Intelligent boundary delineation of regions of interest of an organism from multispectral video streams using perfusion models

Patent metadata

Field	Value
Publication number	US-11276176-B2
Application number	US-201916560734-A
Country	US
Kind code	B2
Filing date	Sep 4, 2019
Priority date	Sep 4, 2019
Publication date	Mar 15, 2022
Grant date	Mar 15, 2022

Abstract

Embodiments for implementing intelligent boundary delineation of a region of interest of an organism in two spatial dimensions in a computing environment by a processor. Time series data of a contrast agent in one or more regions of interest captured from multispectral image streams may be collected. One or more regions of interest having one or more perfusion patterns may be identified from the time series data. Boundaries of the one or more regions of interest may be delineated into at least two spatial dimensions, wherein the boundaries of the one or more regions of interest include one or more selected labels.

First claim

The invention claimed is:

1. A method for implementing intelligent classification of regions of interest of an organism by a processor, comprising:
collecting time series data of a contrast agent in one or more regions of interest from multispectral image streams;
identifying the one or more regions of interest having one or more perfusion patterns from the time series data; and
delineating boundaries of the one or more regions of interest into at least two spatial dimensions, wherein the boundaries of the one or more regions of interest include one or more selected labels; and
generating a description of advection and diffusion fields of one or more characteristics and performances of the contrast agent in the two spatial dimensions of the boundaries of the one or more regions of interest, wherein advection and diffusion coefficients of bio-physical models of dye transport and fluorescence in tissue of the contrast agent in the one or more regions of interest are estimated.

2. The method of claim 1 , further including classifying the one or more perfusion patterns into one of a plurality of classes by applying one or more perfusion models representing spatio-temporal behavior of the contrast agent reflected by the time series data and by using a machine learning operation.

3. The method of claim 1 , further including:
receiving, in real-time, the multispectral image streams of the one or more regions of interest for a selected period of time from an image capturing device;
collecting a corpus of labels for labeling the multispectral image streams;
labeling boundaries of the one or more perfusion patterns in the multispectral image streams; or
labeling one of a plurality of classes of one or more regions of interest in the multispectral image streams, wherein the one of the plurality of classes represents at least a predicted medical diagnosis of the one or more regions of interest.

4. The method of claim 1 , further including assigning a confidence score to the one or more selected labels.

5. The method of claim 1 , further including:
identifying a spatio-temporal behavior of the contrast agent captured from the multispectral image streams, wherein the spatio-temporal behavior includes a fluorescence intensity level and the fluorescence intensity level represents a concentration level of the contrast agent in the one or more regions of interest.

6. The method of claim 1 , further including initiating a machine learning model to perform one or more machine learning operations to train or retrain the one or more machine learning models according to a repository of plurality of multispectral image streams, a corpus of classes or labels of each of the plurality of multispectral image streams, a plurality of time series data, labeled perfusion patters, labeled regions of interest, labeled boundaries associated with the one or more regions of interest, patient profile data, or a combination thereof.

7. A system for implementing intelligent classification of region of interest of an organism, comprising:
one or more computers with executable instructions that when executed cause the system to:
collect time series data of a contrast agent in one or more regions of interest from multispectral image streams;
identify the one or more regions of interest having one or more perfusion patterns from the time series data;
delineate boundaries of the one or more regions of interest into at least two spatial dimensions, wherein the boundaries of the one or more regions of interest include one or more selected labels; and
generate a description of advection and diffusion fields of one or more characteristics and performances of the contrast agent in the two spatial dimensions of the boundaries of the one or more regions of interest, wherein advection and diffusion coefficients of bio-physical models of dye transport and fluorescence in tissue of the contrast agent in the one or more regions of interest are estimated.

8. The system of claim 7 , wherein the executable instructions further classify the one or more perfusion patterns into one of a plurality of classes by applying one or more perfusion models representing spatio-temporal behavior of the contrast agent reflected by the time series data and by using a machine learning operation.

9. The system of claim 7 , wherein the executable instructions further:
receive, in real-time, the multispectral image streams of the one or more regions of interest for a selected period of time from an image capturing device;
collect a corpus of labels for labeling the multispectral image streams;
label boundaries of the one or more perfusion patterns in the multispectral image streams; or
label one of a plurality of classes of the one or more regions of interest in the multispectral image streams, wherein the one of the plurality of classes represents at least a predicted medical diagnosis of the one or more regions of interest.

10. The system of claim 7 , wherein the executable instructions further assign a confidence score to the one of the plurality of classes of the one or more perfusion patterns.

11. The system of claim 7 , wherein the executable instructions further identify a spatio-temporal behavior of the contrast agent captured from the multispectral image streams, wherein the spatio-temporal behavior includes a fluorescence intensity level and the fluorescence intensity level represents a concentration level of the contrast agent in the one or more regions of interest.

12. The system of claim 7 , wherein the executable instructions further initiating a machine learning model to perform one or more machine learning operations to train or retrain the one or more machine learning models according to a repository of plurality of multispectral image streams, a corpus of classes or labels of each of the plurality of multispectral image streams, a plurality of time series data, labeled perfusion patters, labeled regions of interest, labeled boundaries associated with the one or more regions of interest, patient profile data, or a combination thereof.

13. A computer program product for implementing intelligent classification of region of interest of an organism by a processor, the computer program product comprising a non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising:
an executable portion that collects time series data of a contrast agent in one or more regions of interest from multispectral image streams;
an executable portion that identifies the one or more regions of interest having one or more perfusion patterns from the time series data; and
an executable portion that delineates boundaries of the one or more regions of interest into at least two spatial dimensions, wherein the boundaries of the one or more regions of interest include one or more selected labels; and
an executable portion that generates a description of advection and diffusion fields of one or more characteristics and performances of the contrast agent in the two spatial dimensions of the boundaries of the one or more regions of interest, wherein advection and diffusion coefficients of bio-physical models of dye transport and fluorescence in tissue of the contrast agent in the one or more regions of interest are estimated.

14. The computer program product of claim 13 , further including an executable portion that classifies the one or more perfusion patterns into one of a plurality of classes by applying one or more perfusion models representing spatio-temporal behavior of the contrast agent reflected by the ti

Assignees

• IBM

Inventors

• Zhuk Sergiy
• Epperlein Jonathan
• Mac Aonghusa Pol
• Nair Rahul

Classifications

• G06V10/255

  Detecting or recognising potential candidate objects based on visual cues, e.g. shapes · CPC title

• G06V10/143

  Sensing or illuminating at different wavelengths · CPC title

• G06T7/13 Primary

  Edge detection · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06F18/217

  Validation; Performance evaluation; Active pattern learning techniques · CPC title