Patent Issued for Machine learning models for diagnosis suspecting (USPTO 11587678): Clover Health

2023 MAR 09 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- Clover Health (Jersey City, New Jersey, United States) has been issued patent number 11587678, according to news reporting originating out of Alexandria, Virginia, by NewsRx editors.

The patent’s inventors are Goetz, Melanie (Oakland, CA, US), Lauinger, Christopher James (Golden, CO, US).

This patent was filed on March 23, 2020 and was published online on February 21, 2023.

From the background information supplied by the inventors, news correspondents obtained the following quote: “Determining a diagnosis for an undiagnosed disease or condition in a medical patient may be desired. Described herein are improvements in technology and solutions to technical problems that can be used to, among other things, assist in determining diagnoses.”

Supplementing the background information on this patent, NewsRx reporters also obtained the inventors’ summary information for this patent: “Systems and methods for utilizing machine learning models for diagnosis suspecting are described herein. Accurate diagnosing relies on data to drive probabilities for decision making. This process can be optimized and improved through the use of large, complex data sets and machine learning to recognize patterns and provide accurate probabilities of diagnosis. For example, data from disparate sources that may not traditionally be viewed or viewable together for diagnostic purposes (e.g. laboratory data, medical record data, geographical data) can be combined into larger data sets, which can then utilize machine learning models and data analytic pipelines to diagnose diseases that may otherwise go undetected. These diagnoses may then be confirmed, or rejected, by a medical service provider (e.g. physician, nurse practitioner, physician’s assist). In examples, these diagnoses may be surfaced and then presented to a medical service provider (e.g. physician, nurse practitioner, registered dietitian, physician’s assistant), wherein a given diagnosis may be either confirmed or rejected. In these and other examples, the confirmation or rejection of the diagnosis may then be used to update the data and machine learning models to improve diagnosis accuracy for future diagnosis suspecting and surfacing.

“The present innovation is directed to systems and methods that generate machine learning models configured to diagnose one or more diseases or conditions and utilize the trained machine learning models to determine a likelihood that a disease or condition should be diagnosed in a medical patient and then surfacing that diagnoses to a medical service provider. By way of example, machine learning models may be trained on a large data set, wherein individual machine learning models are trained to determine a likelihood that a disease or condition should be diagnosed in a medical patient.

“The data used to train the machine learning models may be received into a database from multiple disparate data sources via a computing network. On their own, the data from the disparate sources are likely to exist in disparate formats and may be formatted into coherent data structures and formats, particularly formats associated with the machine learning models. These data structures may be multi-dimensional data and may include associated meta-data that may be unpacked and formatted within the data structures. Annotations may be included in the disparate data sources and may be formatted. The data structures are further formatted into model features that are configured to be input into machine learning models. Formatting data into model features may include several data manipulations. For example, data may be standardized or normalized to bring data with different scales into a similar scale. New data may be generated in appropriate scales using other techniques appropriate for machine learning models.

“Data standardization is the process of rescaling the features so that they will have the properties of a Gaussian distribution where the mean is equal to zero and the standard deviation is equal to one. Data normalization is the process of rescaling the features such that the range of the data is fixed. For example, the range can be fixed between zero and one, and normalized based on a sigmoid function. In other examples, the range can be fixed between zero and 10, and normalized based on a rectified linear unit function. While these standardization and normalization approaches are discussed, other approaches may be utilized.

“Formatted data and model features may be input into machine learning models. The machine learning models may be individually trained to specific diagnoses. The machine learning models receive the inputted, formatted data and produce output data. The output data includes a probability and/or confidence value that a medical patient may be diagnosed with one or more diagnoses. The output data, including the probability and/or confidence interval data, may then be sent by a first computing device executing a first application and received by a second computing device executing a second application. The computing device executing an application that receives the output data may also receive an indication that the medical patient with whom the output data is associated with will be seen by a medical service provider at a given time. The output data may then be surfaced and displayed as a notification including the potential diagnosis.”

The claims supplied by the inventors are:

“1. A system comprising: one or more processors; and non-transitory computer-readable media storing first computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: generating machine learning models configured to diagnose one or more diseases or conditions, wherein individual ones of machine learning models are trained to determine a likelihood that a disease or condition should be diagnosed for a medical patient; receiving data from multiple disparate sources via a computing network; formatting the data into model features configured to be input into the machine learning models, wherein the individual ones of the machine learning models are trained to receive the model features and output data indicating a probability that the medical patient should be diagnosed with the one or more diseases or conditions; inputting the model features into the machine learning models; generating, utilizing at least the machine learning models, output data indicating a potential diagnosis associated with the one or more diseases or conditions for the medical patient and a confidence value associated with that potential diagnosis; receiving, from a computing device executing an application, an indication that the medical patient will be seen by a medical service provider at a given time; receiving an indication that the medical patient is located at a location associated with the medical service provider at the given time; and sending, to the computing device and based at least in part on receiving the indication that the medical patient is located at the location associated with the medical service provider at the given time, a command configured to cause a device associated with the medical service provider to display a notification including the potential diagnosis.

“2. The system of claim 1, wherein the data includes at least medical records, chart codes, Centers for Medicare & Medicaid Services data, International Codes for Diagnosis data, medication data, and laboratory data.

“3. The system of claim 1, the operations further comprising: determining, based at least in part on machine learning techniques, that a combination of risk factors, which are derived at least in part from the data, is associated with one or more diseases or conditions; and the association exceeds a threshold for a confirmation of diagnosis.

“4. The system of claim 1, wherein the machine learning models include individual machine learning models for at least one of cancer, chronic kidney disease, heart disease, congestive heart failure, vascular disease, morbid obesity, or diabetes.

“5. A method comprising: generating machine learning models configured to diagnose one or more diseases or conditions for a medical patient; receiving data from multiple disparate sources via a computing network; formatting the data into model features configured to be input into machine learning models; inputting the model features into the machine learning models; generating, utilizing at least the machine learning models, output data indicating a potential diagnosis for the medical patient; assigning a confidence value to the output data indicating a diagnosis for the medical patient; receiving, from a computing device executing an application, an indication that the medical patient will be seen by a medical service provider at a given time; receiving an indication that the medical patient is located at a location associated with the medical service provider at the given time; and sending, to the computing device and based at least in part on receiving the indication that the medical patient is located at the location associated with the medical service provider at the given time, a command configured to cause a device associated with the medical service provider to display a notification including the potential diagnosis.

“6. The method of claim 5, wherein the user data includes at least medical records, chart codes, Centers for Medicare & Medicaid Services data, International Codes for Diagnosis data, medication data, and laboratory data.

“7. The method of claim 5, wherein the diagnoses of one or more diseases or conditions associated with the medical patient comprises: determining, based at least in part on machine learning techniques, that a combination of risk factors, which are derived at least in part from the data, is associated with one or more diseases or conditions; and the association exceeds a threshold for a confirmation of diagnosis.

“8. The method of claim 5, wherein the machine learning techniques are based, at least in part, on models trained and by disease groups, wherein the disease groups are at least one of cancer, chronic kidney disease, heart disease, congestive heart failure, vascular disease, morbid obesity, or diabetes.

“9. The method of claim 5, wherein feedback indicating the diagnosis was correct is inputted by a second user, wherein the feedback data is used to hone the model.

“10. The method of claim 5, further comprising: prioritizing the user data prior to training the machine learning models, based upon predefined criteria, wherein the predefined criteria includes, but is not limited to, at least one of documented International Classification of Disease codes, medication for singular disease, or laboratory values that define diagnosis.

“11. The method of claim 5, further comprising: determining an impact of a data type on the output data; determining that the impact satisfies a threshold impact; and prioritizing the data type.

“12. The method of claim 5, further comprising: receiving feedback data over a period of time; inputting feedback data into the machine learning models; receiving an indication of criteria; and updating the machine learning models to determine the diagnosis of one or more diseases or conditions based at least in part on the criteria.

“13. A system comprising: one or more processors; and non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: generating machine learning models configured to diagnose one or more diseases or conditions for a medical patient; receiving data from multiple disparate sources via a computing network; formatting the data into model features configured to be input into machine learning models; inputting the model features into the machine learning models; generating, utilizing at least the machine learning models, output data indicating a potential diagnosis for the medical patient; assigning a confidence value to the output data indicating a diagnosis for the medical patient; receiving, from a computing device executing an application, an indication that the medical patient will be seen by a medical service provider at a given time; receiving an indication that the medical patient is located at a location associated with the medical service provider at the given time; and sending, to the computing device and based at least in part on receiving the indication that the medical patient is located at the location associated with the medical service provider at the given time, a command configured to cause a device associated with the medical service provider to display a notification including the potential diagnosis.

“14. The system of claim 13, wherein the user data includes at least medical records, chart codes, Centers for Medicare & Medicaid Services data, International Codes for Diagnosis data, medication data, and laboratory data.

“15. The system of claim 13, wherein the diagnoses of one or more diseases or conditions associated with the medical patient comprises: determining, based at least in part on machine learning techniques, that a combination of risk factors, which are derived at least in part from the data, is associated with one or more diseases or conditions; and the association exceeds a threshold for a confirmation of diagnosis.

“16. The system of claim 13, wherein the machine learning techniques are based, at least in part, on models trained and by disease groups, wherein the disease groups are at least one of cancer, chronic kidney disease, heart disease, congestive heart failure, vascular disease, morbid obesity, or diabetes.

“17. The system of claim 13, wherein feedback indicating the diagnosis was correct is inputted by a second user, wherein the feedback data is used to hone the model.

“18. The system of claim 13, further comprising: prioritizing the user data prior to training the machine learning models, based upon predefined criteria, wherein the predefined criteria includes, but is not limited to, at least one of documented International Classification of Disease codes, medication for singular disease, or laboratory values that define diagnosis.

“19. The system of claim 13, further comprising: determining an impact of a data type on the output data; determining that the impact satisfies a threshold impact; and prioritizing the data type.

“20. The system of claim 13, further comprising: receiving feedback data over a period of time; inputting feedback data into the machine learning models; receiving an indication of criteria; and updating the machine learning models to determine the diagnosis of one or more diseases or conditions based at least in part on the criteria.”

For the URL and additional information on this patent, see: Goetz, Melanie. Machine learning models for diagnosis suspecting. U.S. Patent Number 11587678, filed March 23, 2020, and published online on February 21, 2023. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(11587678)&db=USPAT&type=ids

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