Patent Application
20240206981
The disclosed technology is generally related to guiding navigation of a medical device in a body lumen using fuzzy logic, medical device parameters, user input, and anonymous patient data.
Current technology includes catheter systems that enable navigation of a catheter tip through a tortuous lumen of the body to a target. This technology involves the real-time movement of a catheter using medical images, e.g., computerized tomography (CT) images, of a target area of the body. Navigation may be presented as a recommended path based on the structure of the anatomy (e.g., an airway bronchiole structure) shown in the medical images. During navigation, however, the catheter systems may be unable to cause the distal portion of the catheter to reach all positions with an acceptable alignment with the target area.
The techniques of this disclosure generally relate to applying fuzzy logic to decision making and decision-guiding user interfaces to improve diagnostic yields and procedure efficiency. The techniques of this disclosure enable improved catheter system performance (e.g., diagnostic or therapy procedure time and success rate) through recommendations of best-practice navigation to the user, e.g., clinician or surgeon, based on the parameters of the physical performance of the medical device.
One aspect of the disclosure is directed to a method including receiving image data corresponding to a branched structure and receiving procedure information. The method also includes determining possible navigation paths for navigating a medical device within the branched structure using machine learning based on the image data and the procedure information; and displaying the possible navigation paths and controls for user selection of one of the possible navigation paths. Other implementations of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods and systems described herein.
Implementations of this aspect of the disclosure may include one or more of the following features. The procedure information may include at least one of medical device physical characteristics, medical device physical behaviors, tissue information, previous procedure data, published peer-reviewed results, or path tortuosity. The method may include displaying maximum distances of the possible navigation paths. The device physical characteristics may include bend radius, stiffness for axial or transverse loads applied to a distal end portion of the medical device, multiplanar or single planar behavior, articulating section length, or number of articulation joints. The method may include determining a probability of success for navigation from all of the possible navigation paths for navigation based on physical characteristics of the medical device and tissue information; and displaying an indication of a probability for success for navigation for each path.
Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium, including software, firmware, hardware, or a combination of them installed on the system that, in operation, causes or cause the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by a data processing apparatus, e.g., a processor, cause the data processing apparatus to perform the actions.
A further aspect of the disclosure is directed to a method including receiving image data and receiving procedure information. The method also includes receiving previous clinician experience data. The method also includes determining possible navigation plans for navigating a medical device within a branched structure based on the image data and the procedure information; determining, for each possible navigation plan, a score for a difficulty to a clinician of carrying out the possible navigation plan using machine learning based on previous clinician experience data; and displaying the score for each possible navigation plan. Other implementations of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods and systems described herein.
Implementations of this aspect of the disclosure may include one or more of the following features. The method may include tracking a position of the medical device during a procedure. The method may include determining that navigation of the medical device is unsuccessful based on the position of the medical device. The method may also include in response to determining that navigation of the medical device is unsuccessful, displaying a suggestion to use a different navigation plan. The method may include receiving previous procedure information. The method may include estimating, for each possible navigation plan, procedure time based on previous procedure information; and displaying the procedure time for each possible navigation plan. The method may include determining procedures similar to a current procedure. The method may include estimating the procedure time based on the previous procedure information for the current procedure and procedures similar to the current procedure.
The method may include determining a success rate or risk information for each navigation plan. The method may include displaying the success rate or the risk information for each navigation plan. The method may include determining positions of critical structures near each of the possible navigation plans. The method may include displaying warning indications for the positions of the critical structures. The method may include determining a success rate of a current clinician relative to success rates of other clinicians. The method may include displaying an indication of the success rate of the current clinician relative to success rates of other clinicians. The method may include receiving selection of a navigation plan from a current clinician.
The method may include determining that a current procedure is ended. The method may include in response to determining that the current procedure is ended, prompting the current clinician to input at least one outcome of the current procedure. The method may also include associating the navigation plan selected by the current clinician with the at least one outcome. The at least one outcome may be a complication during the current procedure or after the current procedure is ended. Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium, including software, firmware, hardware, or a combination of them installed on the system that, in operation, causes or cause the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions.
A further aspect of the disclosure is directed to a navigation system including a human-controlled medical device. The navigation system also includes a position sensor coupled to the human-controlled medical device. The navigation system also includes a processor; and a memory having stored thereon a physical parameter and a behavior parameter of the human-controlled medical device, previous procedure information, and instructions, which, when executed by the processor, cause the processor to: receive preoperative image data; determine possible navigation plans within a branched structure based on the preoperative image data, the physical parameter, the behavior parameter of the human-controlled medical device, and previous procedure information; and display the possible navigation plans and controls for clinician selection of one of the possible navigation plans. Other implementations of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods and systems described herein.
Implementations of this aspect of the disclosure may include one or more of the following features. The physical parameter may be in a look-up table stored on the memory and the behavior parameter may be based on previous performance data. The instructions, when executed by the processor, may cause the processor to determine that a current procedure is ended. The instructions, when executed by the processor, may cause the processor to in response to determining that the current procedure is ended, prompt a current clinician to input procedure performance information of the current procedure. The instructions, when executed by the processor, may cause the processor to store the current procedure performance information in association with the previous procedure information.
The instructions, when executed by the processor, may cause the processor to determine a control mode for the human-controlled medical device based on the preoperative image data. The instructions, when executed by the processor, may cause the processor to display a message requesting a current clinician to confirm whether to proceed in the control mode. The control mode may be a delicate tissue mode, a proximal lesion mode, an off-airway mode, or an ablation mode. The instructions, when executed by the processor, may cause the processor to prompt a current clinician to input patient information relevant to a current procedure. The instructions, when executed by the processor, may cause the processor to receive patient information. The instructions, when executed by the processor, may cause the processor to store the patient information in association with previous procedure information.
Implementations of the described techniques may include hardware, a method or process, or computer software on a computer-accessible medium, including software, firmware, hardware, or a combination of them installed on the system that, in operation, causes or cause the system to perform the actions. One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions.
The details of one or more aspects of the disclosure are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the techniques described in this disclosure will be apparent from the description and drawings, and from the claims.
Ideally, a catheter would be able to access all desired positions within an airway tree or other anatomical pathway. Reaching a desired position near a target, e.g., a lesion, alone is insufficient when sampling and/or treating tissue. Optimal alignment of the working channel is also critical to successful placement of a biopsy or therapy device at a target. However, during navigation, catheter systems may be unable to cause the distal portion of a catheter to achieve all positions with acceptable alignment with target. Some of the latest technology, including robotics, involves high manufacturing costs, large capital expenditures, and reprocessing of device components for limited uses to obtain favorable profit margins.
For catheter devices manufactured according to traditional catheter construction methods, performance approximates more complex devices in many cases but not in a universal way. In addition, catheter device performance may vary based on the skill sets, training, experience, and comfort of the users. Navigation is currently presented as a path based on the structure of the anatomy, such as airway bronchiole structure. Other variables which may impact how a procedure progresses includes tissue integrity, limitations of the catheter system devices (e.g., the scope and catheter), tool stiffnesses, nearby critical anatomical structures (e.g., the pleura or blood vessels greater than minimum size), and the goal of the procedure (e.g., biopsy or ablation).
The techniques of this disclosure generally relate to applying machine learning techniques, such as fuzzy logic (which mimics the logic of human thought), to decision-making and decision-guiding user interfaces to improve diagnostic yields and procedure efficiency. The techniques enable improved catheter system performance (e.g., diagnostic or therapy procedure time and success rate) through recommendations to a user, e.g., a clinician, of best-practice navigation based on the parameters of the physical performance of the medical device, e.g., minimum bend radius, articulating length, propensity to whip in rotation, pushability, torqueability, cross-sectional area, minimum or maximum diameter, and/or non-linear or asymmetric articulation behaviors.
Implementation details of such techniques, systems incorporating such techniques, and methods using the same are described below. However, these implementation details are merely examples of the disclosure, which may be embodied in various forms. Therefore, specific structural and functional details disclosed herein are not to be interpreted as limiting, but merely as a basis for the claims and as a representative basis for allowing one skilled in the art to variously employ this disclosure in virtually any appropriately detailed structure. While the example implementations described below are directed to the bronchoscopy of a patient's airways, those skilled in the art will realize that the same or similar devices, systems, and methods may also be used in other lumen networks, such as, for example, the vascular, lymphatic, and/or gastrointestinal networks.
Further, while described in connection with fuzzy logic, other artificial intelligence and/or machine learning methods may be employed to perform parameter-based path planning without departing from the scope of this disclosure. For example, neural networks, Markov decision processes, and/or others may be employed. Machine learning may include reinforcement learning, which involves rewarding desired behaviors and punishing negative behaviors. The reinforcement learning may assign positive values to the desired behaviors, e.g., medical procedure steps and/or actions that lead to a successful outcome, and may assign negative values to undesired behaviors, e.g., medical procedure steps and/or actions that lead to an unsuccessful outcome. The medical procedure steps and/or actions and the corresponding outcomes may be based on historical medical data.
With reference to
EMN system 10 generally includes an operating table 40 configured to support a patient; a bronchoscope 50 configured for insertion through the patient's mouth and/or nose into the patient's airways; monitoring equipment 60 coupled to bronchoscope 50 for displaying video images received from bronchoscope 50; a tracking system 70 including a tracking module 72, a plurality of reference sensors 74, and an electromagnetic field generator 76; a workstation 80 including software and/or hardware used to facilitate pathway planning of this disclosure, identification of target tissue, navigation to target tissue, and digitally marking the biopsy location.
As illustrated in
Catheter guide assemblies 90, 100 including LG 92 and EWC 96 are configured for insertion through a working channel of bronchoscope 50 into the patient's airways (although the catheter guide assemblies 90, 100 may alternatively be used without bronchoscope 50). LG 92 and EWC 96 are selectively lockable relative to one another via a locking mechanism 99. A six degrees-of-freedom electromagnetic tracking system 70, or any other suitable positioning measuring system, is utilized for performing navigation, although other configurations are also contemplated. Tracking system 70 is configured for use with catheter guide assemblies 90, 100 to track the position of EM sensor 94 as it moves in conjunction with EWC 96 through the airways of the patient, as detailed below.
As shown in
Also shown in
Although navigation is detailed above with respect to EM sensor 94 being included in LG 92 it is also envisioned that EM sensor 94 may be embedded or incorporated within biopsy tool 102 where biopsy tool 102 may alternatively be utilized for navigation without need of LG 92 or the necessary tool exchanges that use of LG 92 requires.
During procedure planning or navigation, workstation 80 may utilize computed tomographic (CT) image data for generating and viewing the 3D model of the patient's airways, which may enable the identification of target tissue on the 3D model (automatically, semi-automatically or manually), and may allow for the selection of a pathway through the patient's airways to the target tissue. More specifically, the CT scans may be processed and assembled into a 3D volume, which is then utilized to generate the 3D model of the patient's airways. The 3D model may be presented on a display monitor associated with workstation 80, or in any other suitable fashion according to some aspects of this disclosure. Using workstation 80, various slices of the 3D volume and views of the 3D model may be presented and/or may be manipulated by a clinician to facilitate identification of a target, generation and display of possible pathways, and selection of a suitable pathway through the patient's airways to access the target. The 3D model may also show marks of the locations where previous biopsies were performed, including the dates, times, and other identifying information regarding the tissue samples obtained. These marks may also be selected as the target to which a pathway can be planned. Once selected, the pathway is saved for use during the navigation procedure.
During navigation, EM sensor 94, in conjunction with tracking system 70, enables tracking of EM sensor 94 and/or biopsy tool 102 as EM sensor 94 or biopsy tool 102 is advanced through the patient's airways.
Turning to
Memory 202 includes any non-transitory computer-readable storage media for storing data and/or software that is executable by processor 204 and which controls the operation of workstation 80. In one implementation, memory 202 may include one or more solid-state storage devices such as flash memory chips. As an alternative to or in addition to the one or more solid-state storage devices, memory 202 may include one or more mass storage devices connected to the processor 204 through a mass storage controller (not shown) and a communications bus (not shown).
Although the description of computer-readable storage media contained herein refers to solid-state storage, it should be appreciated by those skilled in the art that computer-readable storage media can be any available media that can be accessed by the processor 204. That is, computer-readable storage media may include non-transitory, volatile, non-volatile, removable, and/or non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules, or other data. For example, computer-readable storage media includes RAM, ROM, EPROM, EEPROM, flash memory or other solid state memory technology, CD-ROM, DVD, Blu-Ray or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium, which can be used to store the desired information and which can be accessed by workstation 80.
Memory 202 may store application 81 and/or CT/CBCT data 214. Application 81 may, when executed by processor 204, cause display 206 to present user interface 216. The application 81 may include a machine learning application or system as described herein. The user interface 216 may include one or more aspects of the user interfaces described below. Network interface 208 may be configured to connect to a network such as a local area network (LAN) consisting of a wired network and/or a wireless network, a wide area network (WAN), a wireless mobile network, a Bluetooth network, and/or the internet. Input device 210 may be any device by means of which a user may interact with workstation 80, such as, for example, a mouse, keyboard, foot pedal, touch screen, and/or voice interface. Output module 212 may include any connectivity port or bus, such as, for example, parallel ports, serial ports, universal serial busses (USB), or any other similar connectivity port known to those skilled in the art.
Though described in connection with a bronchoscopic implementation, this disclosure is not so limited and other motor-driven medical device or catheter systems, which may include one or more cameras, may be employed without departing from the scope of the disclosure. Similarly, the methods described herein may be executed in connection with one or more robotic- or manually-controlled and robotically-driven catheter systems. In such implementations, the bronchoscope may be replaced in full or in part by a catheter, catheter sleeve, steerable catheter, and other similar devices.
In aspects, various input variables may be stored and accessed by the system to determine possible navigation paths using machine learning techniques including a fuzzy logic system, and display them to the user or clinician for review and/or selection. The fuzzy logic system may include a knowledge database for storing the various input variables. And the various input variables may be used to define fuzzy rules, which may be used by a fuzzy inference engine to determine fuzzy outputs based on inputs conforming to predefined input variables. The input variables may include device physical behaviors. The device physical behaviors may include at least one of bend radius, stiffness for axial and transverse loads applied to the distal tip of the catheter, multiplanar or single planar behavior, articulating section length, and number of articulation joints. The device physical behaviors may be learned using machine learning or may be accessed from a look up table stored in memory.
In aspects, the input variables may include tool physical behaviors. The tool physical behaviors may include physical behaviors associated with at least one of a needle, a brush, forceps, a wire, and an ablation probe. The tool physical behaviors may be stored in an RFID incorporated in or on the medical device, e.g., positioned at a handle of the device, and/or may be received from the user in the form of planning information (e.g., one or more case goals). For example, the user may inform the system of the current tool through, for example, a suitable user interface. A manufacturer or other third party may manage a look up table of tool behaviors.
In aspects, the input variables may include tissue information. The tissue information may include indications of emphysema, the approximate age of tissue (which may be input according to a broad de-identified stratification of the original age data to comply with relevant health laws or regulations, e.g., the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rules), or information regarding normal anatomy, abnormal anatomy, adhesions, or previous resections. The tissue information may include information inferred from preoperative imaging or information entered by the user during path planning. The tissue information may include anonymous data collected from previous cases or console data.
In aspects, the input variables may include inputs from previous case data or published peer reviewed results. For example, the input variables may be based on collected, past data that have led to a successful result. The previous case data may be obtained from recordings of cases or procedures by a catheter system console. The previous case data may also be obtained from user inputs including inputs from experienced account managers or input from current users, e.g., a surgeon or other clinician involved with the case.
In aspects, the input variables may include path tortuosity data. The path tortuosity data may include turn angles and the distance between turns. For example, the path tortuosity data may include a first turn having a first angle, e.g., 90 degrees, a second turn having a second angle, e.g., 193 degrees, and a distance, e.g., 2 cm, between the first turn and the second turn. The path tortuosity data may also indicate bends or turns that are unobtainable by the catheter system. The planning software may determine various path tortuosity data including one or more of: a number of turns, a distance between turns, which may include the length and angle of the airway segment, off-airway distances, and undersized airway distance. Then, the path tortuosity data may be compared to the physical and behavioral data of the medical device to determine possible paths to a target. For example, the path tortuosity data may be compared to the physical and behavioral data of the medical device to determine whether turns within candidate paths to the target are obtainable.
As will be appreciated, other factors in connection with the path tortuosity are the physical limitations of the catheter and tool combination. Certain catheters may have their bend radius greatly reduced by the tools to be deployed therethrough and thus the viability of a given navigation path may be affected by tool choice or catheter choice as well as the choice of their mutual deployment. The machine learning techniques of this disclosure may take into account these and other similar factors.
In aspects, the surgical navigation system may be based on a minimum amount of shared analysis between the user and the surgical navigation system to determine a navigation path to follow. In one implementation, two or more possible navigation paths may be presented on a display. The two or more possible navigation paths may be determined by a machine learning system based on one or more input variables. In some aspects, the possible navigation paths are determined based on various criteria, which may be incorporated into the machine learning, including airway criteria, e.g., “cross no blood vessels larger than 1 mm” or “maximum needle excursion of 1.5 cm.”
In aspects, the system may suggest, based on machine learning, nearby likely airway candidates for achieving alignment.
In aspects, the surgical navigation system may be based on a small amount of shared analysis between the user and the surgical navigation system to determine a navigation path to follow. The two or more possible navigation paths may be presented without weighting and without using machine learning to recommend one of the two or more possible navigation paths. Accordingly, the user can select one of the two or more possible navigation paths based on the user's skill and experience. In some aspects, the surgical navigation system may identify articulation requirements in order to navigate the medical device, e.g., catheter, according to each of the possible navigation paths. For example, during final positioning, the surgical navigation system may identify an articulation sequence to navigate the catheter to a position and/or orientation to successfully aim the tool at a target, such as a tissue structure.
In aspects, the surgical navigation system may be based on a medium amount of shared analysis between the user and the surgical navigation system to determine a navigation path to follow. In one implementation, two or more possible navigation paths may be determined using a machine learning algorithm and may be displayed with an indication of probabilities for or degrees of an acceptable or successful outcome. The two or more possible navigation paths may be determined alone or in combination with potential risks or hazards along with certain details on the identification of those hazards, which is illustrated in
The surgical navigation system, utilizing machine learning such as fuzzy logic may prioritize possible navigation paths based on the input variables described herein including tissue information, physical or behavioral parameters of the intraluminal medical device, and/or previous user experience. For example, the surgical navigation system may take into consideration tissue condition, the arrangement of neighboring airways, proximity of airways to the target, and the feasibility of alignment of the intraluminal medical device (e.g., a catheter and a tool) with the target. In some aspects, the system may prioritize the feasibility of alignment over proximity of the airway to the target.
In analyzing probabilities for successful outcomes from among the possible navigation paths, the system may analyze the following considerations. As illustrated in
The surgical navigation system may also determine, using machine learning, and display a difficulty score for each of the two or more possible navigation paths.
The probability of success score or difficulty for each of the possible paths may be indicated by the color of the displayed path. For example, a first possible path may be colored red to indicate that the first possible path has the lowest probability of success score, a second possible path may be colored green to indicate that the first possible path has the highest probability of success score, and a third possible path may be colored yellow to indicate that the third possible path has a probability of success score between the probability of success scores for the second and third possible paths.
In aspects, the system may determine possible paths that avoid conflicts. One example of a conflict may include a sharp turn, which cannot be navigated by the catheter based on the input characteristics of the catheter. For example, the input characteristic of the catheter is that it cannot be articulated in such a way as to make the sharp turn. In aspects, the surgical navigation system may make suggestions in real time while information is being gathered during the procedure. For example, as illustrated in
In aspects, relevant patient data may be communicated from an electronic health record (EHR) system or a picture archiving and communication system (PACS) to the surgical console. All patient data may be de-identified, roughened, and/or anonymized prior to being communicated to the surgical console to allay privacy concerns.
In aspects, the system may display the expected procedure time to the user. As illustrated in
In aspects, the system may display relative success rates or risks for the top three possible paths through a branched structure of airways. For example, the system may display probability of pneumothorax, blood vessel puncture, or any other risks associated with the procedure. The system may determine positions of critical structures and display the positions of the critical structures. As illustrated in
In aspects, the system may implement a game among users using the result of the analysis of users' performance data. For example, the system may provide feedback to a user on success rates 608 relative to the user's peers, as illustrated in
In aspects, the system may gather data after the procedure. For example, the system may prompt the user to input information regarding the performance of the procedure. The performance information may be self-reported by the surgical navigation system to a medical database. The performance information may include a relationship between the expected plan and the observed outcome. The performance information may be used as a closed-loop input to the surgical navigation system. In aspects, the EMR or PACS may be updated to include intra-operative and post-operative information. For example, the EMR or PACS may be updated to include intra-operative or post-operative information relating to any complications. The intra-operative and post-operative information may be updated at predetermined times, e.g., during the procedure, one week after the procedure, one month after the procedure, one year after the procedure, or at any combination of times.
In aspects, the surgical navigation system may incorporate position sensing information to access natural orifices. The manufacturer of the surgical navigation system may store known structural parameters, behavioral parameters, and/or collected performance data of the access catheter or device in memory. The parameters and/or performance data may be organized in a look-up table. The system may include computer or data processing devices for collecting and anonymizing data from previous procedures. After the procedure, the surgical navigation system may store procedure data and/or prompt the user to input data that may be used for future similar procedures. For example, the surgical navigation system may store an actual case time, which may be entered by a user or may be automatically stored by the system in system memory or in a secured server.
The system may present questions relating to the procedure and enable the user to input an answer, such as a binary yes/no answer, a nonbinary answer (e.g., a scale from 1 to 5), or a textual answer, or a combination of two or more of these answer types. For example, the questions may include one or more of the following questions: “Did you accomplish your navigation goals for this case?,” “Did you accomplish your alignment or target biopsy goals for this case?,” “How do you feel the system-generated assistance performed today?,” or “Any observed complications as of the time of this entry?”
During planning or during a procedure, the user interface may prompt the user based on preoperative imaging. For example, the system may analyze preoperative imaging, and, based on that analysis, the user interface may display the following message including a prompt as follows: “From preoperative imaging, it appears the patient displays compromised lung tissue structure. Would you like to proceed in delicate tissue mode?” The system may analyze the preoperative imaging based on computer-aided diagnosis methods that perform image processing on preoperative imaging to recognize possible issues with tissue structures. The system may analyze the preoperative imaging using machine learning-based techniques.
For procedures involving human-controlled or human-actuated devices, the surgical navigation system may recommend a control mode. The control modes may include a delicate tissue mode, a proximal lesion mode, an off-airway mode, an ablation mode, or any combination of these modes. In the delicate tissue mode, the surgical navigation system may prioritize various navigation strategies. The surgical navigation system may prioritize: straight-on navigation, larger airways for the first portion of the path (e.g., 90%), and routes which do not require large changes in direction during navigation (e.g., a first angle of 90 degrees and a second angle of 270 degrees). The surgical navigation system may identify a particular tool, a particular set of tools, or an order in which a set of tools are used in a procedure. The proximal lesion mode may, for example, prioritize paths which result in greater than three points of estimated contact by a catheter with nearby airway walls, and may prioritize paths which place the catheter tip proximal to the lesion to facilitate performing a procedure on the lesion.
The control modes may include an off-airway mode. The off-airway mode may be recommended in cases where following the airway route would lead to a decreased chance of success. Additionally, or alternatively, the control modes may include an ablation mode, by which the surgical navigation system utilizes fuzzy logic and/or machine learning to determine and recommend possible paths and/or plans for successfully completing an ablation procedure. The surgical navigation system may utilize user inputs leading to hard-coded recommendations for the user and/or learned and/or taught recommendations for the user. The surgical navigation system may prompt a user for a variety of inputs regarding the patient, e.g., medical or surgical history information, via a user interface. The inputs may be acquired during planning or as optional entries during navigation and used in the machine learning techniques (e.g., fuzzy logic algorithms) of this disclosure.
For example, before a procedure starts, the system may display user interface 700 of
The user interface 700 may further include a data field 714, in which text may be entered by the user to identify the abnormal anatomy. The system may prompt the user to input information regarding previous resections. The system may prompt the user to input the approximate age of the patient, e.g., the decade in which the patient's age falls. For example, as illustrated in
To build a knowledge database and implement machine learning, the surgical navigation system may prompt a user to input information relating to the user via a user interface. For example, the system may display user interface 700, in which the user, e.g., a surgeon, may enter information regarding the user in relation to a procedure under the heading “Surgeon Information” 720. The user information may include the user's experience or comfort level with performing a particular procedure or portion of the procedure. For example, the surgical navigation system may prompt a user to input information regarding the user's comfort with off-airway access “tunneling” 724 to the target. The user may input more specific information regarding the user's comfort level for the location or length of the off-airway access tunneling, e.g., the user's comfort levels for the off-airway distances.
The user information may include the user's medical opinion. The user's medical opinion may include the user's medical opinion on the patient's anatomy, such as the peripheral blood vessels 726. The user's medical opinion may indicate the level of care needed to successfully complete the procedure. The artificial intelligence, machine learning, or fuzzy logic methods of this disclosure may then be used to process the patient information 710, the surgeon information 720, and/or anonymized patient and surgeon information to determine possible and/or recommended surgical plans and/or paths, and to display the possible and recommended surgical plans and/or paths to the user.
At block 806, tissue information is received. The tissue information may include indications of abnormal or diseased tissue or tissue with a condition, e.g., emphysema, the approximate age of tissue (which may be input according to a broad de-identified stratification of the original age data to comply with relevant health laws or regulations, e.g., the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rules), and/or information regarding normal anatomy, abnormal anatomy, adhesions, or previous resections. The tissue information may include information inferred from preoperative imaging or information entered by the user during path planning. The tissue information may include anonymous data collected from previous cases or console data.
In aspects, other or different procedure information may be received and used in a machine learning system of this disclosure. The procedure information may include medical device physical behaviors, previous procedure data, published peer-reviewed results, and/or path tortuosity.
At block 808, possible navigation paths for navigating a medical device within the branched structure are determined based on the image data, the physical characteristics of the medical device, and the received tissue information. At block 810, a probability for success of navigation for the possible paths for navigation is determined based on physical characteristics of the intraluminal medical device and the tissue information. Then, before ending at block 814, possible navigation plans, an indication of the probability of success of navigation, and user controls for selecting one of the possible navigation plans are displayed at block 812.
Also, a procedure time for a possible navigation plan may be estimated based on the previous procedure information. At block 905, the method 900 determines whether there is another possible navigation plan. If there is another possible navigation plan, block 904 is performed for the other possible navigation plan. If there is not another possible navigation plan, scores and estimated procedure times for possible navigation plans are displayed at block 906. At block 908, the position of the medical device is tracked. At block 910, it is determined that navigation of the medical device is unsuccessful based on the tracked position of the medical device. Then, before ending at block 914, a suggestion to use a different possible navigation plan, which may be determined using machine learning, is displayed at block 912.
It should be understood that various aspects disclosed herein may be combined in different combinations than the combinations specifically presented in the description and accompanying drawings. It should also be understood that, depending on the example, certain acts or events of any of the processes or methods described herein may be performed in a different sequence, may be added, merged, or left out altogether (e.g., all described acts or events may not be necessary to carry out the techniques). In addition, while certain aspects of this disclosure are described as being performed by a single module or unit for purposes of clarity, it should be understood that the techniques of this disclosure may be performed by a combination of units or modules associated with, for example, a medical device.
In one or more examples, the described techniques may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored as one or more instructions or code on a computer-readable medium and executed by a hardware-based processing unit. Computer-readable media may include non-transitory computer-readable media, which corresponds to a tangible medium such as data storage media (e.g., RAM, ROM, EEPROM, flash memory, or any other medium that can be used to store desired program code in the form of instructions or data structures and that can be accessed by a computer).
Instructions may be executed by one or more processors, such as one or more digital signal processors (DSPs), general purpose microprocessors, application specific integrated circuits (ASICs), field programmable logic arrays (FPGAs), or other equivalent integrated or discrete logic circuitry. Accordingly, the term “processor” as used herein may refer to any of the foregoing structure or any other physical structure suitable for implementation of the described techniques. Also, the techniques could be fully implemented in one or more circuits or logic elements.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2022/031464 | 5/27/2022 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 63194119 | May 2021 | US |
