GPU accelerated perfusion estimation from multispectral videos

Patent metadata

Field	Value
Publication number	US-11526703-B2
Application number	US-202016940525-A
Country	US
Kind code	B2
Filing date	Jul 28, 2020
Priority date	Jul 28, 2020
Publication date	Dec 13, 2022
Grant date	Dec 13, 2022

Abstract

In an approach for classifying regions of tissue captured in multispectral videos into medically meaningful classes using GPU accelerated perfusion estimation, a processor receives one or more multispectral videos of a subject tissue of a patient. A processor extracts one or more fluorescence time series profiles from the one or more multispectral videos. A processor estimates one or more sets of perfusion parameters based on the one or more fluorescence time series profiles. A processor inputs one or more feature vectors into a classifier, wherein the one or more feature vectors are derived the one or more sets of perfusion parameters. A processor receives a classification result for each of the one or more feature vectors, wherein the classification result comprises a set of medically relevant labels for each of the one or more feature vectors with a level of certainty for each label of the set of medically relevant labels.

First claim

What is claimed is:

1. A computer-implemented method comprising:
receiving, by one or more processors, a set of inputs, wherein the set of inputs includes one or more multispectral videos of a subject tissue of a patient;
extracting, by the one or more processors, one or more fluorescence time series profiles from the one or more multispectral videos;
estimating, by the one or more processors, one or more sets of perfusion parameters for advection-diffusion equation based on the one or more fluorescence time series profiles using a scalar transport equation, ∂ t u+∂ x i (A i u)=∂ x i (D∂ x i u)+S, wherein a spatiotemporal fluorescence intensity u(x,t) of the one or more fluorescence time series profiles can be modelled in terms of a spatially varying advective velocity field A(x) and diffusive scalar field D(x), wherein the advective velocity field and the diffusive scalar field are perfusion parameters of the one or more sets of perfusion parameters;
inputting, by the one or more processors, one or more feature vectors into a classifier, wherein the one or more feature vectors are derived the one or more sets of perfusion parameters; and
receiving, by the one or more processors, a classification result output by the classifier for each of the one or more feature vectors, wherein the classification result comprises a set of medically relevant labels for each of the one or more feature vectors with a level of certainty for each label of the set of medically relevant labels.

2. The computer-implemented method of claim 1 , wherein the one or more multispectral videos is a live stream of a multispectral video of the subject tissue directly from a medical imaging device.

3. The computer-implemented method of claim 1 , wherein the set of inputs further includes patient metadata for the patient.

4. The computer-implemented method of claim 3 , wherein each of the one or more feature vectors are derived from one of the one or more sets of perfusion parameters and the patient metadata.

5. The computer-implemented method of claim 1 , wherein extracting the one or more fluorescence time series profiles from the one or more multispectral videos comprises:
extracting, by the one or more processors, the one or more fluorescence time series profiles in a coordinate system fixed to the patient based on the received one or more multispectral videos of the subject tissue, wherein the coordinate system has time along an x-axis and a fluorescence intensity of the subject tissue along a y-axis.

6. The computer-implemented method of claim 1 , wherein extracting the one or more fluorescence time series profiles from the one or more multispectral videos comprises:
extracting, by the one or more processors, a fluorescence time series profile for each region of the subject tissue.

7. The computer-implemented method of claim 1 , wherein estimating the one or more sets of perfusion parameters based on the one or more fluorescence time series profiles comprises:
estimating, by the one or more processors, a set of values for each of the one or more sets of perfusion parameters for each fluorescence time series profile of the one or more fluorescence time series profiles.

8. The computer-implemented method of claim 1 , wherein estimating the one or more sets of perfusion parameters based on the one or more fluorescence time series profiles comprises:
estimating, by the one or more processors, the one or more sets of perfusion parameters for a pre-defined partial differential equation based on the one or more fluorescence time series profiles.

9. The computer-implemented method of claim 1 , further comprising:
solving, by the one or more processors, a boundary value problem (BVP) system to produce the set of values for each of the one or more sets of perfusion parameters, wherein the BVP system is partitioned to one or more graphics processing units (GPUs) to execute portions of the BVP system.

10. A computer program product comprising:
one or more computer readable tangible storage device and program instructions stored on the one or more computer readable tangible storage device, the program instructions comprising:
program instructions to receive a set of inputs, wherein the set of inputs includes one or more multispectral videos of a subject tissue of a patient;
program instructions to extract one or more fluorescence time series profiles from the one or more multispectral videos;
program instructions to estimate one or more sets of perfusion parameters for advection-diffusion equation based on the one or more fluorescence time series profiles using a scalar transport equation, using a scalar transport equation, ∂ t u+∂ x i (A i u)=∂ x i (D∂ x i u)+S, wherein a spatiotemporal fluorescence intensity u(x,t) of the one or more fluorescence time series profiles can be modelled in terms of a spatially varying advective velocity field A(x) and diffusive scalar field D(x), wherein the advective velocity field and the diffusive scalar field are perfusion parameters of the one or more sets of perfusion parameters;
program instructions to input one or more feature vectors into a classifier, wherein the one or more feature vectors are derived the one or more sets of perfusion parameters; and
program instructions to receive a classification result output by the classifier for each of the one or more feature vectors, wherein the classification result comprises a set of medically relevant labels for each of the one or more feature vectors with a level of certainty for each label of the set of medically relevant labels.

11. The computer program product of claim 10 , wherein the program instructions to extract the one or more fluorescence time series profiles from the one or more multispectral videos comprise:
program instructions to extract the one or more fluorescence time series profiles in a coordinate system fixed to the patient based on the received one or more multispectral videos of the subject tissue, wherein the coordinate system has time along an x-axis and a fluorescence intensity of the subject tissue along a y-axis.

12. The computer program product of claim 10 , wherein the program instructions to estimate the one or more sets of perfusion parameters based on the one or more fluorescence time series profiles comprise:
program instructions to estimate the one or more sets of perfusion parameters for a pre-defined partial differential equation based on the one or more fluorescence time series profiles.

13. The computer program product of claim 10 , further comprising:
program instructions to solve a boundary value problem (BVP) system to produce the set of values for each of the one or more sets of perfusion parameters, wherein the BVP system is partitioned to one or more graphics processing units (GPUs) to execute portions of the BVP system.

14. A computer system comprising:
one or more computer processors;
one or more computer readable storage media;
program instructions stored on the computer readable storage media for execution by at least one of the one or more processors, the program instructions comprising:
program instructions to program instructions to receive a set of inputs, wherein the set of inputs includes one or more multispectral videos of a subject tissue of a patient;
program instructions to extract one or more fluorescence time series profiles from the one or more multispectral videos;
program instructions to estimate one or more sets of perfusion parameters for advection-diffusion equation based on the one or more fluorescence time series profiles using a scalar transport equation, ∂ t u+

Assignees

• IBM

Inventors

• Moore Stephen Michael
• Zhuk Sergiy
• Tirupathi Seshu
• Gazzetti Michele
• Macaonghusa Pol

Classifications

• G06F18/24 Primary

  Classification techniques · CPC title

• G06T7/0016

  involving temporal comparison · CPC title

• G06V10/58 Primary

  relating to hyperspectral data · CPC title

• G06T2207/30028

  Colon; Small intestine · CPC title

• G06T2207/20081

  Training; Learning · CPC title