Operation index determination method, device and equipment for endoscopic surgery and storage medium

Abstract

The invention discloses an operation index determination method and device for an endoscopic surgery, equipment and a storage medium, and can be applied to the technical field of surgical operation data processing. The method comprises the following steps: firstly, acquiring a surgical operation data set of a target operation object for an endoscopic surgery; the surgical operation data set comprises a plurality of pieces of surgical operation data sorted according to a time sequence. Each piece of surgical operation data comprises five-dimensional data, namely endoscope action data, endoscope type data, instrument action data, instrument type data and surgical object data. And constructing a hidden Markov model by using the surgical operation data set. In the hidden Markov model, the surgical operation described by the surgical operation data is the observed quantity, and the surgical state is the state quantity, and based on the hidden Markov model, the statistical principle is utilized to carry out quantitative evaluation on the surgical operation indexes, so that the accurate evaluation of the surgical operation when a doctor carries out the endoscopic surgery is realized, and the reference effectiveness of the operation indexes is improved.

Description

Technical Field

[0001] This application relates to the field of surgical operation data processing technology, specifically to a method, apparatus, equipment, and storage medium for determining operational indicators of endoscopic surgery. Background Technology

[0002] Endoscopic surgery, also known as endoscopic minimally invasive surgery, is a surgical procedure that uses an endoscope and related instruments to enter the body through natural cavities or tiny incisions, performing the procedure under direct visualization. Endoscopic surgery is widely used in various fields, including digestive, respiratory, urological, gynecological, and surgical procedures.

[0003] Currently, during endoscopic surgery, the treatment methods and results are recorded. This allows for subsequent evaluation of the surgeon's skill level based on the recorded results. However, evaluating a surgeon's skill level based solely on treatment results is not accurate enough. Summary of the Invention

[0004] In view of this, this application provides a method, apparatus, equipment and storage medium for determining the operational indicators of endoscopic surgery, which can more accurately assess the operational level of endoscopic surgery.

[0005] To solve the above problems, the technical solution provided in this application is as follows:

[0006] In a first aspect, this application provides a method for determining operational indicators of endoscopic surgery, the method comprising:

[0007] Obtain a set of surgical operation data for the target operation object in endoscopic surgery. The set of surgical operation data includes multiple surgical operation data arranged in chronological order. Each surgical operation data is used to describe the surgical operation of the target operation object in the endoscopic surgery. Each surgical operation data includes data in five dimensions, namely endoscopic action data, endoscopic type data, instrument action data, instrument type data, and surgical object data.

[0008] A hidden Markov model is constructed using the surgical operation dataset. The hidden Markov model includes a state transition probability matrix, an observation probability matrix, and an initial state probability vector. The state transition probability matrix is ​​used to describe the probability of surgical state transitions during the endoscopic surgery. The observation probability matrix is ​​used to describe the probability of the target object performing each surgical operation in each surgical state. The initial state probability vector is used to describe the probability of being in each surgical state at the initial moment.

[0009] Based on the Hidden Markov Model, the operational indicators of the target operation object are determined, and the operational indicators are used to measure the operational level of the target operation object in the endoscopic surgery.

[0010] In one possible implementation, determining the operational indicators of the target operation object based on the hidden Markov model includes:

[0011] Determine the state transition volatility based on the hidden Markov model;

[0012] The state transition volatility is used as a proficiency indicator, which measures the proficiency of the target operator in the endoscopic surgery.

[0013] And / or,

[0014] Based on the hidden Markov model, determine the observation discrete entropy;

[0015] The observed discrete entropy is used as an effectiveness index, which is used to measure the effectiveness of the surgical operation performed on the target object for the endoscopic surgery.

[0016] In one possible implementation, the surgical operation data set includes multiple subsets, each subset corresponding to an endoscopic surgery, each subset including multiple surgical operation data items ordered chronologically, and each surgical operation data item corresponding to a surgical operation of an operation frame, wherein the operation frame is obtained by dividing the endoscopic surgery according to time.

[0017] In one possible implementation, the operation frame is divided according to unit operation time;

[0018] or,

[0019] The operation frames are divided according to the changing moments of the surgical operation.

[0020] In one possible implementation, the surgical data is generated in the following manner:

[0021] Obtain surgical record text data;

[0022] The information of the five dimensions is extracted from the surgical record text data;

[0023] Based on the information from the five dimensions and the data specifications, the surgical operation data is generated. The data specifications are used to indicate the format requirements and encoding methods that the generated surgical operation data must meet.

[0024] In one possible implementation, the endoscopic action data includes first action type data and first operation data, wherein the first action type data is used to represent the type of endoscopic action and the first operation data is used to represent the content of the endoscopic action.

[0025] The data on endoscope types are represented using abbreviations for endoscope types;

[0026] The device motion data includes second motion type data and second operation data. The second motion type data is used to represent the motion type of the device, and the second operation data is used to represent the content of the device motion.

[0027] The device type data is represented by abbreviations for device types;

[0028] The surgical object data includes anatomical location data, lesion type data, and lesion content data. The anatomical location data is used to indicate the anatomical location, the lesion type data is used to indicate the type of lesion, and the lesion content data is used to describe the size of the lesion.

[0029] In one possible implementation, the method further includes:

[0030] Determine the relationship between the operational indicators and the indicator thresholds;

[0031] Based on the size relationship, a prompt message is generated, which is used to indicate whether the operation indicator meets the requirements.

[0032] Secondly, this application provides a device for determining operational parameters of endoscopic surgery, the device comprising:

[0033] The acquisition unit is used to acquire a set of surgical operation data for the target operation object in endoscopic surgery. The set of surgical operation data includes multiple surgical operation data arranged in chronological order. Each surgical operation data is used to describe the surgical operation of the target operation object in the endoscopic surgery. Each surgical operation data includes data in five dimensions, namely endoscopic action data, endoscopic type data, instrument action data, instrument type data, and surgical object data.

[0034] The construction unit is used to construct a hidden Markov model using the surgical operation data set. The hidden Markov model includes a state transition probability matrix, an observation probability matrix, and an initial state probability vector. The state transition probability matrix is ​​used to describe the probability of surgical state transitions during the endoscopic surgery. The observation probability matrix is ​​used to describe the probability of the target object performing each surgical operation in each surgical state. The initial state probability vector is used to describe the probability of being in each surgical state at the initial moment.

[0035] The determining unit is used to determine the operational indicators of the target operation object based on the hidden Markov model, wherein the operational indicators are used to measure the operational level of the target operation object in the endoscopic surgery.

[0036] In one possible implementation, the determining unit is specifically used for:

[0037] Determine the state transition volatility based on the hidden Markov model;

[0038] The state transition volatility is used as a proficiency indicator, which measures the proficiency of the target operator in the endoscopic surgery.

[0039] And / or,

[0040] Based on the hidden Markov model, determine the observation discrete entropy;

[0041] The observed discrete entropy is used as an effectiveness index, which is used to measure the effectiveness of the surgical operation performed on the target object for the endoscopic surgery.

[0042] In one possible implementation, the surgical operation data set includes multiple subsets, each subset corresponding to an endoscopic surgery, each subset including multiple surgical operation data items ordered chronologically, and each surgical operation data item corresponding to a surgical operation of an operation frame, wherein the operation frame is obtained by dividing the endoscopic surgery according to time.

[0043] In one possible implementation, the operation frame is divided according to unit operation time;

[0044] or,

[0045] The operation frames are divided according to the changing moments of the surgical operation.

[0046] In one possible implementation, the surgical data is generated in the following manner:

[0047] Obtain surgical record text data;

[0048] The information of the five dimensions is extracted from the surgical record text data;

[0049] Based on the information from the five dimensions and the data specifications, the surgical operation data is generated. The data specifications are used to indicate the format requirements and encoding methods that the generated surgical operation data must meet.

[0050] In one possible implementation, the endoscopic action data includes first action type data and first operation data, wherein the first action type data is used to represent the type of endoscopic action and the first operation data is used to represent the content of the endoscopic action.

[0051] The data on endoscope types are represented using abbreviations for endoscope types;

[0052] The device motion data includes second motion type data and second operation data. The second motion type data is used to represent the motion type of the device, and the second operation data is used to represent the content of the device motion.

[0053] The device type data is represented by abbreviations for device types;

[0054] The surgical object data includes anatomical location data, lesion type data, and lesion content data. The anatomical location data is used to indicate the anatomical location, the lesion type data is used to indicate the type of lesion, and the lesion content data is used to describe the size of the lesion.

[0055] In one possible implementation, the device further includes:

[0056] The prompting unit is used to determine the relationship between the operation indicator and the indicator threshold; and to generate prompting information based on the relationship, the prompting information being used to indicate whether the operation indicator meets the requirements.

[0057] Thirdly, this application provides a device for determining operational parameters of endoscopic surgery, including: a processor, a memory, and a system bus;

[0058] The processor and the memory are connected via the system bus;

[0059] The memory is used to store one or more programs, the one or more programs including instructions that, when executed by the processor, cause the processor to perform the method described in any of the embodiments of the first aspect above.

[0060] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a terminal device, cause the terminal device to perform the method described in any of the embodiments of the first aspect.

[0061] Therefore, this application has the following beneficial effects:

[0062] This application provides a method, apparatus, device, and storage medium for determining operational indicators in endoscopic surgery. The method first acquires a set of surgical operation data for a target patient undergoing endoscopic surgery. This set includes multiple surgical operation data entries ordered chronologically to describe the surgical operations performed on the target patient during endoscopic surgery. Each surgical operation data entry includes five dimensions: endoscopic action data, endoscope type data, instrument action data, instrument type data, and patient data. A Hidden Markov Model (HMM) is then constructed using the surgical operation data set. In the HMM, the surgical operations described by the data are the observed quantities, and the surgical state is the state quantity. Based on the HMM, statistical principles are used to quantitatively evaluate the operational indicators of the surgery, achieving accurate assessment of the surgeon's operations during endoscopic surgery and improving the reference validity of the operational indicators. Attached Figure Description

[0063] Figure 1 This is a schematic diagram illustrating an application scenario of a method for determining operational indicators in endoscopic surgery, as provided in an embodiment of this application.

[0064] Figure 2 A flowchart illustrating a method for determining operational parameters in endoscopic surgery, provided as an embodiment of this application;

[0065] Figure 3 A schematic diagram illustrating the relationship between five dimensions provided in an embodiment of this application;

[0066] Figure 4 A schematic diagram illustrating surgical operation data included in a subset provided for an embodiment of this application;

[0067] Figure 5 A schematic diagram of surgical operation data provided in an embodiment of this application;

[0068] Figure 6 This is a schematic diagram of the endoscopic surgery operation index determination device provided in the embodiments of this application. Detailed Implementation

[0069] To facilitate understanding and explanation of the technical solutions provided in the embodiments of this application, the background technology of this application will be described first.

[0070] The digestive and respiratory tracts are common sites for human diseases. Lesions within these natural cavities can be examined or treated surgically using flexible endoscopy.

[0071] During endoscopic surgery, the surgeon's intraoperative procedures can be manually recorded. The records include the treatment method and its effects. However, current recording methods are rather haphazard, making it difficult to standardize the processing of the recorded data and accurately assess the surgeon's skill level in performing endoscopic surgery based solely on the recorded information.

[0072] Based on this, embodiments of this application provide a method, apparatus, device, and storage medium for determining operational indicators of endoscopic surgery.

[0073] This method first acquires a set of surgical operation data for the target subject undergoing endoscopic surgery. This set includes multiple surgical operation data entries ordered chronologically to describe the surgical operations performed on the target subject during endoscopic surgery. Each surgical operation data entry includes five dimensions: endoscopic action data, endoscope type data, instrument action data, instrument type data, and surgical subject data. A Hidden Markov Model (HMM) is then constructed using this set of surgical operation data. In the HMM, the surgical operations described by the data are the observed variables, and the surgical state is the state variable. Based on the HMM, statistical principles are used to quantify and evaluate the surgical operation indicators, generating the target subject's operational indicators.

[0074] This allows for standardized descriptions and time-series recording of surgical procedures during endoscopic surgery. By dividing surgical procedures into five dimensions and recording the state information of these five dimensions at different times during the procedure, a standardized description of the endoscopic surgery is formed. This creates a recording medium that can be stored and processed by a computer, forming the foundation for the digitization and intelligentization of endoscopic surgery. Based on standardized coding, a Hidden Markov Model (HMM) statistical model for endoscopic surgery is established to achieve a quantitative assessment of the proficiency of the target patient.

[0075] To facilitate understanding of the technical solutions provided in the embodiments of this application, the method for determining the operational indicators of endoscopic surgery provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0076] See Figure 1As shown in the figure, this is a schematic diagram of an application scenario for a method for determining operational indicators of endoscopic surgery provided in an embodiment of this application. The method for determining operational indicators of endoscopic surgery provided in this embodiment can be applied to a computing device 100 with data processing capabilities. The computing device 100 acquires a set of surgical operation data for endoscopic surgery performed by physician A. The set of surgical operation data includes multiple surgical operation data entries ordered chronologically. Each surgical operation data entry describes the surgical operation performed by physician A during endoscopic surgery. Multiple surgical operation data entries ordered chronologically can describe the endoscopic surgical operation process performed by physician A in chronological order. Each surgical operation data entry uses a unified format, including endoscopic action data, endoscopic type data, instrument action data, instrument type data, and surgical object data, totaling five dimensions of data. Using five-dimensional fine-grained surgical operation data allows for a detailed description of physician A's surgical operations during endoscopic surgery and enables standardized encoding of the surgical operation data, facilitating subsequent data processing using the computing device 100.

[0077] The computing device 100 utilizes the acquired surgical operation data set to construct an HMM (Hidden Markov Model). An HMM addresses time-series problems that are difficult to represent in terms of real-world states. By defining a set of states and a set of observations, and establishing a state transition matrix and an observation probability matrix, it achieves statistical modeling of time-series events. The constructed HMM includes a state transition probability matrix, an observation probability matrix, and an initial state probability vector. The state transition probability matrix describes the probability of surgical state transitions during endoscopic surgery. The observation probability matrix describes the probability of the target surgical object performing each surgical operation in each surgical state. The initial state probability vector describes the probability of being in each surgical state at the initial moment. This allows for statistical modeling of endoscopic surgery using the HMM. Furthermore, based on the HMM, the operational indicators of doctor A are determined. This enables accurate quantitative evaluation of doctor A's operational indicators.

[0078] The above scenario diagrams are merely examples and do not limit the application scenarios of this application. As an example, in some possible implementations, the computing device 100 can also acquire text describing surgical operations in natural language and convert the text into five-dimensional fine-grained surgical operation data.

[0079] See Figure 2 As shown, this figure is a flowchart illustrating a method for determining operational indicators in endoscopic surgery according to an embodiment of this application. Figure 2 As shown in the embodiment of this application, a method for determining the operational indicators of endoscopic surgery includes steps S201-S203.

[0080] S201: Obtain the set of surgical operation data for the target operation object for endoscopic surgery.

[0081] The target of the operation refers to the object on which the endoscopic surgery is performed. The target of the operation may be, for example, the doctor performing the endoscopic surgery, other medical staff involved in the endoscopic surgery, or the surgical robot performing the endoscopic surgery; this application does not limit this.

[0082] For the same target procedure, a set of surgical operation data for endoscopic surgery is obtained. This set includes multiple surgical operation records ordered chronologically. These chronologically ordered records describe the process of the endoscopic surgery, enabling a detailed record of the procedure.

[0083] It should be noted that the surgical procedure data set may include surgical procedure data for one endoscopic surgery of the target patient, or it may include surgical procedure data for multiple endoscopic surgeries of the target patient.

[0084] As an example, the surgical procedure dataset comprises multiple subsets. Each subset corresponds to one endoscopic surgery. Each subset contains multiple surgical procedure records ordered chronologically.

[0085] Each surgical operation data point corresponds to a surgical operation within an operation frame. An operation frame is derived from the time division of an endoscopic surgery corresponding to a subset. For example, see... Figure 4 As shown, this figure is a schematic diagram of surgical operation data included in a subset provided in an embodiment of this application. Wherein, ST l This represents the surgical operation data corresponding to the l-th operation frame. S is the total number of operation frames obtained by dividing a single endoscopic surgery corresponding to a subset according to the operation time.

[0086] Furthermore, this application does not limit the method of dividing operation frames. In one possible implementation, operation frames are divided according to unit operation time. That is, each operation frame includes the same amount of time, which is a unit operation time. The unit operation time can be determined based on the average duration of each operation in the surgical procedure. The surgical operation data of each operation frame may correspond to part, one, or multiple surgical operations.

[0087] In another possible implementation, operation frames are divided according to the timing of changes in surgical operations. That is, operation frames are created when the surgical operation on the target object changes. A change in the surgical operation can refer to a change in the type of action or a change in the degree of action. Each operation frame's surgical operation data can correspond to one or more surgical operations. These multiple surgical operations corresponding to the surgical operation data of each operation frame can be performed simultaneously. In this division method, the time span of each operation frame is not fixed; different operation frames may contain different amounts of time.

[0088] The following is a description of the contents included in the surgical procedure data.

[0089] Surgical procedure data describes the surgical procedures performed on the target patient for endoscopic surgery. Each surgical procedure data entry includes five dimensions: endoscopic action data, endoscope type data, instrument action data, instrument type data, and surgical subject data. Endoscopic action data describes the endoscopic actions performed. Endoscope type data describes the types of endoscopes used. Instrument action data describes the actions of the instruments used in the endoscopic procedure. Instrument type data describes the types of instruments used in the endoscopic procedure. Surgical subject data describes the surgical subject, i.e., the object to be treated during the surgery.

[0090] As an example, a surgical procedure data point can be represented as:

[0091] (1)

[0092] in, This is endoscopic motion data, describing the control behavior of the target object on the endoscope. This is data on the types of endoscopes. This refers to the motion data of the instrument, which describes the control behavior of the target object on the instrument. This is data on the types of medical devices. Data for surgical subjects.

[0093] It should be noted that when a single surgical procedure data point involves multiple endoscopic actions, the endoscopic action data can be represented as follows:

[0094] (2)

[0095] in, Let p be the p-th endoscopic action in the l-th operation frame, where p = 1, 2, ..., D. D is the total number of endoscopic actions.

[0096] It should be noted that when a single surgical procedure data point involves multiple instrument movements, the instrument movement data can be represented as follows:

[0097] (3)

[0098] in, Let q be the q-th instrument action in the l-th operation frame. q = 1, 2, ..., E. E is the total number of instrument actions.

[0099] Furthermore, when a single surgical procedure involves multiple instruments, taking a dual-instrument approach as an example, the instrument type data can be represented as follows:

[0100] (4)

[0101] in, , These represent the types of two instruments.

[0102] Correspondingly, the motion data of the equipment can be represented as:

[0103] (5)

[0104] in, express The motion data of the type of instrument. express The motion data of each type of instrument. Similarly, the motion data of each type of instrument can also include multiple instrument actions, which can be represented by formula (3).

[0105] There are certain relationships between the data across the five dimensions. See also Figure 3 As shown in the figure, this is a schematic diagram illustrating the relationship between five dimensions provided in an embodiment of this application. The state of the endoscope can affect the state of the instruments, and the instruments can be controlled simultaneously through endoscopic and instrument movements. Furthermore, the endoscope and instruments can act on the surgical subject simultaneously.

[0106] Specifically, endoscopic motion data is related to the type of endoscope. That is, the type of endoscope affects the endoscopic actions that can be performed. Endoscope type data is also related to instrument type data and surgical subject data. That is, the type of endoscope affects the types of instruments used for endoscopic procedures, and also the surgical subjects that the instruments can operate on. Instrument motion data is related to instrument type data. That is, the type of instrument affects the instrument actions that can be performed.

[0107] The five dimensions described describe the surgical procedure from five aspects: endoscopic actions, endoscope types, instrument actions, instrument types, and surgical subjects. This essentially covers the relevant operational content of each surgical procedure in endoscopic surgery, enabling a fine-grained and accurate description of the surgical operation.

[0108] For the above five dimensions, the data can be quantified based on the characteristics of the data in each dimension, and a standardized description of the data in each dimension can be determined.

[0109] As an example, this application provides a quantification rule for five dimensions of data, which can also be called an encoding rule.

[0110] First, endoscopic movement data.

[0111] Endoscopic motion data includes first motion type data and first operation data. The first motion type data indicates the type of endoscopic motion. The first operation data indicates the content of the endoscopic motion. The specific representation of the first motion type data and the first operation data can be set as needed.

[0112] In one possible implementation, the first action type data is represented by two letters. The first operation data is represented by two numbers. The letter corresponding to each endoscopic action type in the first action type data can be preset. The unit used in the first operation data can also be preset. Furthermore, the first operation data can describe the change compared to the surgical operation in the previous operation frame, or it can be a cumulative amount compared to the initial state without operation; it can be flexibly set as needed, and this application does not limit it in this regard.

[0113] As an example, this application provides the letters corresponding to the first action type data of six endoscopic actions, as well as the data type and unit of the first operation data.

[0114] The endoscopic action is to rotate the large dial. The first action type data is BR, and the first operation data is the rotation angle, in degrees.

[0115] The endoscopic action is to rotate the small dial. The first action type data is SR, and the first operation data is the rotation angle, in degrees.

[0116] The endoscopic action is to rotate the endoscope body. The first action type data is ER, and the first operation data is the rotation angle, in degrees (°).

[0117] The endoscopic action is called endoscopic advancement. The first action type data is AD, and the first operation data is the advance distance, in centimeters.

[0118] The endoscopic action is called endoscopic retreat. The first action type data is RT, and the first operation data is the retreat distance, in centimeters (cm).

[0119] The endoscopic action is water insufflation, the first action type data is IW, and the first operation data is water insufflation time, in seconds.

[0120] For example, "AD1.5" means the endoscope has advanced 1.5 cm. "BR30" means the large dial has been turned 30° clockwise. "IW2.0" means water injection takes 2 seconds.

[0121] Second, data on the types of endoscopes.

[0122] Endoscope type data uses abbreviations for endoscope types. For example, "Ga" represents gastroscopy, "Du" represents duodenoscope, and "Co" represents colonoscope.

[0123] Third, the data on the movement of the equipment.

[0124] The instrument motion data includes second motion type data and second operation data. The second motion type data indicates the type of motion performed by the instrument. The second operation data indicates the content of the motion performed by the instrument.

[0125] The encoding method for instrument movement data can be similar to that for endoscopic movement data. As an example, the second movement type data is represented by two letters. The second operation data is represented by two numbers. The specific representation of the second movement type data and the second operation data can be set as needed.

[0126] In the second action type data, the letter corresponding to each instrument action type can be preset. The unit used in the second operation data can also be preset. Furthermore, the second operation data can describe the change compared to the surgical operation in the previous operation frame, or it can be a cumulative amount compared to the initial state without operation. It can be flexibly set as needed, and this application does not limit it in this regard.

[0127] As an example, embodiments of this application provide some instruments and corresponding instrument action types. See Table 1 for details:

[0128] Table 1

[0129] Device Name Machine action type snare Forward, backward, rotate, open, close biopsy forceps Forward, Backward, Rotate, Open, Close Injection needle Forward, backward, withdraw needle, retract needle, puncture, injection hemostatic clip Forward, backward, rotate, open, close, pull to release Hot coagulation forceps Forward, backward, rotate, open, close, coagulate Cutting knife Forward, backward, rotate, draw the bow, cut Foreign body forceps Forward, backward, close, open support Forward, backward, release balloon Forward, backward, inflate, deflate Stone retrieval basket Forward, backward, rotate, open, contract

[0130] Most of these instruments have forward and backward movements, while some have rotational movements. For these movements, the second operational data can be calibrated according to the corresponding parameters. For example, when the instrument's movement is forward, the second operational data can be calibrated according to the operating distance; when the movement is backward, the second operational data can also be calibrated according to the operating distance; when the movement is rotational, the second operational data can be calibrated according to the rotation angle. Specifically, "AD3.0" represents the instrument moving forward 3.0 cm.

[0131] Some medical devices have specific actions; for example, injection needles have an injection action. These actions can also be calibrated with second operational data according to corresponding parameters. Specifically, when the device is an injection needle and the device action type is injection, the second operational data can be calibrated according to the injection volume.

[0132] Some instruments have opening and closing actions. These actions can be tagged with individual numbers to indicate the open or closed state, or the instrument actions can be divided into multiple levels, with different levels corresponding to different degrees of action, to ensure coding consistency. Specific settings should be configured according to needs.

[0133] For example, when the instrument is a snare, if the action type is opening, the opening action can be divided into 5 levels according to the degree of opening to calibrate the second operation data. Specifically, "Op1.0", "Op2.0", "Op3.0", "Op4.0" and "Op5.0" can be used to represent different degrees of opening of the snare. If the action type is closing, the closing action can be divided into 5 levels according to the size of the closing action to calibrate the second operation data.

[0134] When the device is a support and the device action type is release, the release action can be divided into 5 levels according to the degree of opening to calibrate the second operation data.

[0135] When the device is a balloon and the device action type is inflation, the inflation action can be divided into 5 levels according to the degree of inflation to calibrate the second operation data.

[0136] When the instrument is a biopsy forceps, 1.0 can be used to indicate the open and closed state of the biopsy forceps. For example, "Op1.0" represents the biopsy forceps open and "Cl1.0" represents the biopsy forceps closed. Of course, the open state can also be divided into 5 levels according to the degree of opening, such as "Op1.0", "Op2.0", "Op3.0", "Op4.0" and "Op5.0" to represent different degrees of opening of the biopsy forceps.

[0137] When the instrument is an injection needle, 1.0 can be used to indicate the state of needle withdrawal, needle retraction, and puncture. For example, "Op1.0" represents needle withdrawal, "Cl1.0" represents needle retraction, and "Ac1.0" represents needle puncture.

[0138] When the instrument is a hemostatic clip, 1.0 can be used to indicate the state of the hemostatic clip being pulled apart and released, such as "Ac1.0" which means the hemostatic clip has been pulled apart and released.

[0139] When the instrument is a hot coagulation clamp, 1.0 can be used to indicate the coagulation status of the hot coagulation clamp, such as "Ac1.0" representing hot coagulation clamp coagulation.

[0140] When the instrument is a cutting blade, 1.0 can be used to represent the state of the cutting blade being drawn back and cutting. For example, "Op1.0" represents the cutting blade being drawn back and "Ac1.0" represents the cutting blade cutting.

[0141] Fourth, data on the types of medical devices.

[0142] Instrument type data uses abbreviations for instrument type. For example, "Sn" represents snare, and "Bf" represents endoscopic biopsy forceps.

[0143] In some possible implementations, no instruments may be used during endoscopic surgery. Correspondingly, instrument movement data and instrument type data can be null values. The specific representation of null values ​​can be set according to the representation requirements.

[0144] Fifth, data on surgical subjects.

[0145] Surgical subject data includes anatomical location data, lesion type data, and lesion content data. Anatomical location data indicates the anatomical location. Lesion type data indicates the type of lesion. Lesion content data describes the size of the lesion. Anatomical location data and lesion type data can be represented by two letters. Lesion content data can be represented by two numbers.

[0146] The specific representation of anatomical location data, lesion type data, and lesion content data can be set as needed. The letters corresponding to each anatomical location and each lesion type in the anatomical location and lesion type data can be preset. The units used for lesion content data can also be preset.

[0147] As an example, this application provides a description of some lesion types and corresponding surgical subject data. See Table 2 for details:

[0148] Table 2

[0149] Types of lesions Description of surgical subject data polyp Polyps are further classified according to whether they have a stalk or not: pedunculated polyps are classified as lp; sessile polyps as sp; and subpedunculated polyps as ssp. The lesion content data is the polyp diameter; for example, a polyp with a diameter of 15 mm would have a lesion content data of 15. ulcer Ulcers are classified in detail based on their bleeding status: ulcers in the active phase are categorized as type A, ulcers in the healing phase as type H, and ulcers in the scarring phase as type S. The ulcer content data is the ulcer area. tumor Detailed classification is based on the TNM staging system: Lesion type data corresponding to carcinoma in situ is Tis; lesion type data corresponding to invasion of the submucosa is T1; lesion type data corresponding to invasion of the muscularis propria is T2; lesion type data corresponding to penetration of the serosa is T3; lesion type data corresponding to invasion of adjacent organs is T4. Lesion content data is the tumor diameter; for example, if the tumor diameter is 15 mm, the lesion content data is 15 mm.

[0150] Specifically, the anatomical location data of the gastric antrum can be "ga", and "ga_lp_15" indicates a pedunculated polyp 15 mm in the gastric antrum.

[0151] Taking the above encoding method as an example, see Figure 5 As shown, "This means that when facing a pedunculated polyp of 10 mm in the antrum of the stomach, the endoscope is advanced 1.0 cm and the biopsy forceps are opened to level 3.0."

[0152] The above encoding method is only an example and does not limit the encoding method of surgical operation data in this application. It can be flexibly set according to needs.

[0153] This allows for the creation of an accurate and complete description of the endoscopic surgical procedure based on multiple surgical data points ordered chronologically.

[0154] Furthermore, this application does not limit the source of surgical procedure data. In one possible implementation, the surgical procedure data may be generated directly by humans during the recording of endoscopic surgery. In another possible implementation, the surgical procedure data may be generated based on surgical record text data. For example, surgical record text data is obtained. Five dimensions of information are extracted from the surgical record text data. Based on the five dimensions of information and data specifications, surgical procedure data is generated. Data specifications are used to indicate the format requirements and encoding methods that the generated surgical procedure data must meet. The process of generating surgical procedure data from surgical record text data can be implemented based on an artificial intelligence model, such as a large language model.

[0155] S202: Construct a hidden Markov model using surgical procedure data sets.

[0156] Based on the acquired surgical procedure data set, a Hidden Markov Model (HMM) is constructed. The HMM is a probability-based time series model.

[0157] Hidden Markov Models (HMMs) include a state transition probability matrix, an observation probability matrix, and an initial state probability vector.

[0158] The state transition probability matrix describes the probability of surgical state transitions during endoscopic surgery. The observation probability matrix describes the probability of the target object performing each surgical operation in each surgical state. The initial state probability vector describes the probability of being in each surgical state at the initial moment.

[0159] Surgical status refers to the state of being during surgery. As an example, three surgical statuses are described below.

[0160] Surgery Status 1: During an upper gastrointestinal endoscopy, a polyp was found in the gastric antrum. The polyp was pedunculated and 8mm in diameter. The endoscopist determined that it was suitable to use a snare for cold resection and prepared to change the snare for the resection operation.

[0161] Surgery Status 2: During an upper gastrointestinal endoscopy, a mucosal ulcer was found at the greater curvature of the stomach body with slight bleeding, indicating an active phase. The endoscopist determined that a sample of the ulcer needed to be taken and prepared to use biopsy forceps for the sampling.

[0162] Surgery Status 3: During the removal of polyps in the upper gastrointestinal tract, the polyp is located on the greater curvature of the stomach. The snare has been extended from the endoscope and opened. In order to snare the polyp, the snare needs to be extended further and the large dial needs to be rotated.

[0163] The surgical status encompasses the current treatment methods, preliminary conclusions, ongoing procedures, and subsequent procedures.

[0164] Specifically, all possible surgical states are defined as the set of states Q in the Hidden Markov Model (HMM).

[0165] (6)

[0166] Where n is the total number of possible states.

[0167] Each surgical operation data point is defined as an observation state in the HMM, and the surgical operation data set included in the surgical operation data set is defined as the observation state set V.

[0168] (7)

[0169] Where m is the total number of possible observations.

[0170] The surgical state transitions during endoscopic surgery are defined as the state transition probability matrix in a Hidden Markov Model (HMM).

[0171] (8)

[0172] in, The probability that the surgical scenario at time t transitions from surgical state i to surgical state j is specifically expressed as:

[0173] (9)

[0174] Where i = 1, 2, 3, ..., n. j = 1, 2, 3, ..., n.

[0175] The probability of selecting different surgical procedures for the target patient based on the surgical status is defined as the observation probability matrix B:

[0176] (10)

[0177] in, This represents the target object taking action when the surgical scenario is in the "i" surgical state. The probability of a surgical procedure is specifically expressed as:

[0178] (11)

[0179] Where i = 1, 2, 3, ..., n. k = 1, 2, 3, ..., m.

[0180] Using historical statistics of endoscopic surgeries as the initial state probability vector π, the Hidden Markov Model (HMM) is represented as follows:

[0181] (12)

[0182] S203: Determine the operational indicators of the target operation object based on the Hidden Markov Model.

[0183] Based on the constructed Hidden Markov Model, the surgical procedures performed on the target patient can be quantitatively evaluated, generating operational indicators. These operational indicators are used to measure the patient's proficiency in endoscopic surgery.

[0184] As an example, operational metrics include one or more of proficiency metrics and effectiveness metrics. Proficiency metrics measure the target patient's proficiency in endoscopic surgery. Effectiveness metrics measure the effectiveness of the target patient's surgical procedures during endoscopic surgery.

[0185] For proficiency indicators, the state transition volatility of a Hidden Markov Model can be used. State transition volatility refers to the degree of fluctuation in the probability distribution of transitions between hidden states.

[0186] The state transition volatility is calculated as follows:

[0187] (13)

[0188] Where n is the total number of surgical states, This indicates the standard for switching surgical states during routine procedures. HMM generation can be constructed based on surgical operation data from standard or normative procedures.

[0189] The smaller the value, the closer the change in surgical state under actual operation is to the change in surgical state under standard operation, indicating that the target patient is skilled and accurate in operation. The larger the value, the greater the difference between the actual surgical state under the target operation and the surgical state under the standard operation. This indicates that the target operation subject experiences many unexpected situations during the operation and frequently changes the surgical operation method, suggesting that the target operation subject is unfamiliar with the operation.

[0190] For an indicator of effectiveness, the observation discrete entropy of a Hidden Markov Model can be used. Discrete entropy is an indicator used to measure the amount of uncertainty information in a discrete probability distribution.

[0191] The discrete entropy is calculated as follows:

[0192] (14)

[0193] in, This represents the probability that the surgical scenario at time t transitions from surgical state i to surgical state j. It is the state transition probability. In state The joint probability under the given probability of occurrence; r is the base, which can be set according to the requirements, for example, it can be 2. The discrete entropy is calculated by first solving for the discrete entropy of different operations during the transition from surgical state i, using one-dimensional conditional probabilities. Then, the discrete entropy of all states is summed with the weighted probability of occurrence to obtain the discrete entropy of the two-dimensional state transition matrix. A higher discrete entropy value means higher uncertainty in each surgical operation transition, reflecting redundancy and more ineffective operations of the target operation object. A lower discrete entropy value indicates higher determinism in surgical operation transitions during the operation, suggesting more precise and efficient operations of the target operation object, with a higher content of effective actions.

[0194] Furthermore, after obtaining the operational indicators, it is possible to compare them with indicator thresholds. The indicator thresholds can be operational indicators calculated using an Hidden Markov Model (HMM) constructed based on surgical operation data from standard or regulated procedures. Alternatively, the indicator thresholds can also be manually set based on the surgical situation. This application does not limit the method for determining the indicator thresholds.

[0195] Determine the relationship between the operational metric and its threshold. Based on this relationship, generate a prompt message. This prompt message indicates whether the operational metric meets the requirements.

[0196] Therefore, using five-element fine-grained surgical operation data to describe surgical procedures enables the formation of a standardized coding language that can be processed by computers, which is the foundation for the digital and intelligent development of medicine. By using the surgical operation data as an observation sequence and the surgical state as a hidden state sequence, a Hidden Model (HMM) is constructed. Using the state transition matrix and observation probability matrix of the constructed HMM, the operational indicators of the endoscopic surgery of the target patient can be calculated, enabling a more accurate assessment of whether the surgical skill level of the target patient is standardized and whether the accuracy rate is high.

[0197] Based on the method embodiments described above for determining operational parameters of endoscopic surgery, this application provides a device for determining operational parameters of endoscopic surgery. See also... Figure 6 As shown in the attached diagram, the device for determining the operational parameters of this endoscopic surgery will be described below.

[0198] The acquisition unit 601 is used to acquire a set of surgical operation data for the target operation object in endoscopic surgery. The set of surgical operation data includes multiple surgical operation data arranged in chronological order. Each surgical operation data is used to describe the surgical operation of the target operation object in the endoscopic surgery. Each surgical operation data includes data in five dimensions, namely endoscopic action data, endoscopic type data, instrument action data, instrument type data, and surgical object data.

[0199] The construction unit 602 is used to construct a hidden Markov model using the surgical operation data set. The hidden Markov model includes a state transition probability matrix, an observation probability matrix, and an initial state probability vector. The state transition probability matrix is ​​used to describe the probability of surgical state transitions during the endoscopic surgery. The observation probability matrix is ​​used to describe the probability of the target object performing each surgical operation in each surgical state. The initial state probability vector is used to describe the probability of being in each surgical state at the initial moment.

[0200] The determining unit 603 is used to determine the operation index of the target operation object according to the hidden Markov model, the operation index being used to measure the operation level of the target operation object for the endoscopic surgery.

[0201] In one possible implementation, the determining unit 603 is specifically used for:

[0202] Determine the state transition volatility based on the hidden Markov model;

[0203] The state transition volatility is used as a proficiency indicator, which measures the proficiency of the target operator in the endoscopic surgery.

[0204] And / or,

[0205] Based on the hidden Markov model, determine the observation discrete entropy;

[0206] The observed discrete entropy is used as an effectiveness index, which is used to measure the effectiveness of the surgical operation performed on the target object for the endoscopic surgery.

[0207] In one possible implementation, the surgical operation data set includes multiple subsets, each subset corresponding to an endoscopic surgery, each subset including multiple surgical operation data items ordered chronologically, and each surgical operation data item corresponding to a surgical operation of an operation frame, wherein the operation frame is obtained by dividing the endoscopic surgery according to time.

[0208] In one possible implementation, the operation frame is divided according to unit operation time;

[0209] or,

[0210] The operation frames are divided according to the changing moments of the surgical operation.

[0211] In one possible implementation, the surgical data is generated in the following manner:

[0212] Obtain surgical record text data;

[0213] The information of the five dimensions is extracted from the surgical record text data;

[0214] Based on the information from the five dimensions and the data specifications, the surgical operation data is generated. The data specifications are used to indicate the format requirements and encoding methods that the generated surgical operation data must meet.

[0215] In one possible implementation, the endoscopic action data includes first action type data and first operation data, wherein the first action type data is used to represent the type of endoscopic action and the first operation data is used to represent the content of the endoscopic action.

[0216] The data on endoscope types are represented using abbreviations for endoscope types;

[0217] The device motion data includes second motion type data and second operation data. The second motion type data is used to represent the motion type of the device, and the second operation data is used to represent the content of the device motion.

[0218] The device type data is represented by abbreviations for device types;

[0219] The surgical object data includes anatomical location data, lesion type data, and lesion content data. The anatomical location data is used to indicate the anatomical location, the lesion type data is used to indicate the type of lesion, and the lesion content data is used to describe the size of the lesion.

[0220] In one possible implementation, the device further includes:

[0221] The prompting unit is used to determine the relationship between the operation indicator and the indicator threshold; and to generate prompting information based on the relationship, the prompting information being used to indicate whether the operation indicator meets the requirements.

[0222] Based on the method embodiment provided above, this application provides an endoscopic surgery operation indicator determination device, including: a processor, a memory, and a system bus;

[0223] The processor and the memory are connected via the system bus;

[0224] The memory is used to store one or more programs, the one or more programs including instructions, which, when executed by the processor, cause the processor to perform the method for determining operational indicators of endoscopic surgery as described in any of the above embodiments.

[0225] Based on the method embodiments described above, this application provides a computer-readable storage medium storing instructions that, when executed on a terminal device, cause the terminal device to perform the method for determining endoscopic surgical operation indicators as described in any of the above embodiments.

[0226] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

[0227] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0228] It should also be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0229] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0230] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for determining operational parameters in endoscopic surgery, characterized in that, The method includes: Obtain a set of surgical operation data for the target operation object in endoscopic surgery. The set of surgical operation data includes multiple surgical operation data arranged in chronological order. Each surgical operation data is used to describe the surgical operation of the target operation object in the endoscopic surgery. Each surgical operation data includes data in five dimensions, namely endoscopic action data, endoscopic type data, instrument action data, instrument type data, and surgical object data. A hidden Markov model is constructed using the surgical operation dataset. The hidden Markov model includes a state transition probability matrix, an observation probability matrix, and an initial state probability vector. The state transition probability matrix is ​​used to describe the probability of surgical state transitions during the endoscopic surgery. The observation probability matrix is ​​used to describe the probability of the target object performing each surgical operation in each surgical state. The initial state probability vector is used to describe the probability of being in each surgical state at the initial moment. Based on the Hidden Markov Model, the operational indicators of the target operation object are determined, and the operational indicators are used to measure the operational level of the target operation object in the endoscopic surgery.

2. The method according to claim 1, characterized in that, The step of determining the operational indicators of the target operation object based on the hidden Markov model includes: Determine the state transition volatility based on the hidden Markov model; The state transition volatility is used as a proficiency indicator, which measures the proficiency of the target operator in the endoscopic surgery. And / or, Based on the hidden Markov model, determine the observation discrete entropy; The observed discrete entropy is used as an effectiveness index, which is used to measure the effectiveness of the surgical operation performed on the target object for the endoscopic surgery.

3. The method according to claim 1, characterized in that, One surgical operation data point corresponds to a surgical operation frame, which is a surgical operation divided into time segments for the endoscopic surgery.

4. The method according to claim 3, characterized in that, The operation frames are divided according to the unit operation time; or, The operation frames are divided according to the changing moments of the surgical operation.

5. The method according to claim 1, characterized in that, The surgical data was generated in the following way: Obtain surgical record text data; The information of the five dimensions is extracted from the surgical record text data; Based on the information from the five dimensions and the data specifications, the surgical operation data is generated. The data specifications are used to indicate the format requirements and encoding methods that the generated surgical operation data must meet.

6. The method according to claim 1, characterized in that, The endoscopic motion data includes first motion type data and first operation data. The first motion type data is used to represent the type of endoscopic motion, and the first operation data is used to represent the content of the endoscopic motion. The data on endoscope types are represented using abbreviations for endoscope types; The device motion data includes second motion type data and second operation data. The second motion type data is used to represent the motion type of the device, and the second operation data is used to represent the content of the device motion. The device type data is represented by abbreviations for device types; The surgical object data includes anatomical location data, lesion type data, and lesion content data. The anatomical location data is used to indicate the anatomical location, the lesion type data is used to indicate the type of lesion, and the lesion content data is used to describe the size of the lesion.

7. The method according to claim 1, characterized in that, The method further includes: Determine the relationship between the operational indicators and the indicator thresholds; Based on the size relationship, a prompt message is generated, which is used to indicate whether the operation indicator meets the requirements.

8. A device for determining operational parameters in endoscopic surgery, characterized in that, The device includes: The acquisition unit is used to acquire a set of surgical operation data for the target operation object in endoscopic surgery. The set of surgical operation data includes multiple surgical operation data arranged in chronological order. Each surgical operation data is used to describe the surgical operation of the target operation object in the endoscopic surgery. Each surgical operation data includes data in five dimensions, namely endoscopic action data, endoscopic type data, instrument action data, instrument type data, and surgical object data. The construction unit is used to construct a hidden Markov model using the surgical operation data set. The hidden Markov model includes a state transition probability matrix, an observation probability matrix, and an initial state probability vector. The state transition probability matrix is ​​used to describe the probability of surgical state transitions during the endoscopic surgery. The observation probability matrix is ​​used to describe the probability of the target object performing each surgical operation in each surgical state. The initial state probability vector is used to describe the probability of being in each surgical state at the initial moment. The determining unit is used to determine the operational indicators of the target operation object based on the hidden Markov model, wherein the operational indicators are used to measure the operational level of the target operation object in the endoscopic surgery.

9. A device for determining operational parameters in endoscopic surgery, characterized in that, include: Processor, memory, system bus; The processor and the memory are connected via the system bus; The memory is used to store one or more programs, the one or more programs including instructions that, when executed by the processor, cause the processor to perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on a terminal device, cause the terminal device to perform the method described in any one of claims 1-7.

Images