# US 2022/0277456 A1 · Method and apparatus for analysing intracoronary images

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

- Number: US 2022/0277456 A1
- Document ID: US20220277456A1
- Status: Published application
- Kind: A1
- Application number: 17/606,996
- Filed: 2020-04-30
- Published: 2022-09-01
- Earliest claimed: 2019-05-01
- Assignees: Cambridge Enterprise Limited
- Applicants: Cambridge Enterprise Limited
- Inventors: Adam WOOLF; Martin BENNETT
- Web page: https://patentgrep.com/patent/US20220277456A1

## CPC

- G06T 7/0012: {Biomedical image inspection}
- G06T 7/0016: {involving temporal comparison}
- A61B 5/0066: {Optical coherence imaging}
- G06V 10/40: Extraction of image or video features
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/766: using regression, e.g. by projecting features on hyperplanes
- G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
- G06V 10/82: using neural networks
- G16H 30/20: for handling medical images, e.g. DICOM, HL7 or PACS
- A61B 2576/00: Medical imaging apparatus involving image processing or analysis (A61B1/00009, A61B6/52 and A61B8/52 take precedence)
- G06T 2207/10101: Optical tomography; Optical coherence tomography [OCT]
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T 2207/30048: Heart; Cardiac
- G06T 2207/30101: Blood vessel; Artery; Vein; Vascular
- G06T 2207/30168: Image quality inspection
- G06V 2201/03: Recognition of patterns in medical or anatomical images
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Embodiments of the present techniques provide apparatus and methods for analysing intracoronary images, for example to predict the likelihood of a disease, disease presentation or event, and/or to track performance of a drug or other treatment. The method may comprise: for each image in the set of images of a coronary artery: classifying the image, using a first neural network, for the presence or absence of diseased tissue; when the image is classified as having diseased tissue present, classifying the image, using a second neural network, for the presence or absence of an artefact; determining whether to analyse the image based on the classifying steps; when the image is to be analysed, analysing the image by identifying, using a third neural network, one or more features of interest in a coronary artery tissue; and measuring each identified feature of interest.

## Claims

### Claim 1

An automated computer-implemented method for analysing a set of images of a coronary artery, the method comprising:
for each image in the set of images:
classifying the image, using a first neural network, for the presence or absence of diseased tissue;
when the image is classified as having diseased tissue present, classifying the image, using a second neural network, for the presence or absence of an artefact;
determining whether to analyse the image based on the classifying steps;
when the image is to be analysed, analysing the image by identifying, using a third neural network, one or more features of interest in a coronary artery tissue; and
measuring each identified feature of interest.

### Claim 2 (depends on 1)

The method as claimed in claim 1, further comprising, prior to classifying the image using the second neural network, sampling the coronary artery in the image into a plurality of samples and inputting a set of samples into the second neural network.

### Claim 3 (depends on 2)

The method as claimed in claim 2, wherein classifying the image comprises:
determining a proportion of coronary artery tissue which is detectable in each sample in the input set; and
classifying the image based on the determined proportion.

### Claim 4 (depends on 2)

The method as claimed in claim 2, further comprising:
training the second neural network to classify the image by analysing a plurality of training dataset samples.

### Claim 5 (depends on 2)

The method as claimed in claim 2, wherein when an image is classified as having an artefact present, further classifying the artefact as correctable or non-correctable.

### Claim 6 (depends on 5)

The method as claimed in claim 5, wherein classifying the image comprises:
determining a proportion of coronary artery tissue which is detectable in each sample in the input set; and
classifying the image based on the determined proportion, and
wherein the image is classified as having a correctable artefact when the determined proportion is over 50%.

### Claim 7 (depends on 5)

The method as claimed in claim 5,
comprising, prior to analysing the image, correcting the image which has been classified as having a correctable artefact, preferably wherein correcting an image classified as having a correctable artefact comprises applying a variational-autoencoder generative adversarial network to the identified image to recover detail of the coronary artery tissue underneath the identified artefact.

### Claim 8

(canceled)

### Claim 9 (depends on 1)

The method as claimed in claim 1, wherein when the image is classified as having an absence of diseased tissue, replacing the image with blank information, and/or
wherein measuring each identified feature of interest comprises measuring any one or more of: fibrous tissue thickness, fibrous tissue intensity, plaque composition, lumen area, and diseased lumen circumference.

### Claim 10

(canceled)

### Claim 11 (depends on 1)

The method as claimed in claim 1,
further comprising, prior to analysing the image, sampling the coronary artery in the retained image into a plurality of samples; and arranging the samples in a linear representation of the coronary artery,
preferably wherein analysing the image further comprises: identifying, using a regression bounding box technique, an interface between fibrous tissue and necrotic or calcific tissue in each of the plurality of samples, and
further preferably wherein measuring the identified feature of interest comprises measuring the distance between the interface and an edge of the sample.

### Claim at position 12

12-13. (canceled)

### Claim 14 (depends on 1)

The method as claimed in claim 1, wherein the set of images of coronary arteries are optical coherence tomography images.

### Claim 15 (depends on 1)

The method as claimed in claim 1 wherein each image is of a coronary artery of a patient, and the method further comprises:
determining, using the measurement of each identified feature of interest, the likelihood of the patient having presenting with a particular manifestation of coronary artery disease.

### Claim 16 (depends on 15)

The method as claimed in claim 15, determining the likelihood of the patient presenting with a particular manifestation of coronary artery disease comprises using a fourth neural network.

### Claim 17 (depends on 16)

The method as claimed in claim 16, the fourth neural network receives as inputs the measurement of each identified feature of interest for each analysed image and blank information for each image which is not analysed.

### Claim 18 (depends on 15)

The method as claimed in claim 15, wherein the coronary artery disease is myocardial infarction and the identified feature of interest is fibrous tissue thickness.

### Claim 19 (depends on 1)

The method as claimed in claim 1, wherein the set of images comprises a first subset of images of a coronary artery of a patient captured at a first time and a second subset of images of the coronary artery of the patient at a second, subsequent time, and the method further comprises:
measuring a first set of measurements for each identified feature of interest for the first subset of images;
measuring a second set of measurements for each identified feature of interest for the second subset of images; and
determining, using the first and second sets of measurements, any change in the coronary artery.

### Claim 20 (depends on 19)

The method as claimed in claim 19, wherein the first time is a first stage in treatment and the second time is a second stage in the treatment and further comprising determining the efficacy of the treatment based on any determined change.

### Claim 21

A method of diagnosis of a patient comprising:
receiving a plurality of images of a coronary artery of the patient;
for each of the received images:
classifying the image, using a first neural network, for the presence or absence of diseased tissue;
when the image is classified as having diseased tissue present, classifying the image, using a second neural network, for the presence or absence of an artefact;
determining whether to analyse the image based on the classifying; and when the image is to analysed:
analysing the processed image, using a third neural network, by identifying an interface between fibrous tissue and necrotic or calcific tissue; and
measuring a fibrous tissue thickness; and
diagnosing the patient, using a fourth neural network, based on the measured fibrous tissue thickness.

### Claim 22 (depends on 1)

A non-transitory data carrier carrying code which, when implemented on a processor, causes the processor to carry out the method of claim 1.

### Claim 23

An apparatus for analysing a set of images of a coronary artery, the apparatus comprising:
an imaging device for capturing a set of images of a coronary artery; and
at least one processor, coupled to memory, arranged to:
for each image in the set of images:
classify the image, using a first neural network, for the presence or absence of diseased tissue;
when the image is classified as having diseased tissue present, classifying the image, using a second neural network, for the presence or absence of an artefact;
determine whether to analyse the image based on the classifying; and when the image is analysed:
analyse the retained image, using a third neural network, by identifying one or more features of interest in a coronary artery tissue; and measuring each identified feature of interest.

### Claim 24 (depends on 23)

The apparatus as claimed in claim 23, wherein the imaging device is an optical coherence tomography device.
