Adaptive instrument kinematic model optimization for optical shape sensed instruments

Patent metadata

Field	Value
Publication number	US-11067387-B2
Application number	US-201515117496-A
Country	US
Kind code	B2
Filing date	Feb 9, 2015
Priority date	Feb 28, 2014
Publication date	Jul 20, 2021
Grant date	Jul 20, 2021

Abstract

A shape sensing system comprises a processor coupled to a memory storage device and a predictive model module stored in the memory storage device and configured to receive shape sensing measurements and predict next shape sensing measurements based upon current state data for a shape sensing enabled device, the predictive model module including a Kalman filter and at least one reference model employed by the Kalman filter. The Kalman filter employs the reference model(s) and the current state data to predict new shape data by weighting the reference model(s) versus the current state data to improve shape data measurements.

First claim

The invention claimed is:

1. A shape sensing system, comprising:
an elongated flexible medical instrument configured for insertion into an internal region of a subject, the medical instrument including one of a catheter, a guidewire, a probe, an endoscope, a robot or a balloon device, the medical instrument being outfitted with an optical fiber shape sensing device configured to output shape sensing measurements indicative of measured 3D shapes of the elongated flexible medical instrument as the elongated flexible medical instrument flexes and changes shape during insertion, the optical fiber shape sensing device including a shape sensing multi-core fiber with a central core and at least three outer cores helically wound around the central core;
a display configured to display an image of the internal region of the subject;
a processor coupled to a memory storage device;
a predictive model module stored in the memory storage device and configured to receive the measured 3D shapes of the elongated flexible medical instrument during insertion and predicted 3D shapes of the elongated flexible medical instrument based upon a current measured 3D shape of the elongated flexible medical instrument, the predictive model module including:
a Kalman filter;
at least one kinematic model including mass-spring-damper model of the elongated flexible medical instrument employing mechanical and material parameters of the elongated flexible medical instrument which provides the predicted 3D shapes of the elongated flexible medical instrument to the Kalman filter, the Kalman filter weightingly combines the current measured 3D shape of the elongated flexible medical instrument and one or more of the predicted 3D shapes to generate an improved 3D shape of the elongated flexible medical instrument;
a comparison module configured to generate confidence scores for the measured 3D shape of the elongated flexible medical instrument, the confidence scores being used to adjust the weighting with which the measured 3D shape and the predicted next 3D shapes are combined; and
wherein the processor is configured to control the display to display the image of the internal region overlaid with the improved 3D shape of the elongated flexible medical instrument.

2. The system as recited in claim 1 , wherein the mass-spring-damper model is based on stiffness, material density, and geometry of the elongated flexible medical instrument.

3. The system as recited in claim 1 , wherein the at least one kinematic model also includes a device-specific noise model for a given shape sensing enabled device indicative of noise which affects an accuracy of the measured 3D shapes.

4. The system as recited in claim 1 , further comprising a model parameter adaptation module configured to update the at least one kinematic model in accordance with shapes of the elongated flexible medical instrument from another source.

5. The system as recited in claim 1 , wherein the at least one kinematic model includes 3D shape data stored in a library of 3D shapes from previously acquired training cases including different phases of interventional procedures.

6. The system as recited in claim 1 , wherein the predictive model further includes a dynamic model of physical laws of motion to estimate spatio-temporally-varying quantities.

7. A method for determining a 3D shape of an elongated, flexible shape sensing enabled interventional medical device, the method comprising during performance of an interventional procedure on a subject using the elongated, flexible shape sensing enabled interventional medical device:
(a) measuring the 3D shape of the elongated, flexible shape sensing enabled interventional medical device along its length and determining a confidence score regarding an accuracy of the measured 3D shape,
(b) obtaining a dynamic model for at least one modeled 3D shape of the elongated, flexible shape sensing enabled interventional medical device wherein the dynamic model models shape based on mechanical and material parameters of the interventional medical device, and a confidence score for each modeled 3D shape;
(c) inputting the measured 3D shape and the at least one modeled 3D shape into a Kalman filter;
(d) computing a weighted average in accordance with the confidence score of the measured 3D shape and the at least one predicted 3D shape to predict new state data of an improved 3D shape of the elongated, flexible shape sensing enabled interventional medical device;
(e) overlaying a rendering of the improved measured 3D shape of the elongated, flexible shape sensing enabled interventional medical device on an internal image of the subject; and
(f) repeating steps (a)-(e) as the elongated, flexible shape sensing enabled interventional medical device moves longitudinally in the subject,
wherein computing the weighted average includes additional dynamic models, and
wherein the weighting is adjusted in accordance with the confidence score.

8. The method as recited in claim 7 , further including:
(g) receiving another source 3D shape of the elongated, flexible shape sensing enabled interventional medical device from another source; and
(h) adjusting the dynamic model data and the confidence value based on the another source 3D shape.

9. The method as recited in claim 7 , wherein the dynamic model is based on at least stiffness, material density, and geometry of the interventional medical device.

10. The method as recited in claim 7 , wherein the dynamic model includes a mass-spring-damper model.

11. A shape sensing system comprising:
a memory storing dynamic model data for a plurality of elongated, flexible shape sensing enabled interventional devices configured for being inserted in an internal region of a subject, the dynamic model data being based on physical and mechanical properties of the elongated, flexible shape sensing enabled interventional devices;
a selected one of the elongated, flexible shape sensing enabled interventional medical devices;
a display device; and
one or more processors configured to:
(a) measure the 3D shape of the selected elongated, flexible shape sensing enabled interventional medical device along its length and determining a confidence score regarding an accuracy of the measured 3D shape;
(b) retrieve the dynamic model data for the selected elongated, flexible shape sensing enabled interventional medical device and modeling the 3D shape for the selected elongated, flexible shape sensing enabled interventional medical device;
(c) input the measured 3D shape and the at least one modeled 3D shape of the selected elongated, flexible shape sensing enabled interventional medical device into a Kalman filter;
(d) compute a weighted average in accordance with the confidence score of the measured 3D shape and the modeled 3D shape to predict new state data of an improved 3D shape of the selected elongated, flexible shape sensing enabled interventional medical device;
(e) overlay a rendering of the improved measured 3D shape of the elongated, flexible shape sensing enabled interventional medical device on an internal image of the subject; and
(f) repeat steps (a)-(e) during insertion of the elongated, flexible shape sensing enabled interventional medical device in the internal region of the subject,
wherein the memory further stores additional dynamic models and the one or more processors are further configured to compute the weighted average including the additional dynamic models, and
wherein the one or more processors are further configured to adjust the weighting in accordance with the confidence score.

12. The system as recited in claim 11 , wherein the dynamic model data is ba

Assignees

• Koninklijke Philips Nv

Inventors

• Marell Milan Jan Henri
• Van Belzen Femke
• Chan Raymond

Classifications

• G01B11/18 Primary

  using photoelastic elements · CPC title

• G01B21/045

  Correction of measurements (G01B9/02055 takes precedence) · CPC title

• A61B34/20

  Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis · CPC title

• A61B2034/102

  Modelling of surgical devices, implants or prosthesis · CPC title

• A61B2034/2061

  using shape-sensors, e.g. fiber shape sensors with Bragg gratings · CPC title