Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12100517-B2
Publication Date
2024-09-24
Expiration Date
Abstract
A model-assisted system for identifying candidates for a cohort based on a biomarker may include at least one processor. The processor may be programmed to access a database from which information associated with a population of individuals can be derived; provide, to a generalized biomarker model, a first biomarker associated with a cohort, the generalized biomarker model being trained based on one or more second biomarkers using the information, wherein the first biomarker is different from the one or more second biomarkers; receive, from the generalized biomarker model, a first output indicating a first group of the population of individuals exceeding a first likelihood threshold of having been tested for the first biomarker; and determine, based on the first output, whether an individual from among the first group of the population of individuals is a candidate for the cohort.
Core Innovation
The invention provides a model-assisted system and computer-implemented method that identify candidates for a cohort using a biomarker. The system accesses a database from which information associated with a population of individuals can be derived, and provides a first biomarker associated with a cohort to a generalized biomarker model and to a biomarker specific model. The generalized biomarker model is trained based on one or more second biomarkers using the information, and the biomarker specific model is trained based on the first biomarker using the information, wherein the first biomarker is different from the one or more second biomarkers.
The system receives, from the generalized biomarker model, a first output indicating a first group of the population of individuals exceeding a first likelihood threshold of having been tested for the first biomarker. Based on the first output, the system determines whether an individual from among the first group is a candidate for the cohort. The system also receives, from the biomarker specific model, a third output indicating a third group of the population exceeding the first likelihood threshold of having been tested for the first biomarker.
To validate the approach, the system verifies the accuracy of the generalized biomarker model by comparing the first output to the third output. The document further describes a generalized biomarker model system for cohort candidate identification that can process population medical records and train a generalized biomarker model using one or more second biomarkers tokenized from text and feature-vectorized from structured and unstructured data, optionally via OCR. The model output is used to form a likelihood-based group of individuals exceeding a threshold for having been tested for the first biomarker, and candidate membership can be determined and optionally verified using individual medical records.
Claims Coverage
The partial document provides three independent claims, each centered on training a generalized model using different biomarkers or characteristics, producing likelihood-threshold group outputs, identifying candidate cohort membership, and verifying accuracy by comparing outputs from the generalized model and a biomarker- or characteristic-specific model. Across the independent claims, the core inventive flow contains four main inventive features.
Accessing population-associated information for cohort candidate identification
Access a database from which information associated with a population of individuals can be derived, and provide the derived information to a generalized biomarker model and to a biomarker specific model for identifying candidates for a cohort.
Training generalized and biomarker-specific models using different biomarkers
Provide, to a generalized biomarker model and to a biomarker specific model, a first biomarker associated with a cohort, wherein the generalized biomarker model is trained based on one or more second biomarkers using the information and the biomarker specific model is trained based on the first biomarker using the information, wherein the first biomarker is different from the one or more second biomarkers.
Using likelihood-threshold generalized output to determine candidate membership
Receive, from the generalized biomarker model, a first output indicating a first group of the population of individuals exceeding a first likelihood threshold of having been tested for the first biomarker, and determine, based on the first output, whether an individual from among the first group is a candidate for the cohort.
Verifying generalized-model accuracy by comparing to biomarker-specific output
Receive, from the biomarker specific model, a third output indicating a third group of the population of individuals exceeding the first likelihood threshold of having been tested for the first biomarker, and verify the accuracy of the generalized biomarker model by comparing the first output to the third output.
Generalizing the model-assisted workflow to characteristics and characteristic-specific accuracy comparison
Provide, to a generalized model and to a characteristic specific model, a first characteristic associated with a cohort, wherein the generalized model is trained based on one or more second characteristics using the information and the characteristic specific model is trained based on the first characteristic using the information, wherein the first characteristic is different from the one or more second characteristics, receive outputs based on a first likelihood threshold of being associated with the first characteristic, determine candidate cohort membership based on the generalized model output, and verify accuracy by comparing the first output to the third output from the characteristic specific model.
Across the independent claims, the document covers a model-assisted cohort candidate identification workflow that trains a generalized model on one or more second biomarkers or characteristics different from the cohort biomarker or characteristic, generates likelihood-threshold group outputs from the generalized model to determine candidate membership, and verifies accuracy by comparing the generalized-model output to an output from a biomarker- or characteristic-specific model.
Stated Advantages
Verifying the accuracy of the generalized biomarker model by comparing the first output to the third output.
Documented Applications
Identifying candidates for a cohort based on a biomarker using population medical records accessed from a database and a generalized biomarker model trained using one or more second biomarkers.
Cohort candidate identification by applying a generalized biomarker model to produce a likelihood-based group of individuals exceeding a threshold for having been tested for a first biomarker, with optional verification using individual medical records.
Interested in licensing this patent?