Navigation of tubular networks
Inventors
Mintz, David S. • Ghoreyshi, Atiyeh • Jeevan, Prasanth • Xu, Yiliang • Yang, Gehua • Leotta, Matthew Joseph • Stewart, Charles V.
Assignees
Auris Health Inc • Kitware Inc
Publication Number
US-10169875-B2
Publication Date
2019-01-01
Expiration Date
2036-09-16
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Methods and apparatuses provide improved navigation through tubular networks such as lung airways by providing improved estimation of location and orientation information of a medical instrument (e.g., an endoscope) within the tubular network. Various input data such as image data, EM data, and robot data are used by different algorithms to estimate the state of the medical instrument, and the state information is used to locate a specific site within a tubular network and/or to determine navigation information for what positions/orientations the medical instrument should travel through to arrive at the specific site. Probability distributions together with confidence values are generated corresponding to different algorithms are used to determine the medical instrument's estimated state.
Core Innovation
The invention provides methods and apparatuses for improved navigation through tubular networks, such as lung airways, by enabling enhanced estimation of location and orientation information for a medical instrument (e.g., an endoscope) within the tubular network. Various types of input data—such as image data, electromagnetic (EM) data, and robot data—are processed by different algorithms to estimate the instrument's state. This estimated state information is then used to accurately locate a specific site within the network and to generate navigation details for appropriately moving the instrument to the target location.
The background identifies that conventional navigation techniques, even those employing robotic bronchoscopes or 3D anatomical models derived from CT scans, can lead to inaccurate motion estimation of medical devices inside a patient's body. This results in imprecise localization of the device, potentially misleading medical professionals during procedures. The invention addresses the need for improved navigation techniques that provide more reliable and precise real-time state estimation for instruments inside complex branched anatomical structures.
The disclosed system integrates multiple algorithm modules utilizing EM-based, image-based, and robot-based approaches. Estimated states generated by these modules are combined, along with associated probability distributions and confidence values, to improve the accuracy of navigation within the tubular network. The apparatus may include a flexible or rigid endoscopic tool equipped with sensors and controlled by a robotic manipulator, with input data continually processed by a navigation configuration system and processor to deliver real-time movement and location/orientation estimates.
The system accommodates pre-operative planning by generating 3D models from CT scans, enabling automated path planning and virtual navigation to target areas. During procedures, real-time sensor data—including images, EM readings, and robotic movement—is analyzed to update navigation guidance and instrument localization. By using probability distributions and confidence values from different algorithmic modules, the system offers a robust framework for precise and adaptive endoluminal navigation.
Claims Coverage
The independent claims focus on a system and computer readable storage medium for determining the estimated state of an elongated medical instrument within a tubular network using multiple sensor modalities and algorithmic processing.
Combining estimated states from multiple sensor data sources
A system uses a set of processors and program instructions to: - Access sensor data from at least two different sets of sensors (e.g., one set at the proximal portion of the instrument and another within the instrument itself), each providing distinct types of data regarding the manipulation and positioning of an elongated medical instrument. - Determine a first estimated state from data of the first sensors and a second estimated state from data of the second sensors. - Establish an overall estimated state of the medical instrument by combining both estimated states.
Integration of a third sensor modality for instrument state estimation
- The system can access a third set of sensor data from sensors different from the first and second sets (e.g., another data stream providing further positioning information). - The system determines a third estimated state using this data and combines it with the first and second estimated states for an aggregate instrument state estimation.
Utilization of confidence values in combining estimated states
- The system associates each estimated state (from each sensor data set) with a confidence value representing the degree of confidence in each individual estimate. - The final determined instrument state is based on the estimated states and their corresponding confidence values.
Sensor data types: robotic data, image data, and electromagnetic (EM) data
- The system is explicitly configured to access sensor data that comprises: - Robotic data regarding physical manipulation (such as pitch, roll, yaw, and cable retraction or insertion distance), - Image data captured by an imaging device near the instrument tip (potentially as a time series), and/or - EM data recorded by sensors located near the instrument tip and at least one external EM sensor/generator.
Pre-operative model integration and EM registration
- The system can generate a 3D model of the tubular network based on CT scans. - Registration between an EM system's coordinates and the 3D model's coordinates is performed, so EM-derived estimated states align with model-based navigation.
Discrete probability distribution for state representation
- The estimated state (e.g., which branch the instrument tip is in) can be represented as a discrete probability distribution, with each possible branch assigned a likelihood value indicating the chance the tip is in that branch.
Methods for updating estimated state using prior state information
- When determining an estimated state (first and/or second), the system utilizes data about a prior estimated state (from a prior time instant) to inform the current calculation, supporting recursive/continuous state estimation.
Non-transitory computer readable storage medium
- A non-transitory storage medium stores instructions for accessing multiple sensor data types, determining estimated states from each, and combining them for overall instrument tip state estimation, including features such as confidence value association, robotic data access, EM data access, image data access, and probability/discrete representations.
In summary, the inventive features cover a multi-modal, confidence-weighted system and method for determining the estimated state of an elongated medical instrument when navigating tubular anatomical networks, including explicit use of diverse sensor data, probabilistic state representation, confidence values, and integration of pre-operative models and EM registration.
Stated Advantages
Improved estimation of location and orientation information of a medical instrument within tubular networks, resulting in enhanced accuracy for real-time navigation.
Integration of multiple sensor modalities and algorithms provides more robust and reliable navigation compared to approaches relying on a single data source.
The use of probability distributions and confidence values allows dynamic weighting of different sensor data and algorithms, improving localization accuracy and permitting error correction during procedures.
Real-time feedback and display of navigation information facilitate more convenient operations for physicians and enable pre-operative planning and simulation.
Documented Applications
Navigation of medical instruments, such as endoscopes and bronchoscopes, through tubular networks in a patient's body, including lung airways.
Assisting in surgical procedures involving bronchoscopy, including examination, biopsy, and therapeutic operations within lung airways.
Pre-operative planning by generating and displaying 3D models of anatomical tubular networks and simulating navigation paths and tool alignment.
Interested in licensing this patent?