# US 2022/0221373 A1 · Method and device for determining the shape of an optical waveguide, and device for producing training data for a neural network

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0221373 A1
- Document ID: US20220221373A1
- Status: Published application
- Kind: A1
- Application number: 17/694,151
- Filed: 2022-03-14
- Published: 2022-07-14
- Earliest claimed: 2019-09-12
- Assignees: Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.
- Applicants: Fraunhofer-Gesellschaft Zur FöRderung Der Angewandten Forschung E.V.
- Inventors: WOLFGANG SCHADE
- Web page: https://patentgrep.com/patent/US20220221373A1

## CPC

- G01M 11/088: {of optical fibres; Mechanical features associated with the optical testing of optical fibres}
- A61B 34/20: Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
- G01B 11/18: {using photoelastic elements}
- G01L 1/245: {using microbending}
- G01L 1/246: {using integrated gratings, e.g. Bragg gratings}
- G01M 11/31: {with a light emitter and a light receiver being disposed at the same side of a fibre or waveguide end-face, e.g. reflectometers}
- G06N 3/0499: Feedforward networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- A61B 2034/2061: {using shape-sensors, e.g. fiber shape sensors with Bragg gratings}
- G01M: TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G01B: MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01L: MEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE (weighing G01G)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A method for determining the shape of an optical waveguide (1) having a plurality of fiber Bragg gratings (15) includes the following steps: coupling light (20) of a light source (2) into the optical waveguide (1), coupling the reflected light out of the optical waveguide (1), determining a spectrum (35) of the reflected light by measuring the intensity (I) versus the wavelength (1), the spectrum (35) being fed to a self-learning neural network (4) and the shape of the optical waveguide (1) being determined by the neural network. A device for determining the shape of an optical waveguide (1) may be used in a catheter or an endoscope or a biopsy needle or an aerodynamic profiled element. A device for producing training data for a neural network to implement the above is also contemplated.

## Claims

### Claim 1

A method for determining the shape of an optical waveguide having a plurality of fiber Bragg gratings, comprising:
coupling light of a light source into the optical waveguide,
coupling the reflected light out of the optical waveguide,
determining spectrum data of the reflected light by measuring the intensity versus the wavelength,
providing the spectrum data to a self-learning neural network having been preconditioned by training data, and
determining the shape of the optical waveguide by means of said self-learning neural network.

### Claim 2 (depends on 1)

The method according to claim 1, comprising:
preconditioning said self-learning neural network with training data comprising a plurality of data sets, each data set comprising measured spectrum data and an associated shape of the optical waveguide.

### Claim 3 (depends on 2)

The method according to claim 2, wherein:
the training data contains more than 600 data sets.

### Claim 4 (depends on 2)

The method according to claim 2, wherein preconditioning said self-learning neural network comprises generating the training data by repeatedly:
deforming the optical waveguide into a deformed shape,
detecting the deformed shape of the optical waveguide, and
measuring spectrum data associated with the deformed shape.

### Claim 5 (depends on 4)

The method according to claim 4, comprising:
deforming the optical waveguide with a robot arm.

### Claim 6 (depends on 5)

The method according to claim 5, wherein:
the shape of the optical waveguide is detected from any position of the robot arm, by electronic image processing of at least one image of the optical waveguide.

### Claim 7 (depends on 1)

The method according to claim 1, comprising:
attaching the optical waveguide to one from the group consisting of: a catheter, an endoscope, a biopsy needle and a deformable aerodynamic structure, prior to determining the shape of said optical waveguide.

### Claim 8 (depends on 1)

The method according to claim 1, wherein the shape of said optical waveguide comprises any of its curvature and its torsion.

### Claim 9

A device for determining the shape of an optical waveguide, comprising
at least one optical waveguide having a plurality of fiber Bragg gratings,
at least one light source adapted to couple light into the optical waveguide,
at least one spectrometer adapted to determine spectrum data of the light reflected in the optical waveguide by measuring the intensity versus the wavelength of the reflected light,
wherein
the device further comprises a self-learning neural network having been preconditioned by training data and having at least one input and at least one output, the input being configured to receive said spectrum data and the output being configured to provide data representing the shape of the optical waveguide.

### Claim 10 (depends on 9)

The device according to claim 9, wherein:
said self-learning neural network has been preconditioned with training data comprising of a plurality of data sets, each data set comprising measured spectrum data and an associated shape of the optical waveguide.

### Claim 11

A device for generating training data for a neural network, comprising
at least one optical waveguide having a plurality of fiber Bragg gratings,
at least one light source configured to couple light into the optical waveguide,
at least one spectrometer configured to determine spectrum data of the light reflected in the optical waveguide by measuring the intensity versus the wavelength of the reflected light,
an apparatus configured to deform the optical waveguide,
an apparatus configured to detect the shape of the optical waveguide, and
a memory configured to store data representing the shape of the optical waveguide and the spectrum.

### Claim 12 (depends on 11)

The device according to claim 11, wherein:
the apparatus configured to deform the optical waveguide comprises a robot arm.

### Claim 13 (depends on 11)

The device according to claim 11, wherein:
the apparatus configured to detect the shape of the optical waveguide is configured to detect a position of the robot arm.

### Claim 14 (depends on 11)

The device according to claim 11, wherein:
the apparatus configured to detect the shape of the optical waveguide comprises a camera configured to produce a data stream representing an image of the optical waveguide.
