# US 2020/0402239 A1 · Systems and methods for determining blood vessel conditions

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

- Number: US 2020/0402239 A1
- Document ID: US20200402239A1
- Status: Published application
- Kind: A1
- Application number: 17/015,070
- Filed: 2020-09-08
- Published: 2020-12-24
- Earliest claimed: 2018-05-25
- Assignees: Shenzhen Keya Medical Technology Corporation
- Applicants: Shenzhen Keya Medical Technology Corporation
- Inventors: Xin Wang; Youbing Yin; Kunlin Cao; Yuwei Li; Junjie Bai; Xiaoyang Xu
- Web page: https://patentgrep.com/patent/US20200402239A1

## CPC

- G06T 7/0014: {using an image reference approach}
- A61B 6/03: Computed tomography [CT]
- A61B 6/504: {for diagnosis of blood vessels, e.g. by angiography}
- A61B 6/507: {for determination of haemodynamic parameters, e.g. perfusion CT}
- A61B 6/5205: {involving processing of raw data to produce diagnostic data}
- A61B 6/5217: {extracting a diagnostic or physiological parameter from medical diagnostic data}
- G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade}
- G06F 18/295: {Markov models or related models, e.g. semi-Markov models; Markov random fields; Networks embedding Markov models}
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06T 17/00: Three-dimensional [3D] modelling for computer graphics
- G06V 10/454: {Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]}
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/82: using neural networks
- G06V 10/85: {Markov-related models; Markov random fields}
- G16H 50/30: for calculating health indices; for individual health risk assessment
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T 2207/30104: Vascular flow; Blood flow; Perfusion
- G06V 40/14: Vascular patterns
- 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)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- 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

The disclosure relates to systems and methods for evaluating a blood vessel. The method includes receiving image data of the blood vessel acquired by an image acquisition device, and predicting, by a processor, blood vessel condition parameters of the blood vessel by applying a deep learning model to the acquired image data of the blood vessel. The deep learning model maps a sequence of image patches on the blood vessel to blood vessel condition parameters on the blood vessel, where in the mapping the entire sequence of image patches contribute to the blood vessel condition parameters. The method further includes providing the blood vessel condition parameters of the blood vessel for evaluating the blood vessel.

## Claims

### Claim 1

A method for evaluating a blood vessel, comprising:
receiving image data of the blood vessel acquired by an image acquisition device;
predicting, by a processor, blood vessel condition parameters of the blood vessel by applying a deep learning model to the acquired image data of the blood vessel, wherein the deep learning model maps a sequence of image patches on the blood vessel to blood vessel condition parameters of the blood vessel, wherein in the mapping the entire sequence of image patches contribute to the blood vessel condition parameters; and
providing the blood vessel condition parameters of the blood vessel for evaluating the blood vessel.

### Claim 2 (depends on 1)

The method of claim 1, wherein the deep learning model includes a convolutional neural network, a recursive neural network, and a conditional random field model connected in series, wherein parameters of the convolutional neural network, the recursive neural network, and the conditional random field network are jointly optimized.

### Claim 3 (depends on 2)

The method of claim 2, wherein the parameters include at least one of a number of network layers in the convolutional neural network or a number of nodes of each network layer.

### Claim 4 (depends on 2)

The method of claim 2, wherein the parameters include at least one of a number of hidden layers in the recursive neural network, a learning rate, or an initial value.

### Claim 5 (depends on 1)

The method of claim 1, wherein the deep learning model includes a plurality of convolutional neural networks connected in parallel, wherein each convolutional neural network receives a different sequence of image patches derived from the image data of the blood vessel, wherein the deep learning model further includes a connection layer that connects outputs of the plurality of convolutional neural networks.

### Claim 6 (depends on 5)

The method of claim 5, wherein the plurality of convolutional neural networks include a first convolutional neural network receiving a sequence of 2D image patches and a second convolutional neural network receiving a sequence of 3D image patches.

### Claim 7 (depends on 5)

The method of claim 5, wherein the plurality of convolutional neural networks include a first convolutional neural network receiving a sequence of image patches corresponding to a first projection angle on the blood vessel and a second convolutional neural network receiving a sequence of image patches corresponding to a second projection angle on the blood vessel.

### Claim 8 (depends on 1)

The method of claim 1, wherein the deep learning model comprises a forward long short-term memory network and a backward long short-term memory network connected in parallel.

### Claim 9 (depends on 1)

The method of claim 1, wherein the deep learning model is trained using training images of blood vessels and corresponding blood vessel condition parameters obtained by applying computer simulations to the training images or by performing measurements to the blood vessels.

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
reconstructing a geometric model of the blood vessel using the acquired image data of the blood vessel;
extracting a center line of the blood vessel from the geometric model of the blood vessel;
intercepting a plurality of image patches along the extracted center line of the blood vessel; and
mapping the plurality of image patches to the blood vessel condition parameters using the deep learning model.

### Claim 11 (depends on 1)

The method of claim 1, wherein the blood vessel condition parameter includes at least one of fractional flow reserve, blood flow, blood flow velocity, micro vascular resistance, blood flow pressure drop and vascular stenosis.

### Claim 12 (depends on 1)

The method of claim 1, further comprising:
determining a treatment based on the blood vessel condition parameters of the blood vessel.

### Claim 13

A system for evaluating a blood vessel, comprising:
a processor, configured to:
receive image data of the blood vessel acquired by an image acquisition device; and
predict blood vessel condition parameters of the blood vessel by applying a deep learning model to the acquired image data of the blood vessel, wherein the deep learning model maps a sequence of image patches on the blood vessel to blood vessel condition parameters on the blood vessel, wherein in the mapping the entire sequence of image patches contribute to the blood vessel condition parameters; and
a display configured to display the blood vessel condition parameters of the blood vessel for evaluating the blood vessel.

### Claim 14 (depends on 13)

The system of claim 13, wherein the deep learning model includes a convolutional neural network, a recursive neural network, and a conditional random field model connected in series, wherein parameters of the convolutional neural network, the recursive neural network, and the conditional random field network are jointly optimized.

### Claim 15 (depends on 13)

The system of claim 13, wherein the deep learning model includes a plurality of convolutional neural networks connected in parallel, wherein each convolutional neural network receives a different sequence of image patches derived from the image data of the blood vessel, wherein the deep learning model further includes a connection layer that connects outputs of the plurality of convolutional neural networks.

### Claim 16 (depends on 15)

The system of claim 15, wherein the plurality of convolutional neural networks include a first convolutional neural network receiving a sequence of image patches corresponding to a first projection angle on the blood vessel and a second convolutional neural network receiving a sequence of image patches corresponding to a second projection angle on the blood vessel.

### Claim 17 (depends on 13)

The system of claim 13, wherein the deep learning model is trained using training images of blood vessels and corresponding blood vessel condition parameters obtained by applying computer simulations to the training images or by performing measurements to the blood vessels.

### Claim 18 (depends on 13)

The system of claim 13, wherein the processor is further configured to:
reconstruct a geometric model of the blood vessel using the acquired image data of the blood vessel;
extract a center line of the blood vessel from the geometric model of the blood vessel;
intercept a plurality of image patches along the extracted center line of the blood vessel; and
map the plurality of image patches to the blood vessel condition parameters using the deep learning model.

### Claim 19

A non-transitory computer readable storage medium storing computer executable instructions thereon, which, when executed by a processor, implement a method for evaluating a blood vessel, comprising:
receiving image data of the blood vessel acquired by an image acquisition device;
predicting, by a processor, blood vessel condition parameters of the blood vessel by applying a deep learning model to the acquired image data of the blood vessel, wherein the deep learning model maps a sequence of image patches on the blood vessel to blood vessel condition parameters on the blood vessel, wherein in the mapping the entire sequence of image patches contribute to the blood vessel condition parameters; and
providing the blood vessel condition parameters of the blood vessel for evaluating of the blood vessel.

### Claim 20 (depends on 19)

The non-transitory computer readable storage medium of claim 19, wherein the deep learning model includes a convolutional neural network, a recursive neural network, and a conditional random field model connected in series, wherein parameters of the convolutional neural network, the recursive neural network, and the conditional random field network are jointly optimized.
