snap shots, such as: analyzing the intestinal ulcer area in the snap shot image;

determining a snap image meeting the image requirements from the checked image track diagram and taking the snap image as a target image;

only the target image is reserved in the inspection image track diagram;

determining an image processing template corresponding to the image processing requirement from a preset image processing template library; the image processing template library is provided with image processing templates corresponding to different image processing requirements, and the image processing templates can be used for processing the snap-shot images according to the image processing requirements by the system, for example: if the image processing requirement is to increase the image brightness, the image parameters which are set for the snapshot image when the image brightness is increased are arranged in the image processing template, and the system can realize the brightness increase of the snapshot image by contrast with the image processing template;

Performing image processing on the target image based on the image processing template;

determining an image analysis template corresponding to the image analysis requirement from a preset image analysis template library; the image analysis template library is provided with image analysis templates corresponding to different image analysis requirements, and the image analysis templates can be used for the system to compare and execute analysis meeting the image analysis requirements on the snap-shot images, for example: the image analysis requirement is to analyze the intestinal ulcer area in the snap shot image, and then a strategy for identifying and extracting the intestinal ulcer area image from the snap shot image is arranged in the image analysis template (the strategy can be set by a technician based on an image identification technology, and can be realized), and the system can identify and extract the intestinal ulcer area image from the snap shot image and calculate the image area thereof as the intestinal ulcer area by comparing with the image analysis template;

based on the image analysis template, performing image analysis on the target image to obtain an image analysis result;

mapping the image analysis result into a preset information frame, associating the information frame with a corresponding target image, and setting the information frame beside the corresponding target image. After the image analysis result is mapped into the information frame, the information frame can display the image analysis result.

The working principle and the beneficial effects of the technical scheme are as follows:

according to the embodiment of the invention, the preprocessing requirement library, the image processing template library and the image analysis template library are introduced, so that the inspection image track map is preprocessed rapidly based on the endoscope image optimization requirement, and the preprocessing efficiency is improved.

In one embodiment, the medical endoscopic image optimization system further comprises:

a position adjustment module for including:

acquiring the center position of the information frame and the opposite direction of the information frame; the right direction of the information frame is the right direction of the front surface of the information frame;

constructing a third direction vector based on the center position of the information frame and the opposite direction of the information frame;

acquiring a viewing position and a viewing direction of a user viewing information frame; in general, a doctor views the examination image track diagram through a terminal (such as a computer) and pulls the examination image track diagram so that the doctor can see the content which the doctor wants to view when viewing the examination image track diagram from the central position of the terminal screen, therefore, the viewing position can be the central position of the terminal screen, and the viewing direction is the direction vertically inwards from the central position of the terminal screen;

constructing a fourth direction vector based on the viewing position and the viewing direction;

Calculating a second vector angle between the third direction vector and the fourth direction vector;

acquiring a second linear distance between the center position of the information frame and the image center position of the target image associated with the information frame;

and performing position adjustment on the information frame until the position adjustment is stopped when the second vector included angle is equal to a preset second included angle value and the second linear distance is equal to a preset second distance value. Wherein the preset second included angle value is 180 degrees, and the preset second distance value is 0.8 cm.

The working principle and the beneficial effects of the technical scheme are as follows:

when the second vector included angle is equal to a preset second included angle value, the information frame can be opposite to the face of the doctor, and the doctor can look at the content of the information frame. Generally, when a doctor views the inspection image track map, the doctor drags the inspection image track map to view the snap-shot images from different angles, but the analysis result of each snap-shot image, namely the associated information frame doctor, is required to view at any time (most of the analysis is based on the content displayed by the information frame to perform global analysis), so that the second vector included angle is ensured to be equal to the preset second included angle value, the practical application requirement is met, the rationality is improved, and the doctor is more humanized. In addition, the second linear distance is ensured to be equal to a preset second distance value, so that the information frame is close to the corresponding snap-shot image, and a doctor can visually know that the information frame is an analysis annotation on the snap-shot image.

The embodiment of the invention provides a medical endoscope image optimization method, which is shown in fig. 2 and comprises the following steps:

step S1: generating an inspection image trajectory graph of the endoscope when the endoscope is ready to be withdrawn from the body of the patient;

step S2: predicting an endoscopic image optimization requirement of a user;

step S3: based on the optimization requirement of the endoscope image, preprocessing the examination image track diagram, and outputting and displaying the preprocessing result.

In step S1, an inspection image trajectory chart of an endoscope is generated, including:

acquiring a detection track, a snap-shot image and corresponding snap-shot parameters when the endoscope enters the body of a patient; the snapshot parameters include: a snapshot position and a snapshot direction;

determining the track point position corresponding to the snapshot position from the detection track;

constructing a first direction vector based on the track point position and the snapshot direction;

randomly setting the snap-shot images beside the track point positions;

acquiring the center position of an image of the snap shot image and the opposite direction of the image;

constructing a second direction vector based on the image center position and the image facing direction;

calculating a first vector included angle between the first direction vector and the second direction vector;

acquiring a first linear distance between a track point position and an image center position;

Position adjustment is carried out on the snap shot image until the position adjustment is stopped when the first vector included angle is equal to a preset first included angle value and the first linear distance is equal to a preset first distance value;

an inspection image trajectory graph is generated based on the penetration trajectory.

Step S2: predicting an endoscopic image optimization need of a user, comprising:

acquiring a generated voice fragment and corresponding voice start-stop time when a user enters the body of a patient by using an endoscope;

determining the time interval position corresponding to the starting and ending time of the voice from a preset time axis;

setting the voice segment at the time interval position;

acquiring snapshot time of a snapshot image;

determining a first time position corresponding to the snapshot time from a time axis;

a second time position which accords with the time position condition within a preset time period before the first time position on the time axis;

determining a local human voice segment between a leftmost second time position and a first time position on a time axis;

carrying out semantic extraction on the local voice segments to obtain first voice semantics;

matching the first human voice semantics with second human voice semantics in a preset characterization semantic library;

when the matching is met, taking the matched first human voice semantic as a first target semantic, taking the matched second human voice semantic as a second target semantic, and acquiring a preset representation requirement item and a preset representation type corresponding to the second target semantic; the characterization types include: individual and combined characterizations;

When the characterization type of the second target semantic is single characterization, taking the characterization requirement item as an endoscope image optimization requirement;

when the representation type of the second target semantics is the combination representation, acquiring a preset combination representation semantics library corresponding to the second target semantics;

matching any third human voice semantics in the combined representation semantic library with first human voice semantics except the first target semantics;

when the two images are matched and met, taking the representation requirement item as an endoscope image optimization requirement;

wherein the time position condition includes:

and checking that the track corresponding to the second time position exists in the image track graph.

Step S3: based on the optimization requirement of the endoscope image, preprocessing the examination image track map comprises the following steps:

determining a preprocessing requirement corresponding to the endoscope image optimization requirement from a preset preprocessing requirement library; the pretreatment requirements include: image requirements, image processing requirements, and image analysis requirements;

determining a snap image meeting the image requirements from the checked image track diagram and taking the snap image as a target image;

only the target image is reserved in the inspection image track diagram;

determining an image processing template corresponding to the image processing requirement from a preset image processing template library;

Performing image processing on the target image based on the image processing template;

determining an image analysis template corresponding to the image analysis requirement from a preset image analysis template library;

based on the image analysis template, performing image analysis on the target image to obtain an image analysis result;

mapping the image analysis result into a preset information frame, associating the information frame with a corresponding target image, and setting the information frame beside the corresponding target image.

The medical endoscope image optimization method further comprises the following steps:

acquiring the center position of the information frame and the opposite direction of the information frame;

constructing a third direction vector based on the center position of the information frame and the opposite direction of the information frame;

acquiring a viewing position and a viewing direction of a user viewing information frame;

constructing a fourth direction vector based on the viewing position and the viewing direction;

calculating a second vector angle between the third direction vector and the fourth direction vector;

acquiring a second linear distance between the center position of the information frame and the image center position of the target image associated with the information frame;

and performing position adjustment on the information frame until the position adjustment is stopped when the second vector included angle is equal to a preset second included angle value and the second linear distance is equal to a preset second distance value.

It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims (6)

1. A medical endoscopic image optimization system, comprising:

an inspection image track map generation module for generating an inspection image track map of an endoscope when the endoscope is ready to be withdrawn from the body of a patient;

the endoscope image optimization demand prediction module is used for predicting the endoscope image optimization demand of a user;

the preprocessing module is used for preprocessing the inspection image track graph based on the endoscope image optimization requirement and outputting and displaying a preprocessing result;

the examination image trajectory graph generating module generates an examination image trajectory graph of the endoscope, and performs the following operations:

acquiring a detection track, a snap-shot image and corresponding snap-shot parameters when the endoscope enters the body of the patient; the snapshot parameters include: a snapshot position and a snapshot direction;

determining the track point position corresponding to the snapshot position from the detection track;

Constructing a first direction vector based on the track point position and the snapshot direction;

randomly setting the snap-shot images beside the track point positions;

acquiring the image center position and the image facing direction of the snap shot image;

constructing a second direction vector based on the image center position and the image facing direction;

calculating a first vector angle between the first direction vector and the second direction vector;

acquiring a first linear distance between the track point position and the image center position;

position adjustment is carried out on the snap shot image until position adjustment is stopped when the first vector included angle is equal to a preset first included angle value and the first linear distance is equal to a preset first distance value;

generating the inspection image track map based on the penetration track;

the endoscope image optimization demand prediction module predicts the endoscope image optimization demand of a user and executes the following operations:

acquiring a generated human voice fragment and corresponding human voice start-stop time when the user enters the body of the patient by using the endoscope;

determining the time interval position corresponding to the starting and ending time of the voice from a preset time axis;

Setting the voice section at the time interval position;

acquiring the snapshot time of the snapshot image;

determining a first time position corresponding to the snapshot time from the time axis;

a second time position which accords with time position conditions in a preset time period before the first time position on the time axis;

determining a local human voice segment between the leftmost second time position and the first time position from the time axis;

carrying out semantic extraction on the local voice segments to obtain first voice semantics;

matching the first human voice semantics with second human voice semantics in a preset characterization semantic library;

when the first human voice semantics are matched and matched as first target semantics, the second human voice semantics are matched and matched as second target semantics, and a preset characterization requirement item and a preset characterization type corresponding to the second target semantics are obtained; the characterization type includes: individual and combined characterizations;

when the representation type of the second target semantic is the single representation, taking the representation requirement item as the endoscope image optimization requirement;

When the representation type of the second target semantic is the combined representation, acquiring a preset combined representation semantic library corresponding to the second target semantic;

matching any third human voice semantics in the combined representation semantic library with the first human voice semantics except the first target semantics;

when the two images are matched and matched, the representation requirement item is used as the endoscope image optimization requirement;

wherein the time position condition includes:

and the track stopping point corresponding to the second time position is arranged in the checking image track graph.

2. The medical endoscopic image optimization system of claim 1, wherein said preprocessing module preprocesses said examination image trajectory graph based on said endoscopic image optimization requirement, performing the following operations:

determining a preprocessing requirement corresponding to the endoscope image optimization requirement from a preset preprocessing requirement library; the preprocessing requirements include: image requirements, image processing requirements, and image analysis requirements;

determining the snap-shot image meeting the image requirement from the checking image track graph and taking the snap-shot image as a target image;

Only the target image is retained in the inspection image trajectory graph;

determining an image processing template corresponding to the image processing requirement from a preset image processing template library;

performing image processing on the target image based on the image processing template;

determining an image analysis template corresponding to the image analysis requirement from a preset image analysis template library;

based on the image analysis template, performing image analysis on the target image to obtain an image analysis result;

mapping the image analysis result into a preset information frame, associating the information frame with the corresponding target image, and setting the information frame beside the corresponding target image.

3. A medical endoscopic image optimization system according to claim 2, further comprising:

a position adjustment module for including:

acquiring the center position of the information frame and the opposite direction of the information frame;

constructing a third direction vector based on the center position of the information frame and the opposite direction of the information frame;

acquiring a viewing position and a viewing direction of the information frame viewed by the user;

constructing a fourth direction vector based on the viewing position and the viewing direction;

Calculating a second vector angle between the third direction vector and the fourth direction vector;

acquiring a second linear distance between the information frame center position and the image center position of the target image associated with the information frame;

and carrying out position adjustment on the information frame until the position adjustment is stopped when the second vector included angle is equal to a preset second included angle value and the second linear distance is equal to a preset second distance value.

4. A method for optimizing a medical endoscopic image, comprising:

step S1: generating an inspection image trajectory of the endoscope when the endoscope is ready to be withdrawn from the body of the patient;

step S2: predicting an endoscopic image optimization requirement of a user;

step S3: preprocessing the inspection image track graph based on the endoscope image optimization requirement, and outputting and displaying a preprocessing result;

in the step S1, generating an inspection image trajectory chart of the endoscope includes:

acquiring a detection track, a snap-shot image and corresponding snap-shot parameters when the endoscope enters the body of the patient; the snapshot parameters include: a snapshot position and a snapshot direction;

Determining the track point position corresponding to the snapshot position from the detection track;

constructing a first direction vector based on the track point position and the snapshot direction;

randomly setting the snap-shot images beside the track point positions;

acquiring the image center position and the image facing direction of the snap shot image;

constructing a second direction vector based on the image center position and the image facing direction;

calculating a first vector angle between the first direction vector and the second direction vector;

acquiring a first linear distance between the track point position and the image center position;

position adjustment is carried out on the snap shot image until position adjustment is stopped when the first vector included angle is equal to a preset first included angle value and the first linear distance is equal to a preset first distance value;

generating the inspection image track map based on the penetration track;

the step S2: predicting an endoscopic image optimization need of a user, comprising:

acquiring a generated human voice fragment and corresponding human voice start-stop time when the user enters the body of the patient by using the endoscope;

determining the time interval position corresponding to the starting and ending time of the voice from a preset time axis;

Setting the voice section at the time interval position;

acquiring the snapshot time of the snapshot image;

determining a first time position corresponding to the snapshot time from the time axis;

a second time position which accords with time position conditions in a preset time period before the first time position on the time axis;

determining a local human voice segment between the leftmost second time position and the first time position from the time axis;

carrying out semantic extraction on the local voice segments to obtain first voice semantics;

matching the first human voice semantics with second human voice semantics in a preset characterization semantic library;

when the first human voice semantics are matched and matched as first target semantics, the second human voice semantics are matched and matched as second target semantics, and a preset characterization requirement item and a preset characterization type corresponding to the second target semantics are obtained; the characterization type includes: individual and combined characterizations;

when the representation type of the second target semantic is the single representation, taking the representation requirement item as the endoscope image optimization requirement;

When the representation type of the second target semantic is the combined representation, acquiring a preset combined representation semantic library corresponding to the second target semantic;

matching any third human voice semantics in the combined representation semantic library with the first human voice semantics except the first target semantics;

when the two images are matched and matched, the representation requirement item is used as the endoscope image optimization requirement;

wherein the time position condition includes:

and the track stopping point corresponding to the second time position is arranged in the checking image track graph.

5. The method for optimizing a medical endoscopic image as defined in claim 4, wherein said step S3: preprocessing the inspection image trajectory graph based on the endoscope image optimization requirement, including:

determining a preprocessing requirement corresponding to the endoscope image optimization requirement from a preset preprocessing requirement library; the preprocessing requirements include: image requirements, image processing requirements, and image analysis requirements;

determining the snap-shot image meeting the image requirement from the checking image track graph and taking the snap-shot image as a target image;

only the target image is retained in the inspection image trajectory graph;

Determining an image processing template corresponding to the image processing requirement from a preset image processing template library;

performing image processing on the target image based on the image processing template;

determining an image analysis template corresponding to the image analysis requirement from a preset image analysis template library;

based on the image analysis template, performing image analysis on the target image to obtain an image analysis result;

mapping the image analysis result into a preset information frame, associating the information frame with the corresponding target image, and setting the information frame beside the corresponding target image.

6. The medical endoscopic image optimization method according to claim 5, further comprising:

acquiring the center position of the information frame and the opposite direction of the information frame;

constructing a third direction vector based on the center position of the information frame and the opposite direction of the information frame;

acquiring a viewing position and a viewing direction of the information frame viewed by the user;

constructing a fourth direction vector based on the viewing position and the viewing direction;

calculating a second vector angle between the third direction vector and the fourth direction vector;

Acquiring a second linear distance between the information frame center position and the image center position of the target image associated with the information frame;

and carrying out position adjustment on the information frame until the position adjustment is stopped when the second vector included angle is equal to a preset second included angle value and the second linear distance is equal to a preset second distance value.