Systems and methods for optimizing clinical workflow using a prioritization engine
Inventors
Limoli, Craig • Larkin, D. Justin • Lee, Ryan
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11031111-B1
Publication Date
2021-06-08
Expiration Date
Abstract
The disclosed subject matter is directed to systems and methods for processing multiple sources of data into a clinical data prioritization and visualization framework that enhances and expedites clinical workflow.
Core Innovation
A computer-implemented self-enhancing clinical data prioritization and visualization system retrieves information relating to a clinical context of a patient visit from a plurality of information sources. A combiner module supplies the retrieved information to a prioritization module to generate priority estimates for different clinical information, and the priority estimates determine placement of a plurality of visual blocks representing clinical data points within a context-driven interface.
The prioritization module performs experience-driven statistical inference to generate priority estimates and uses machine interpretation of clinical guidelines to adjust the priority estimates for different clinical information. The adjustment is performed by applying a plurality of document-driven rules to adjust the priority estimates generated by the experience-driven statistical inference, and the guideline interpretation refines the ordering of displayed content.
The system encodes features of retrieved information relating to the clinical context into binary features input into a machine learning model. The machine learning model generates priority estimates of observed data points, is iteratively adjusted using a loss calculated by comparing the priority estimates against a priority index comprising historical physician behaviors and patient outcomes, and after iterative adjustment is complete, the prioritization module determines priority estimates for observed and unobserved data points in real time to be used for machine interpretation of clinical guidelines.
Claims Coverage
The independent claims are clm-00001 and clm-00012. They cover retrieval of clinical-context information, experience-driven statistical inference producing priority estimates, machine interpretation of clinical guidelines using document-driven rules to adjust priority estimates, binary-feature encoding with iterative model adjustment against a priority index, real-time priority estimates for observed and unobserved data points, and display of prioritized visual blocks in a customizable context-driven interface.
Clinical-context information retrieval for a patient visit
A combiner module retrieves information relating to a clinical context of a patient visit from a plurality of information sources.
Experience-driven statistical inference for priority estimates
The prioritization module performs experience-driven statistical inference to generate priority estimates for different clinical information, and the priority estimates determine placement of a plurality of visual blocks representing clinical data points.
Machine interpretation of clinical guidelines to adjust priority estimates
The prioritization module adjusts the priority estimates by performing machine interpretation of clinical guidelines and applying a plurality of document-driven rules to the priority estimates generated by the experience-driven statistical inference.
Binary-feature encoding and iterative loss-based model adjustment against a priority index
The prioritization module encodes features of the retrieved information into binary features for input into a machine learning model, generates priority estimates of observed data points, and iteratively adjusts the model using a loss calculated by comparing the priority estimates against a priority index comprising historical physician behaviors and patient outcomes.
Real-time priority for observed and unobserved data points for guideline interpretation
After iterative adjustment is complete, the system determines priority estimates for observed and unobserved data points in real time to be used for the machine interpretation of clinical guidelines.
Context-driven interface display with prioritized and customizable visual blocks
The display module generates a context-driven interface comprising a plurality of visual blocks arranged in a prioritized order based on the adjusted priority estimates, and the interface is customized by selecting and rearranging the visual blocks.
Overall, the independent claims require retrieval of clinical-context information, experience-driven statistical inference with binary-feature encoding and iterative loss-based model adjustment against a historical physician behavior and patient outcome priority index, document-driven rules for machine interpretation of clinical guidelines, real-time priority estimates for observed and unobserved data points, and a customizable context-driven interface with prioritized visual blocks.
Stated Advantages
Documented Applications
No documented applications found
Interested in licensing this patent?