Patent Application Titled “Privacy-Preserving Computing On Subject Data Used To Develop Artificial Intelligence Tools” Published Online (USPTO 20230259654): Genentech Inc.

2023 SEP 04 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- According to news reporting originating from Washington, D.C., by NewsRx journalists, a patent application by the inventors Dridi, Abdesslem (South San Francisco, CA, US); Jalal, Niaz Ahsan (South San Francisco, CA, US), filed on April 24, 2023, was made available online on August 17, 2023.

The assignee for this patent application is Genentech Inc. (South San Francisco, California, United States).

Reporters obtained the following quote from the background information supplied by the inventors: “Modern computing paradigms, including cloud computing, data parallel cluster computing, and high performance computing, combined with a widely available variety of machine learning and deep learning algorithmic architectures, have created an environment in which a vast array of artificial intelligence (AI) applications can be developed to solve problems in almost any industry, if enough data is available to optimize the underlying algorithm properly. It is now clear that access to data is a primary barrier to the development of AI applications. In fact, in many industries, it is necessary to use data from a variety of sources in order to create AI that is robust, generalizable, and unbiased. A specific challenge is that, in general, the owners of data often cannot or will not share the data or allow the data to leave their control. This is understandable since data often contains highly sensitive private and/or personal data and can be regulated in ways that make it difficult or impossible to share. These challenges are particularly difficult to overcome in the development of healthcare AI.

“In healthcare AI, data-driven technology solutions are being developed to further personalize healthcare all while reducing costs. Healthcare providers are innovating solutions for automating and streamlining the process of analyzing subject data to determine a medical prediction. Machine-learning (ML) techniques may be used for a number of healthcare-related predictions, such as disease diagnosis and prognosis as well as for predicting treatment efficacy. Because medical data typically contains private/identification data for the subjects from which it is generated, government regulation (e.g., Health Insurance Portability and Accountability Act (HIPAA), “good practice” quality guidelines and regulations (GxP), and General Data Protection Regulation (GDPR) compliance becomes a unique challenge for healthcare providers looking into machine learning for medical analysis. Training ML models can involve a large amount of data, so it can be difficult to access a sufficient amount of data that is de-identified and/or anonymize to train effective ML models. Accordingly, there is a need for advances in compliant software platforms, built to provide accurate medical predictions while ensuring the confidentiality, availability and integrity of protected healthcare information.”

In addition to obtaining background information on this patent application, NewsRx editors also obtained the inventors’ summary information for this patent application: “In various embodiments, a computer-implemented method is provided that comprises: receiving subject data regarding a first subject from a first computing device associated with the first subject; performing a de-identifying operation, an anonymizing operation, or both on the subject data to generate processed subject data; storing the processed subject data in a processed data store accessible to the local cloud server; sending a batch of data to a remote cloud server, the batch of data comprising the processed subject data; receiving a production model from the remote cloud server, the production model including parameters derived in part from the processed subject data; receiving subsequent data regarding a second subject from a second computing device associated with the second subject; inputting the subsequent data into the production model to analyze the subsequent data and generate an inference or prediction from the analysis of the subsequent data; and sending the inference or the prediction to the second computing device, a third computing device, or both for use in one or more operations performed by the second computing device, the third computing device, or a combination thereof.

“In some embodiments, the local cloud server is physically located in a same geographic region as the subject.

“In some embodiments, the same geographic region is a same country.

“In some embodiments, the subject data is health care data comprising individually identifiable health information and the subsequent data is subsequent healthcare data comprising individually identifiable health information.

“In some embodiments, the same geographic region collectively shares a set of data regulations regarding use and storage of the individually identifiable health information.

“In some embodiments, the de-identifying operation, the anonymizing operation, or both are performed on the individually identifiable health information of the subject data based on the set of data regulations.

“In some embodiments, the first computing device is the same or different device as the second computing device.

“In some embodiments, the first computing device is a clinical device sensor, a handheld portable device, or a combination thereof.

“In some embodiments, the second computing device is a clinical device sensor, a handheld portable device, or a combination thereof.

“In some embodiments, the first subject is the same or different subject as the second subject.

“In some embodiments, the processed data store is not accessible to the remote cloud server.

“In some embodiments, sending the processed subject data as a part of the batch of data to the remote cloud server occurs responsive to the local cloud server having not received a request for deletion of the processed subject data prior to the sending the processed subject data.

“In some embodiments, the method further comprises: prior to performing the de-identifying operation, the anonymizing operation, or both on the subject data, storing the subject data in a raw data store accessible to the local cloud server; receiving a request to delete the subject data from the remote cloud server; and in response to receiving the request to delete the subject data, deleting the subject data from the raw data store.

“In some embodiments, the processed subject data is not deleted from the processed data store.

“In some embodiments, the sending the processed subject data as a part of the batch of data occurs at a periodic or stochastic timing such that the batch of data includes data from multiple other subjects captured since a previous sending of data to the remote cloud server.

“In some embodiments, the inference or the prediction are generated with respect to a diagnosis, a prognosis, a treatment or therapy, identification of a treatment or therapy protocol, detection or determination of a disease state, identification or detection of a biomarker, a reduction in treatment or therapy non-adherence, a reduction in operational cost, image analysis, marketing of a treatment or therapy, automation of an administrative task, assistance with a medical procedure, or any combination thereof.

“In some embodiments, the one or more operations include communicating or displaying the inference or the prediction, analysis of the =inference or the prediction, providing a treatment or therapy, initiating a treatment or therapy protocol, measuring a biomarker, providing a notice or reminder for a treatment or therapy, obtaining healthcare data, reporting a diagnosis or a prognosis, analyzing an image, providing marketing regarding a treatment or therapy, control of a medical device or a medical procedure, or any combination thereof.

“In some embodiments, the method further comprises sending a subsequent batch of data to the remote cloud server, the subsequent batch of data comprising the inference or the prediction.

“In some embodiments, the method further comprises: performing the de-identifying operation, the anonymizing operation, or both on the subsequent data to generate processed subsequent data; storing the processed subsequent data in the processed data store accessible to the local cloud server; and sending a batch of data to the remote cloud server, the batch of data comprising the processed subsequent data.

“In some embodiments, the method further comprises: performing the de-identifying operation, the anonymizing operation, or both on the inference or the prediction to generate processed output data; storing the processed output data in the processed data store accessible to the local cloud server; and sending a batch of data to the remote cloud server, the batch of data comprising the processed output data.

“In some embodiments, the method further comprises: receiving response data at the local cloud server from the second computing device, the third computing device, or both generated in response to receiving the inference or the prediction or using the inference or the prediction in the one or more operations; performing the de-identifying operation, the anonymizing operation, or both on the response data to generate processed response data; storing the processed response data in the processed data store; and sending a batch of data to the remote cloud server, the batch of data comprising the processed response data.

“In some embodiments, the method further comprises: receiving a new production model from the remote cloud server, the new production model including parameters derived in part from the processed response data; and replacing the production model with the new production model, wherein the replacing includes deleting the production model from the local cloud server.

“In various embodiments, a computer-implemented method is provided that comprises: receiving processed subject data associated with a plurality of different subjects from a local cloud server, the processed subject data having been de-identified, anonymized, or both; associating the processed subject dataset with a versioned dataset; determining an expiration date for the versioned dataset; storing the versioned dataset in a version data store accessible to the remote cloud server, the versioned dataset stored in association with the expiration date; training a production model using the versioned dataset; storing, in the versioned data store, an association between the versioned dataset and the production model trained with the version dataset; and sending the production model to the local cloud server for use in analyzing subsequent data and generating an inference or prediction from the analysis of the subsequent data.

“In some embodiments, the local cloud server is physically located in a same geographic region as the subjects.

“In some embodiments, the same geographic region is a same country.

“In some embodiments, the processed subject data is health care data comprising individually identifiable health information and the subsequent data is subsequent healthcare data comprising individually identifiable health information.

“In some embodiments, the remote cloud server is physically located in a same or different geographic region as the local cloud server.

“In some embodiments, the same or different geographic region is a same or different country.

“In some embodiments, the version data store is not accessible to the local cloud server.

“In some embodiments, the same geographic region collectively shares a set of data regulations regarding use and storage of the individually identifiable health information.

“In some embodiments, the individually identifiable health information of the processed subject data has been de-identified, anonymized, or both based on the set of data regulations.

“In some embodiments, the expiration date is determined based a date of creation of the versioned dataset, a date of receipt of the processed subject data, an expiry of an informed consent form associated with the processed subject dataset, or any combination thereof.

“In some embodiments, the method further comprises: storing the production model in the versioned data store accessible to the remote cloud server; and in response to the expiration date passing, deleting or removing the versioned dataset and the production model from the versioned data store.”

There is additional summary information. Please visit full patent to read further.”

The claims supplied by the inventors are:

“1. A computer-implemented method carried out by a local cloud server comprising: receiving subject data regarding a first subject from a first computing device associated with the first subject; performing a de-identifying operation, an anonymizing operation, or both on the subject data to generate processed subject data; storing the processed subject data in a processed data store accessible to the local cloud server; sending a batch of data to a remote cloud server, the batch of data comprising the processed subject data; receiving a production model from the remote cloud server, the production model including parameters derived in part from the processed subject data; receiving subsequent data regarding a second subject from a second computing device associated with the second subject; inputting the subsequent data into the production model to analyze the subsequent data and generate an inference or prediction from the analysis of the subsequent data; and sending the inference or the prediction to the second computing device, a third computing device, or both for use in one or more operations performed by the second computing device, the third computing device, or a combination thereof.

“2. The computer-implemented method of claim 1, wherein the local cloud server is physically located in a same geographic region as the subject.

“3. The computer-implemented method of claim 2, wherein the same geographic region is a same country.

“4. The computer-implemented method of claim 2, wherein the subject data is health care data comprising individually identifiable health information and the subsequent data is subsequent healthcare data comprising individually identifiable health information.

“5. The computer-implemented method of claim 4, wherein the same geographic region collectively shares a set of data regulations regarding use and storage of the individually identifiable health information.

“6. The computer-implemented method of claim 4, wherein the de-identifying operation, the anonymizing operation, or both are performed on the individually identifiable health information of the subject data based on the set of data regulations.

“7. The computer-implemented method of claim 1, wherein sending the processed subject data as a part of the batch of data to the remote cloud server occurs responsive to the local cloud server having not received a request for deletion of the processed subject data prior to the sending the processed subject data.

“8. The computer-implemented method of claim 1, further comprising: prior to performing the de-identifying operation, the anonymizing operation, or both on the subject data, storing the subject data in a raw data store accessible to the local cloud server; receiving a request to delete the subject data from the remote cloud server; and in response to receiving the request to delete the subject data, deleting the subject data from the raw data store.

“9. The computer-implemented method of claim 1, wherein the sending the processed subject data as a part of the batch of data occurs at a periodic or stochastic timing such that the batch of data includes data from multiple other subjects captured since a previous sending of data to the remote cloud server.

“10. The computer-implemented method of claim 1, wherein the inference or the prediction are generated with respect to a diagnosis, a prognosis, a treatment or therapy, identification of a treatment or therapy protocol, detection or determination of a disease state, identification or detection of a biomarker, a reduction in treatment or therapy non-adherence, a reduction in operational cost, image analysis, marketing of a treatment or therapy, automation of an administrative task, assistance with a medical procedure, or any combination thereof.

“11. The computer-implemented method of claim 1, wherein the one or more operations include communicating or displaying the inference or the prediction, analysis of the inference or the prediction, providing a treatment or therapy, initiating a treatment or therapy protocol, measuring a biomarker, providing a notice or reminder for a treatment or therapy, obtaining healthcare data, reporting a diagnosis or a prognosis, analyzing an image, providing marketing regarding a treatment or therapy, control of a medical device or a medical procedure, or any combination thereof.

“12. The computer-implemented method of claim 1, further comprising sending a subsequent batch of data to the remote cloud server, the subsequent batch of data comprising the inference or the prediction.

“13. The computer-implemented method of claim 1, further comprising: performing the de-identifying operation, the anonymizing operation, or both on the subsequent data to generate processed subsequent data; storing the processed subsequent data in the processed data store accessible to the local cloud server; and sending a batch of data to the remote cloud server, the batch of data comprising the processed subsequent data.

“14. The computer-implemented method of claim 1, further comprising: performing the de-identifying operation, the anonymizing operation, or both on the inference or the prediction to generate processed output data; storing the processed output data in the processed data store accessible to the local cloud server; and sending a batch of data to the remote cloud server, the batch of data comprising the processed output data.

“15. The computer-implemented method of claim 1, further comprising: receiving response data at the local cloud server from the second computing device, the third computing device, or both generated in response to receiving the inference or the prediction or using the inference or the prediction in the one or more operations; performing the de-identifying operation, the anonymizing operation, or both on the response data to generate processed response data; storing the processed response data in the processed data store; and sending a batch of data to the remote cloud server, the batch of data comprising the processed response data.

“16. The computer-implemented method of claim 15, further comprising: receiving a new production model from the remote cloud server, the new production model including parameters derived in part from the processed response data; and replacing the production model with the new production model, wherein the replacing includes deleting the production model from the local cloud server.

“17. A computer-implemented method carried out by a remote cloud server comprising: receiving processed subject data associated with a plurality of different subjects from a local cloud server, the processed subject data having been de-identified, anonymized, or both; associating the processed subject dataset with a versioned dataset; determining an expiration date for the versioned dataset; storing the versioned dataset in a version data store accessible to the remote cloud server, the versioned dataset stored in association with the expiration date; training a production model using the versioned dataset; storing, in the versioned data store, an association between the versioned dataset and the production model trained with the version dataset; and sending the production model to the local cloud server for use in analyzing subsequent data and generating an inference or prediction from the analysis of the subsequent data.

“18. The computer-implemented method of claim 17, wherein the local cloud server is physically located in a same geographic region as the subjects.

“19. The computer-implemented method of claim 18, wherein the same geographic region is a same country.

“20. The computer-implemented method of claim 18, wherein the processed subject data is health care data comprising individually identifiable health information and the subsequent data is subsequent healthcare data comprising individually identifiable health information.”

For more information, see this patent application: Dridi, Abdesslem; Jalal, Niaz Ahsan. Privacy-Preserving Computing On Subject Data Used To Develop Artificial Intelligence Tools. U.S. Patent Application Number 20230259654, filed April 24, 2023 and posted August 17, 2023. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(20230259654)&db=US-PGPUB&type=ids

(Our reports deliver fact-based news of research and discoveries from around the world.)