CN-122029611-A - Usage prediction machine learning model for content selection

Abstract

Techniques for improving machine learning are provided. Usage information for a plurality of patients is accessed, the usage information indicating participation of each of the plurality of patients in respiratory therapy. Content delivery information for a plurality of patients is accessed. One or more parameters of the machine learning model are updated based on the usage information and the content delivery information to generate an updated machine learning model. A content selection is generated for a patient of the plurality of patients using the updated machine learning model, wherein a first content selection is delivered to a first patient.

Inventors

• YAN YANG
• CHEN JIAMING
• Andrew John Will
• Nathan James. Bartlett
• WANG LU

Assignees

• 瑞思迈数字医疗公司

Dates

Publication Date

20260512

Application Date

20240828

Priority Date

20230828

Claims (20)

1. 1.A method, comprising: accessing first usage information for a plurality of patients, the first usage information indicating participation in respiratory therapy by each of the plurality of patients; Accessing first content delivery information for the plurality of patients; updating one or more parameters of a machine learning model based on the first usage information and the first content delivery information to generate an updated machine learning model, and A first content selection is generated for a first patient using the updated machine learning model, wherein the first content selection is delivered to the first patient.
2. 2. The method of claim 1, further comprising, for each respective patient of the plurality of patients: generating respective content selections for the respective patient using the updated machine learning model, and The respective content selection is provided, wherein content is delivered to the respective patient in accordance with the respective content selection.
3. 3. The method of claim 1, wherein: Performing an update of the one or more parameters of the machine learning model on a first day, an The first usage information indicates a duration during which the first patient was engaged in the respiratory therapy during a first night immediately preceding the first day.
4. 4. The method of claim 3, wherein the first content delivery information indicates therapeutic content that was delivered to the first patient during a second day immediately preceding the first day.
5. 5. A method as claimed in claim 3, further comprising, on a second day immediately following the first day: Accessing second usage information of the plurality of patients, the second usage information indicating, for each respective patient of the plurality of patients, a respective duration for which the respective patient was engaged in the respiratory therapy during a second night immediately after the first night, and Updating the one or more parameters of the updated machine learning model based on the second usage information to generate a second updated machine learning model.
6. 6. The method of claim 1, further comprising accessing demographic information of the plurality of patients, wherein the one or more parameters of the machine learning model are further updated based on the demographic information.
7. 7. The method of claim 1, wherein: Updating the one or more parameters of the machine learning model includes, for each respective patient of the plurality of patients, generating a respective feature vector based on respective usage information of the respective patient and respective content delivery information of the respective patient, The respective identity of the respective patient and the respective content selection indicated by the respective content delivery information are encoded using a one-time-hot encoding.
8. 8. The method of claim 7, further comprising converting the feature vector to a sparse matrix data format, wherein the one or more parameters of the machine learning model are updated based on the feature vector in the sparse matrix data format being used as an input to the machine learning model.
9. 9. The method of claim 1, wherein generating the first content selection comprises: generating predicted usage information using the updated machine learning model for each respective content selection of a plurality of content selections, and The first content selection is selected based on the predicted usage information.
10. 10. A non-transitory computer-readable medium comprising computer-executable instructions that, when executed by one or more processors of a processing system, cause the processing system to perform operations comprising: accessing first usage information for a plurality of patients, the first usage information indicating participation in respiratory therapy by each of the plurality of patients; Accessing first content delivery information for the plurality of patients; updating one or more parameters of a machine learning model based on the first usage information and the first content delivery information to generate an updated machine learning model, and A first content selection is generated for a first patient using the updated machine learning model, wherein the first content selection is delivered to the first patient.
11. 11. The non-transitory computer-readable medium of claim 10, the operations further comprising, for each respective patient of the plurality of patients: generating respective content selections for the respective patient using the updated machine learning model, and The respective content selection is provided, wherein content is delivered to the respective patient in accordance with the respective content selection.
12. 12. The non-transitory computer-readable medium of claim 10, wherein: Updating the one or more parameters of the machine learning model is performed on a first day, The first usage information indicates a duration of time during which the first patient was engaged in the respiratory therapy during a first night immediately preceding the first day, and The first content delivery information indicates treatment content that was delivered to the first patient during a second day immediately preceding the first day.
13. 13. The non-transitory computer-readable medium of claim 12, the operations further comprising, on a second day immediately after the first day: Accessing second usage information of the plurality of patients, the second usage information indicating, for each respective patient of the plurality of patients, a respective duration for which the respective patient was engaged in the respiratory therapy during a second night immediately after the first night, and Updating the one or more parameters of the updated machine learning model based on the second usage information to generate a second updated machine learning model.
14. 14. The non-transitory computer-readable medium of claim 10, wherein: Updating the one or more parameters of the machine learning model includes, for each respective patient of the plurality of patients, generating a respective feature vector based on respective usage information of the respective patient and respective content delivery information of the respective patient, The respective identity of the respective patient and the respective content selection indicated by the respective content delivery information are encoded using a one-time-hot encoding.
15. 15. The non-transitory computer-readable medium of claim 14, further comprising converting the feature vector to a sparse matrix data format, wherein the one or more parameters of the machine learning model are updated based on the feature vector in the sparse matrix data format being used as input to the machine learning model.
16. 16. The non-transitory computer-readable medium of claim 10, wherein generating the first content selection comprises: generating predicted usage information using the updated machine learning model for each respective content selection of a plurality of content selections, and The first content selection is selected based on the predicted usage information.
17. 17. A system, comprising: A memory including computer executable instructions, and One or more processors configured to execute the computer-executable instructions and cause the system to perform operations comprising: accessing first usage information for a plurality of patients, the first usage information indicating participation in respiratory therapy by each of the plurality of patients; Accessing first content delivery information for the plurality of patients; updating one or more parameters of a machine learning model based on the first usage information and the first content delivery information to generate an updated machine learning model, and A first content selection is generated for a first patient using the updated machine learning model, wherein the first content selection is delivered to the first patient.
18. 18. The system of claim 17, wherein: Updating the one or more parameters of the machine learning model is performed on a first day, The first usage information indicates a duration of time during which the first patient was engaged in the respiratory therapy during a first night immediately preceding the first day, and The first content delivery information indicates treatment content that was delivered to the first patient during a second day immediately preceding the first day.
19. 19. The system of claim 18, the operations further comprising, on a second day immediately after the first day: Accessing second usage information of the plurality of patients, the second usage information indicating, for each respective patient of the plurality of patients, a respective duration for which the respective patient was engaged in the respiratory therapy during a second night immediately after the first night, and Updating the one or more parameters of the updated machine learning model based on the second usage information to generate a second updated machine learning model.
20. 20. The system of claim 17, wherein: Updating the one or more parameters of the machine learning model includes, for each respective patient of the plurality of patients, generating a respective feature vector based on respective usage information of the respective patient and respective content delivery information of the respective patient, Encoding a respective identity of the respective patient and a respective content selection indicated by the respective content delivery information using a one-time thermal encoding, and The operations further include converting the feature vector to a sparse matrix data format, wherein the one or more parameters of the machine learning model are updated based on the feature vector in the sparse matrix data format being used as an input to the machine learning model.

Description

Usage prediction machine learning model for content selection Cross Reference to Related Applications The present application claims the benefit and priority of U.S. provisional patent application No. 63/579,217 filed on 8/28 of 2023, the entire contents of which are incorporated herein by reference. Technical Field The present disclosure relates generally to machine learning, and more particularly to using machine learning to predict respiratory therapy usage. Many people suffer from sleep-related and/or breath-related diseases such as, for example, periodic Limb Movement Disorder (PLMD), restless Leg Syndrome (RLS), sleep Disordered Breathing (SDB) such as Obstructive Sleep Apnea (OSA) and Central Sleep Apnea (CSA), cheyne-stokes respiration (CSR), respiratory insufficiency, obesity Hypoventilation Syndrome (OHS), chronic Obstructive Pulmonary Disease (COPD), neuromuscular disease (NMD) and chest wall disease. These diseases are usually treated using respiratory therapy systems. Each respiratory therapy system typically has a respiratory therapy device that is connected to a user interface (e.g., mask) via a conduit and optional connectors. The user wears a user interface and supplies a flow of compressed air from the respiratory therapy device via the conduit. The user interface is typically a user interface for a particular category and type of user, such as a direct or indirect connection of the user interface type, as well as a full face mask, partial face mask, nasal mask, or nasal pillow of the user interface type. The user interface is typically a specific model number produced by a specific manufacturer, such as AIRFITTM F manufactured by ResMed, except for a specific category and type. In some cases, patient usage of the treatment system may be collected or monitored in order to assess user compliance with the treatment. For example, the duration of use (e.g., the length of time a user wears the user interface at a given night) may be determined. In general, various factors may affect compliance and use cases, and predicting or identifying compliance (or improving use cases) may be difficult. While many patients will benefit from increased therapeutic use (e.g., use their masks for a longer period of time every night), it is often difficult or impossible to effectively predict and improve use. Improved systems and techniques for predicting use and thereby improving treatment and outcome are needed. Disclosure of Invention According to some implementations of the present disclosure, a method includes accessing first usage information of a plurality of patients, the first usage information indicating participation in respiratory therapy by each of the plurality of patients (participation), accessing first content delivery information of the plurality of patients, updating one or more parameters of a machine learning model based on the first usage information and the first content delivery information to generate an updated machine learning model, and generating a first content selection for a first patient of the plurality of patients using the updated machine learning model, wherein the first content selection is delivered to the first patient. According to some implementations of the present disclosure, a system includes a control system and a memory. The control system includes one or more processors. The memory has machine-readable instructions stored thereon. The control system is coupled to the memory and when machine-executable instructions in the memory are executed by at least one of the one or more processors of the control system, implement any of the methods disclosed herein. Other aspects provide a processing system configured to perform the foregoing method and the methods described herein, a non-transitory computer-readable medium comprising instructions that, when executed by one or more processors of the processing system, cause the processing system to perform the foregoing method and the methods described herein, a computer program product embodied on a computer-readable storage medium, the computer program product comprising code for performing the foregoing method and the methods further described herein, and a processing system comprising means for performing the foregoing method and the methods further described herein. The above summary is not intended to represent each implementation or every aspect of the present disclosure. Additional features and advantages of the present disclosure will be apparent from the detailed description and drawings set forth below. Drawings Fig. 1 depicts an example environment for predicting therapeutic use and selecting content in accordance with some implementations of the present disclosure. FIG. 2 depicts an example workflow for predicting usage based on content consumption using machine learning in accordance with some implementations of the present disclosure. FIG. 3 is a flowchart depicting an example method for predicting usage and pro