Patent Application Titled “Surgical Path Planning Using Artificial Intelligence For Feature Detection” Published Online (USPTO 20220142709): Patent Application
2022 MAY 30 (NewsRx) -- By a
No assignee for this patent application has been made.
Reporters obtained the following quote from the background information supplied by the inventors: “Degenerative disease of the spine is a major cause of disability in the aging population. Vertebral pathology such as degenerative disc disease or abnormal motion of spinal segments relative to each other often results in debilitating pain. Fusion of adjacent vertebrae with insertion of man-made interbodies to replace a degenerated disc can relieve a significant amount of the preoperative pain. The success of spinal fusion operations may range from 60-70% and is dependent on a number of factors, including the skill of the surgeon. Other contributing factors to the success of a spinal fusion are individual characteristics of the patient, the number and location of vertebrae to be fused, the type of hardware selected, and the surgical approach. The five primary lumbar interbody fusion (LIF) approaches are: anterior (ALIF), lateral or extreme lateral interbody fusion (LLIF or XLIF), oblique lumbar interbody fusion/anterior to psoas (OLIF/ATP), transforaminal (TLIF or MI-TLIF), and posterior (PLIF). Each of these approaches has advantages and disadvantages, and varying rates of success. In the context of this application, the term ‘approach’ refers to a specific surgical method or technique, in contrast to a physical path, or a movement closer to a target. This distinction is important, in order to distinguish this use of the term “approach”, from the alternative meaning of a particular path chosen from the entry point on the skin to reach the surgical target, which is also discussed in the application.
“The specific approach selected for a spinal fusion may depend on both the clinical status and physical characteristics of the patient, as well as surgeon preference. The fact that there are so many possible approaches suggests that no single approach is ideal for each application. A given approach may be preferable for a thin patient as compared to one who is grossly overweight, or for a specific vertebral level, or for a patient with pre-existing conditions such as asthma or other respiratory insufficiency. The learning curve for spinal fusion is difficult and requires months or years of practice to become proficient in a given technique or approach. Thus, the multiplicity of options may lead to suboptimal outcomes, as it is not possible for a single surgeon to master all approaches with equal efficiency. It is thus probable that some patients have less than optimal outcomes because of the limits of human expertise. Typically, a surgeon selects the approach that he/she has most experience in performing and feels the most comfortable with, rather than basing the choice of approach on the most likely to succeed, out of the several options possible for a given patient.
“Documents pertaining to planning the surgical approach to spinal fusion, or for planning the path or positioning of an instrument to be inserted into a patient, include:
“
““Lumbar interbody fusion: techniques, indications and comparison of interbody fusion options including PLIF, TLIF, MI-TLIF, OLIF/ATP, LLIF and ALIF.” R J Mobbs, K Phan, G Malham, K Seex, and P J Rao. J Spine Surg. 2015 December; 1(1): 2-18. doi: 10.3978/j.issn.2414-469X.2015.10.05.
““Path Planning for Automation of Surgery Robot based on Probabilistic Roadmap and Reinforcement Learning,” D Baek, M Hwang, H Kim and D S Kwon, presented at the 15th
“
In addition to obtaining background information on this patent application, NewsRx editors also obtained the inventor’s summary information for this patent application: “Embodiments of the present disclosure generally relate to methods of planning robotically-controlled insertion of a surgical instrument or hardware to be implanted. Though the methodology is applicable to any region of surgical intervention, these embodiments use the example of spinal surgery to illustrate the methods and techniques proposed, spinal surgery being a particularly complicated field because of the problems of access to the surgical site, and the sensitivity of the organs close to the operating site, such as the spinal cord, spinal nerves, and the major blood vessels in the vicinity.
“A large amount of clinical data has been acquired over the past several decades regarding the success of various approaches to interbody fusion. Individual surgeons generally select the method with which they are most familiar. Some approaches may be more suitable for a given type of patient, based on a variety of factors such as vertebral levels to be fused, body type, age, and other clinical factors. For various approaches, the patient needs to be positioned on the operating table in a supine orientation; for other approaches, the patient is positioned in a lateral or prone position. Once a particular approach has been selected and the operation is in progress, it is difficult if not impossible to change the type of procedure. Thus, the selection of the optimal approach pre-operatively is important. It would thus be beneficial to automate the methodology for selecting the best technique for spinal fusion, such that the specific approach chosen and the unique path from origin to target region, would be selected to achieve the best possible outcome for a given patient, regardless of the individual experience and skills of the surgeon. To make this determination, algorithms of artificial intelligence (AI) may be applied to a number of data sources, comprising for example, clinical outcomes from previous operations, clinical data from the patient, and other derived and calculated outcomes. These clinical data may be compiled from insurance databases, health organizations, or other sources. The processed data from the AI analysis may be stored in available storage such as in the cloud, through a cloud application programming interface, for access by the system when analyzing data from a new patient, in order to make a determination of the best surgical approach and plan for the procedure.
“The system is capable of looking at the tissues that the tool passes through on its way from the entry point on the skin to the target point on or between the vertebrae, and to recognize which tissues are permissible to move and with how much pressure, which ones should not be moved or even touched, which are safe to cut or even remove, and how much is removable. These determinations are made using AI algorithms, as further detailed in the Detailed Description hereinbelow. In one exemplary implementation, the system is used for planning an operation for interbody insertion. As is mentioned in the Background Section, several anatomical approaches exist for insertion of an artificial disc or interbody. These methods include gaining access to the vertebral column from anterior, lateral, or posterior anatomical approaches. Some of these approaches require removal of the vertebral lamina, or the vertebral facet, in order to find a path to insert the interbody. The system therefore needs to plan how much of the bone is necessary and safe to remove in order to be able to insert the interbody. Also, if during the course of the operation, the location of the vertebral bodies shift relative to one another, or relative to the tracking system, the system must take this shift into account, to determine whether and how to change the position of the tool and the interbody to be inserted. The optimal path is not necessarily linear. In the field of robotics, finding the optimal path for interbody insertion involves use of an algorithm for planning motion through a milieu of obstructions or impediments which would negate the use of a simple direct insertion motion. Such algorithms are known as the navigation problem or the piano mover’s problem, for finding a sequence of valid steps that moves the robotic end effector from the source outside the body to its destination in the intervertebral disc space, through the environment with the obstructions. In this case, the problem to be solved is for inserting the interbody, held and maneuvered by the end effector or the surgical tool,
“In planning the surgical procedure, the plan must take into account the amount of spinal lamina or pedicle to be removed in order to allow insertion of the interbody between the vertebral bodies. As removing bone weakens the spinal column, the quantity and location of removable bone must be determined with consideration of multiple factors in addition to simplifying the ease of interbody insertion.
“Two main factors require consideration during the planning stage of planning the tool or instrument path. The first factor is the spinal column and how much bone must be removed, if at all, at the site of interbody insertion. The second factor is the approach to be taken. Each of the five generally known lumbar interbody fusion approaches has advantages and disadvantages, depending on a number of factors such as vertebral levels to be fused, concurrent medical conditions of the patient, prior abdominal surgery with adhesions, anatomical aberrations, or spondylolisthesis. The known approaches to spinal surgery requiring access to the intervertebral disc space, can be divided into two major categories: anterior, which access the spinal column through the soft tissues of the abdomen and reach directly to the intervertebral space; and posterior, which require dissection of the posterior paraspinal muscles and removal of vertebral lamina and/or spinal processes to access the intervertebral disc space. Implementations of the methods and systems disclosed in this disclosure relate to planning access from the anterior direction, while planning access from the posterior direction is disclosed in co-pending
“In an anterior approach, the surgeon should perform a segmentation analysis to know which tissues can be moved, and which not, which can be removed or incised, and which not. For each tissue, the system should learn the allowed procedures that can be performed on each tissue and how to carry out that procedure. The system learns about each tissue by both preoperative programming and by intraoperative exploration. For the preoperative stage, each tissue is identified and labeled with known parameters, such as density, friability, vascularity, essentiality for life, movability and other factors. The system calculates a score that determines if and how much the tissue can be moved, cut, poked, retracted, or the like, as further discussed in the above mentioned co-pending
“The details of one or more aspects of the disclosure are set forth in the accompanying drawings and the description below. It is to be understood that the scope of the methods revealed in this disclosure are applicable to many types of surgical path planning, comprising at least operations of the abdominal and thoracic cavities, as well as procedures requiring deep tissue dissection and careful avoidance of critical structures not visible by direct and open observation or by endoscopic photography, such as of the neck, throat, and skull base. One exemplary implementation is thus described in detail, specifically in regard to planning the optimal surgical approach for spinal interbody fusion, with other possibilities suggested in brief. Additional features, objects, and advantages of the techniques will be apparent from the description and drawings, and from the claims.
“There is thus provided in accordance with an exemplary implementation of the devices described in this disclosure, a method for automatically planning a tool path through tissues to a surgical target in the spine of a subject, the method comprising: (a) performing image segmentation on a set of three-dimensional medical images of a region of possible paths to the surgical target, to define boundaries of tissues located in the region of the possible paths; (b) selecting an entry point for a first planned path from the entry point to the surgical target; © using information relating to a combination of known tissue parameters to assign to each tissue in the first path, a weighting value for tissue traversability; (d) using the weighting values for tissue traversability, calculate a cost function of the first path, taking into account: (i) the sum of weighted values for tissue traversability of all tissues along the planned first path; and (ii) the clinical status of the subject; (e) repeating steps (b) to (d) for additional entry points and associated paths; and (f) selecting the path with the minimum cost function. The method may further comprise the steps of: (g) providing the planned tool path to a robotic controller, and (h) using the robotic controller to move the tool through the tissue according to the planned tool path to reach the spinal surgical target.”
There is additional summary information. Please visit full patent to read further.”
The claims supplied by the inventors are:
“1. A system for planning a safe path for robotic execution of a surgical procedure on the spine of a subject, comprising: at least one processor executing instructions stored on at least one non-transitory storage medium to cause the at least one processor to: a) access a database containing analyses of outcomes of surgical procedures on the spine of patients in a reference population; b) using the outcomes accessed in the database of patients having a similar clinical profile to that of the subject, select the surgical procedure most likely to produce a desired outcome for the subject; c) select from a set of known surgical approaches, the surgical approach expected to be optimal for executing the selected surgical procedure; and d) using a segmented three-dimensional image set annotated with predetermined tissue traversability of tissues in the region of the selected surgical approach, plan paths for robotic access of at least one surgical tool to the spine, for execution of the selected surgical procedure, wherein the optimal planned path is one which minimizes interaction of the at least one surgical tool with tissues having unfavorable traversability data.
“2. The system according to claim 1, wherein data of the predetermined tissue traversability data is collected from sources comprising at least some of scientific literature, recordings of tissue properties in prior surgical procedures, and experimental data.
“3. The system according to any of claims 1, wherein the tissue traversability data comprises information on known tissue properties including at least some of density, friability, vascularity, removability, compressibility, essentiality for life, movability, capsule or fascia fragility, or relative risk of penetration.
“4. The system according to any of claims 1, wherein the tissue traversability is assigned a weighting according to a combination of the known tissue properties.
“5. The system according to claim 4, wherein the weightings of the tissues encountered in a planned path are combined to generate a score for that path, such that the path with the most favorable score is selected as the optimal path.
“6. The system according to any of claims 1, wherein the planned path is selected to achieve at least one of a) the ability to provide adequate access to the surgical site, b) the shortest operating time to reach the site, or c) the ability of the patient to tolerate the requirements of the surgical procedure.
“7. The system according any of claims 1, wherein the database data are classified according to surgical approaches used and the clinical profile of the patients, and wherein the clinical profile comprises at least some of age, gender, BMI, concurrent bone disease, coexisting medical conditions, level of intervertebral disc disease, or clinical risk indices.
“8. The system according to any of claims 1, wherein if intervertebral disc removal is indicated for the selected surgical approach, the surgical procedure on the spinal column comprises one of artificial intervertebral disc replacement or spinal fusion with interbody insertion.
“9. A system for robotic execution of a planned procedure path using a preselected surgical approach on a subject, comprising: a) at least one processor executing instructions stored on at least one non-transitory storage medium to cause the at least one processor to implement robotic execution of the planned procedure path on the subject; b) a memory comprising the planned procedure path, and tissue traversability data that indicate a risk of interacting with each specific tissue along the planned path; and c) at least one sensor configured to provide input to the processor to update the tissue traversability data intraoperatively; wherein the input is used to update the planned procedure path intraoperatively to avoid tissues with unfavorable traversability data.
“10. The system according to claim 9, wherein the at least one sensor is at least one of an externally situated internal imaging device, a pressure detection sensor, a Doppler flow sensor, an endoscopic camera, a mechanical tonometer, a digital indurometer, a fibrometer, or an ultrasound probe.
“11. The system according to any of claims 9, wherein the tissue traversability data comprise quantitative information on known tissue properties: at least some of density, friability, vascularity, removability, compressibility, essentiality for life, movability, capsule or fascia fragility, or relative risk of penetration.
“12. The system according to claim 9, wherein the tissue traversability data comprise a series of numerical ratings for each tissue, wherein each numerical rating corresponds to one of the known tissue properties.
“13. The system according to any of claims 9, wherein the processor is configured to use at least one of training logic, inference logic, artificial intelligence algorithms, machine learning, or computer logic to execute the planned procedure path using the preselected surgical approach.
“14. The system according to any of claims 9, wherein the sensor input is used to update the planned procedure path intraoperatively in order to enhance at least one of the safety or the efficiency of the robotic execution.
“15. A system for selecting a specific surgical procedure to be performed on a subject having a clinical condition for which intervertebral disc removal is indicated, comprising: at least one non-transitory storage medium for storing instructions; and at least one processor executing the instructions stored on the at least one non-transitory storage medium, the processor performing: i) classify clinical parameters data and surgical outcome data from patients in a reference population, each patient having undergone a surgical procedure for intervertebral disc removal using any one of a set of known surgical approaches; ii) match clinical parameters of the subject to a subgroup of the reference population having an equivalent clinical condition to that of the subject; and iii) based on the classified outcome data of patients in the matched subgroup, select the specific surgical procedure and surgical approach predicted to result in an optimal outcome for the subject.
“16. The system according to claim 15, wherein the clinical conditions for which intervertebral disc removal is indicated are at least one of herniated disk, intervertebral disc disease, spinal stenosis related to disc disease, or spondylolisthesis.
“17. The system according to either claim 15, wherein the clinical data and outcome data of the reference population are derived from at least one of hospital records, health maintenance organization records, insurance company records, or the records of a surgical practice.
“18. The system according to any of claims 15, wherein the outcome data comprise at least some of time to recovery, extent of recovery, level of independence in activities of daily living, reported level of pain, Oswestry disability index score, range of mobility of the affected vertebral segment, and motor function.
“19. The system according to any of claims 15, wherein the known surgical approaches comprise anterior approach, oblique approach, lateral approach, posterior approach, and transverse approach.
“20. The system according to any of claims 15, wherein the processor assigns the subject to the appropriate surgical procedure and the preferred surgical approach using iterative processing to determine the combination of surgical procedure and surgical approach most likely to result in an optimized outcome for the subject.”
For more information, see this patent application: Zucker, Ido. Surgical Path Planning Using Artificial Intelligence For Feature Detection. Filed
(Our reports deliver fact-based news of research and discoveries from around the world.)
Patent Issued for Built-in legal framework file management (USPTO 11328089): International Business Machines Corporation
Loans in Areas Having Special Flood Hazards; Interagency Questions and Answers Regarding Flood Insurance
Advisor News
Annuity News
Health/Employee Benefits News
Life Insurance News