Granted Patent B1

US 12,333,413 · App. 18/592,680 · Granted Jun 17, 2025

Inventors: Eunjung Lee (Rochester, MN); Francisco Lopez-Jimenez (Rochester, MN); Paul A. Friedman (Rochester, MN); Itzhak Zachi Attia (Rochester, MN); Jae K. Oh (Rochester, MN)

Assignee: Mayo Foundation for Medical Education and Research

G06N3/0464A61B5/349A61B8/0883G16H50/20G16H30/40

Abstract

Claims (49)

1. An apparatus for artificial intelligence-supported diagnostic assessment, wherein the apparatus comprises: a sensor for sensing a time series input;

at least a processor, wherein the processor is configured to:

receive a plurality of multi-channel sensor readings of physiological data;

generate training data correlating each of the plurality of multi-channel sensor readings with a plurality of diagnostic labels;

train a neural network using the plurality of diagnostic labels;

update the training data as a function of new training examples;

retrain the neural network as a function of the updated training data, wherein the retraining occurs when a preconfigured threshold of new training examples is exceeded;

receive the time series input describing user physiological data from at least the sensor;

input the time series input into the trained neural network;

generate diagnostic data as a function of the time series input and the trained neural network, wherein the diagnostic data includes a classification under a plurality of groups, wherein the groups are comprised of:

normal LV function;

LVSD only;

LVDD only; and

both LVSD and LVDD;

determine prognostic data as a function of the diagnostic data; and

output the prognostic data; and a display device, wherein the output of the prognostic data is displayed on the display device.

2. The apparatus of claim 1 , wherein the time series input is comprised of electrocardiogram (ECG) data.

3. The apparatus of claim 2 , wherein the plurality of multi-channel sensor readings further comprises a plurality of ECG readings, and the plurality of training data further comprises the plurality of ECG readings, and generating the plurality of training data further comprises:

receiving echocardiogram data correlated to each ECG reading of the plurality of ECG readings; and

generating the plurality of diagnostic labels using the echocardiogram data.

4. The apparatus of claim 1 , wherein the neural network is a multi-output convolutional neural network.

5. The apparatus of claim 1 , wherein the neural network is trained using general-use lead data, wherein the general-use lead data refers to the lead data used for a diagnostic tool.

6. The apparatus of claim 1 , wherein the neural network is trained using reduced general-use lead data, wherein the reduced general-use lead data refers to a reduction in the leads used in acquiring the lead data.

7. The apparatus of claim 1 , wherein the neural network is trained as a function of lead data gathered from 12-lead ECGs performed within 14 days of transthoracic echocardiography, wherein the plurality of multi-channel sensor readings of physiological data are comprised of the 12-lead ECG data and the lead data correlated to each 12-lead ECG datum.

8. The apparatus of claim 1 , wherein the diagnostic data includes an assessment of the LVDD.

9. A method for training an artificial intelligence-supported diagnostic assessment tool, wherein the method comprises: sensing a time series input via a sensor;

receiving a plurality of multi-channel sensor readings of physiological data;

generating training data correlating each of the plurality of multi-channel sensor readings with a plurality of diagnostic labels;

training a neural network using the plurality of diagnostic labels;

updating the training data as a function of new training examples;

retraining the neural network as a function of the updated training data, wherein the retraining occurs when a preconfigured threshold of new training examples is exceeded;

receiving the time series input describing user physiological data from at least the sensor;

inputting the time series input into the trained neural network;

generating diagnostic data as a function of the time series input and the trained neural network, wherein the diagnostic data includes a classification under a plurality of groups, wherein the groups are comprised of:

normal LV function;

LVSD only;

LVDD only; and

both LVSD and LVDD;

determining prognostic data as a function of the diagnostic data; and

outputting the prognostic data; and displaying the output of the prognostic data on a display device.

10. The method of claim 9 , wherein the time series input is comprised of electrocardiogram (ECG) data.

11. The method of claim 10 , wherein the plurality of multi-channel sensor readings further comprises a plurality of ECG readings, and the plurality of training data further comprises the plurality of ECG readings, and generating the plurality of training data further comprises:

receiving echocardiogram data correlated to each ECG reading of the plurality of ECG readings; and

generating the plurality of diagnostic labels using the echocardiogram data.

12. The method of claim 9 , wherein the neural network is a multi-output convolutional neural network.

13. The method of claim 9 , wherein the neural network is trained using general-use lead data, wherein the general-use lead data refers to the lead data used for a diagnostic tool.

14. The method of claim 9 , wherein the neural network is trained using reduced general-use lead data, wherein the reduced general-use lead data refers to a reduction in the leads used in acquiring the lead data.

15. The method of claim 9 , wherein the neural network is trained using lead data gathered from 12-lead ECGs performed within 14 days of transthoracic echocardiography, wherein the plurality of multi-channel sensor readings of physiological data are comprised of the 12-lead ECG data and the lead data correlated to each 12-lead ECG datum.

16. The method of claim 9 , wherein the diagnostic data includes an assessment of both the LVSD and the LVDD.

Assignments (1)

References Cited (17)