# US 2017/0281011 A1 · Systems and methods of image processing to determine flow characteristics

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

- Number: US 2017/0281011 A1
- Document ID: US20170281011A1
- Status: Published application
- Kind: A1
- Application number: 15/624,424
- Filed: 2017-06-15
- Published: 2017-10-05
- Earliest claimed: 2012-09-12
- Assignees: HeartFlow, Inc.
- Applicants: HeartFlow, Inc.
- Inventors: Timothy FONTE; Gilwoo CHOI; Leo GRADY; Michael SINGER
- Web page: https://patentgrep.com/patent/US20170281011A1

## CPC

- A61B 5/026: Measuring blood flow {(A61B3/1233, A61B3/1241 take precedence)}
- A61B 5/0022: {Monitoring a patient using a global network, e.g. telephone networks, internet}
- A61B 5/02007: {Evaluating blood vessel condition, e.g. elasticity, compliance}
- A61B 5/0205: Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
- A61B 5/021: Measuring pressure in heart or blood vessels
- A61B 5/107: Measuring physical dimensions, e.g. size of the entire body or parts thereof
- A61B 5/14535: {for measuring haematocrit}
- A61B 5/14546: {for measuring analytes not otherwise provided for, e.g. ions, cytochromes}
- A61B 5/7267: {involving training the classification device}
- A61B 5/7278: {Artificial waveform generation or derivation, e.g. synthesizing signals from measured signals}
- A61B 5/7282: {Event detection, e.g. detecting unique waveforms indicative of a medical condition (cough events A61B5/0823; seizures A61B5/4094; sleep apnoea A61B5/4818)}
- A61B 5/742: {using visual displays (displays for heart-related electrical signals, e.g. ECG, A61B5/339)}
- A61B 5/743: {Displaying an image simultaneously with additional graphical information, e.g. symbols, charts, function plots}
- A61B 6/032: {Transmission computed tomography [CT]}
- A61B 6/463: {characterised by displaying multiple images or images and diagnostic data on one display}
- 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 6/5217: {extracting a diagnostic or physiological parameter from medical diagnostic data}
- A61B 6/563: {involving image data transmission via a network}
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/00: Machine learning
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- A61B 5/024: Measuring pulse rate or heart rate
- A61B 2560/0475: Special features of memory means, e.g. removable memory cards
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Systems and methods are disclosed for determining individual-specific blood flow characteristics. One method includes acquiring, for each of a plurality of individuals, individual-specific anatomic data and blood flow characteristics of at least part of the individual's vascular system; executing a machine learning algorithm on the individual-specific anatomic data and blood flow characteristics for each of the plurality of individuals; relating, based on the executed machine learning algorithm, each individual's individual-specific anatomic data to functional estimates of blood flow characteristics; acquiring, for an individual and individual-specific anatomic data of at least part of the individual's vascular system; and for at least one point in the individual's individual-specific anatomic data, determining a blood flow characteristic of the individual, using relations from the step of relating individual-specific anatomic data to functional estimates of blood flow characteristics.

## Claims

### Claim at position 1

1-31. (canceled)

### Claim 32

A method for personalized non-invasive assessment of renal artery stenosis for a patient, comprising:
receiving medical image data of the patient;
extracting patient-specific renal arterial geometry of the patient from the medical image data;
extracting features from the patient-specific renal arterial geometry of the patient; and
computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries.

### Claim 33 (depends on 32)

The method of claim 32, further comprising:
receiving at least one of non-invasive physiological measurements of the patient or demographic data of the patient,
wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing the hemodynamic index for the one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features from the patient-specific renal arterial geometry; and
based on features corresponding to the at least one of the non-invasive physiological measurements of the patient or the demographic data of the patient using the trained machine-learning based surrogate model.

### Claim 34 (depends on 32)

The method of claim 32, further comprising
receiving one or more biomarkers of the patient,
wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing the hemodynamic index for the one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features from the patient-specific renal arterial geometry and based on features corresponding to the one or more biomarkers of the patient using the trained machine-learning based surrogate model.

### Claim 35 (depends on 32)

The method of claim 32, wherein the trained machine-learning based surrogate model is trained based on renal blood flow simulation results obtained from renal blood flow simulations performed using a respective patient-specific model of renal circulation.

### Claim 36 (depends on 32)

The method of claim 32, wherein extracting features from the patient-specific renal arterial geometry of the patient further comprises:
extracting a plurality of geometric measurements for one or more renal artery stenosis regions in the patient-specific renal arterial geometry of the patient.

### Claim 37 (depends on 32)

The method of claim 32, wherein the one or more locations of interest correspond to one or more renal artery stenosis locations,
computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
automatically computing the hemodynamic index for the one or more renal artery stenosis locations in the patient-specific renal arterial geometry based on the extracted features using the trained machine-learning based surrogate model.

### Claim 38 (depends on 32)

The method of claim 32, wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing the hemodynamic index for the one or more locations of interest in the patient-specific renal arterial geometry in response to a user input identifying the one or more locations of interest.

### Claim 39 (depends on 32)

The method of claim 32, wherein the hemodynamic index is renal fractional flow reserve (rFFR) at hyperemia.

### Claim 40 (depends on 32)

The method of claim 32, wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing a plurality of hemodynamic indices for each of the one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a respective one of a plurality of trained machine-learning based surrogate models trained based on the features extracted from the synthetically generated renal arterial tree geometries.

### Claim 41 (depends on 32)

The method of claim 32, further comprising: displaying a visualization of the patient-specific renal arterial geometry color coded based on the hemodynamic index computed for the one or more locations of interest in the patient-specific renal arterial geometry.

### Claim 42

A method of training a data-driven surrogate model for predicting a hemodynamic index for non-invasive assessment renal artery stenosis, comprising:
generating a plurality of synthetic renal arterial geometries having renal artery stenosis regions with varying geometries; performing blood flow simulations for the plurality of synthetic renal arterial geometries;
computing hemodynamic index values at a plurality of locations in each of the plurality of synthetic renal arterial geometries based on the blood flow simulations;
extracting geometric features from the plurality of synthetic renal arterial geometries; and
training a surrogate model to map the geometric features extracted from the plurality of synthetic renal arterial geometries to the hemodynamic index values computed at the plurality of locations in each of the plurality of synthetic renal arterial geometries using a machine learning algorithm.

### Claim 43 (depends on 42)

The method of claim 42, wherein performing blood flow simulations for the plurality of synthetic renal arterial geometries comprises:
for each of the synthetic renal arterial geometries, performing a blood flow simulation using a corresponding patient-specific multi-scale model of renal arterial circulation including a lumped model.

### Claim 44

An apparatus for personalized non-invasive assessment of renal artery stenosis for a patient, comprising:
receiving medical image data of the patient;
extracting patient-specific renal arterial geometry of the patient from the medical image data;
extracting features from the patient-specific renal arterial geometry of the patient; and
computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries.

### Claim 45 (depends on 44)

The apparatus of claim 44, wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing the hemodynamic index for the one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features from the patient-specific renal arterial geometry and based on features corresponding to at least one of the non-invasive physiological measurements of the patient or the demographic data of the patient using the trained machine-learning based surrogate model.

### Claim 46 (depends on 44)

The apparatus of claim 44, wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing the hemodynamic index for the one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features from the patient-specific renal arterial geometry and based on features corresponding to one or more biomarkers of the patient using the trained machine-learning based surrogate model.

### Claim 47 (depends on 44)

The apparatus of claim 44, wherein the trained machine-learning based surrogate model is trained based on the features extracted from the synthetically generated renal arterial tree geometries and renal blood flow simulation results obtained from renal blood flow simulations performed using a respective patient-specific multi-scale model of renal circulation including a model for each of the synthetically generated renal arterial geometries.

### Claim 48 (depends on 44)

The apparatus of claim 44, wherein the hemodynamic index is renal fractional flow reserve (rFFR) at rest or at hyperemia.

### Claim 49 (depends on 44)

The apparatus of claim 44, wherein computing a hemodynamic index for one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a trained machine-learning based surrogate model trained based on features extracted from synthetically generated renal arterial tree geometries comprises:
computing a plurality of hemodynamic indices for each of the one or more locations of interest in the patient-specific renal arterial geometry based on the extracted features using a respective one of a plurality of trained machine-learning based surrogate models trained based on the features extracted from the synthetically generated renal arterial tree geometries.

### Claim 50 (depends on 44)

The apparatus of claim 44, further comprising: displaying a visualization of the patient-specific renal arterial geometry color coded based on the hemodynamic index computed for the one or more locations of interest in the patient-specific renal arterial geometry.
