Methods, systems, and apparatuses for associating dielectric properties with a patient model

Patent metadata

Field	Value
Publication number	US-12471852-B2
Application number	US-202017136841-A
Country	US
Kind code	B2
Filing date	Dec 29, 2020
Priority date	Dec 31, 2019
Publication date	Nov 18, 2025
Grant date	Nov 18, 2025

Abstract

Methods, systems, and apparatuses are described for associating dielectric properties with a patient model.

First claim

What is claimed is:

1 . A method comprising:
receiving, for one patient of a plurality of patients, a plurality of samples of tissues, wherein each sample of tissue of the plurality of samples of tissue is from a different location of the one patient; determining a plurality of sets of image data associated with the plurality of patients, wherein each patient is associated with a set of image data derived from imaging a portion of the patient, wherein each set of image data comprises a plurality of voxels; labeling at least one voxel of the set of image data associated with the one patient of the plurality of patients with dielectric property information based on measured dielectric properties of a sample of tissue of the plurality of samples of tissues of the one patient; determining, based on a first portion of the plurality of sets of image data and the dielectric property information associated with each voxel of the plurality of voxels that correspond to the first portion of the plurality of sets of image data, a plurality of features for a predictive model; training, based on the plurality of features and the first portion of the plurality of sets of image data, the predictive model, wherein the predictive model is configured to determine dielectric property information for each voxel of a plurality of voxels associated with an image; testing, based on a second portion of the plurality of sets of image data, the predictive model; and outputting, based on the testing, the predictive model.

2 . The method of claim 1 ,
wherein each location of the one patient is associated with a different region of the portion of the patient, wherein each location of the one patient corresponds to a voxel of the plurality of voxels of the set of image data associated with the patient, the method further comprising: determining, for each sample of tissue of the plurality of samples of tissue from each patient, based on one or more measured properties of the sample tissue, dielectric properties; labeling, based on determining the dielectric properties for each sample of tissue of the plurality of samples of tissue from each patient, the corresponding voxel of the respective the set of image data with the dielectric property information; and labeling, for each voxel of the respective the set of image data labeled with the dielectric property information, based on connected-component labeling, one or more voxels within the region associated with the voxel with dielectric property information that matches the labeled dielectric property information of the voxel.

3 . The method of claim 1 , wherein the dielectric property information comprises one or more of conductivity information or relative permittivity information.

4 . The method of claim 1 , wherein the plurality of sets of image data associated with the plurality of patients comprise image data associated with one or more of magnetic resonance imaging (MRI), radiography, ultrasound, elastography, photoacoustic imaging, positron emission tomography, echocardiography, magnetic particle imaging, or functional near-infrared spectroscopy.

5 . The method of claim 1 , further comprising:
determining, for a new patient, a new set of image data, wherein the new set of image data comprises a plurality of voxels; presenting, to the predictive model, the new image data set; and determining, by the predictive model, for each voxel of the plurality of voxels of the new set of image data, dielectric property information.

6 . The method of claim 1 , wherein determining the plurality of features for the predictive model comprises determining the plurality of features for the predictive model based on a feature selection technique comprising one or more of a filter method, a wrapper method, or an embedded method.

7 . The method of claim 1 , wherein training the predictive model comprises one or more of discriminant analysis, a decision tree, a statistical algorithm, or a neural network.

8 . A method of using the predictive model of claim 1 .

9 . An apparatus comprising:
one or more processors; and memory storing processor-executable instructions that, when executed by the one or more processors, cause the apparatus to:
receive, for one patient of a plurality of patients, a plurality of samples of tissues, wherein each sample of tissue of the plurality of samples of tissue is from a different location of the one patient;
determine a plurality of sets of image data associated with the plurality of patients, wherein each patient is associated with a set of image data derived from imaging a portion of the patient, wherein each set of image data comprises a plurality of voxels;
label at least one voxel of the set of image data associated with the one patient of the plurality of patients with dielectric property information based on measured dielectric properties of a sample of tissue of the plurality of samples of tissues of the one patient;
determine, based on a first portion of the plurality of sets of image data and the dielectric property information associated with each voxel of the plurality of voxels that correspond to the first portion of the plurality of sets of image data, a plurality of features for a predictive model;
train, based on the plurality of features and the first portion of the plurality of sets of image data, the predictive model, wherein the predictive model is configured to determine dielectric property information for each voxel of a plurality of voxels associated with an image;
test, based on a second portion of the plurality of sets of image data, the predictive model; and
output, based on the testing, the predictive model.

10 . The apparatus of claim 9 ,
wherein each location of the one patient is associated with a different region of the portion of the patient, wherein each location of the one patient corresponds to a voxel of the plurality of voxels of the set of image data associated with the patient, wherein the processor-executable instructions, when executed by the one or more processors, further cause the apparatus to: determine, for each sample of tissue of the plurality of samples of tissue from each patient, based on one or more measured properties of the sample tissue, dielectric properties; label, based on determining the dielectric properties for each sample of tissue of the plurality of samples of tissue from each patient, the corresponding voxel of the respective the set of image data with the dielectric property information; and label, for each voxel of the respective the set of image data labeled with the dielectric property information, based on connected-component labeling, one or more voxels within the region associated with the voxel with dielectric property information that matches the labeled dielectric property information of the voxel.

11 . The apparatus of claim 9 , wherein the dielectric property information comprises one or more of conductivity information or relative permittivity information.

12 . The apparatus of claim 9 , wherein the plurality of sets of image data associated with the plurality of patients comprise image data associated with one or more of magnetic resonance imaging (MRI), radiography, ultrasound, elastography, photoacoustic imaging, positron emission tomography, echocardiography, magnetic particle imaging, or functional near-infrared spectroscopy.

13 . The apparatus of claim 9 , wherein the processor-executable instructions, when executed by the one or more processors, further cause the apparatus to:
determine, for a new patient, a new set of image data, wherein the new set of image data comprises a plurality o

Assignees

• Novocure Gmbh

Inventors

• Shamir Reuven Ruby
• Bomzon Zeev
• Urman Noa

Classifications

• G06T2207/30096

  Tumor; Lesion · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title

• G06T2207/20081

  Training; Learning · CPC title

• G06T2207/20024

  Filtering details · CPC title

• G06T2207/10132

  Ultrasound image · CPC title