Apple Inc.

Zhe Li

Christopher Daryl Soli

Kevin M Lynch

Pablo F Caro

Nicholas D Felton

Aroon Pahwa

Matthew Wagner Crowley

Evan Russell Doll

Paul W Salzman

Charmian Bondoc Naguit

Dmitri Cavander

Heather E Daniel

Jason B Morley

Patrick W Granvold

Mark E Pennell

David W Padgett

Jorge Fernando Pozas Trevino

Siji Rachel Tom

Ruchi Goswami

Christian Schroeder

Pratik Solanki

John P Clark

James D Dusseau

Brian A Carlsen

Jeffrey D LaMarche

Suhail Ahmad

Karim Benhmida

A user device may use a curated health sub-ontology to index a user's health record data sourced from multiple electronic health record (EHR) systems. As part of indexing the health record data, the user device may create a personal health ontology that is specific to the user's health record data.

Customized presentation of health record data

A user device can be used to generate medical term expressions, which represent medical terms of a health record. The user device can identify a medical concept present in the health record based on a medical term expression. The user device can generate a node in a personalized relational graph that corresponds to the medical concept. One or more sub-nodes can be added to the node. Responsive to a request, a user interface is presented that identifies the medical concept and some of the additional information.

Techniques for anonymized searching of medical providers

A user interface may be used to present a graph that represents observable values from health record data stored on a user device. A personal health ontology stored on the user device may be accessed. At least one observable value of the observable values may be determined to be of a type of data for which a filter exists by at least accessing the personal health ontology. A filter may be added to the least one observable value. The user interface including the graph may be presented and the filter may be applied to the at least one observable value.

Generation of customized personal health ontologies

A computer system may generate a curated health sub-ontology. This may include receiving a set of selection criteria corresponding to a curated health ontology. A subset of concept nodes may be determined using the set of selection criteria. Generating the curated health sub-ontology may include identifying the subset of concept nodes and combining the subset of concept nodes to define the curated health sub-ontology. The curated health sub-ontology may be transmitted to a consuming system.

A term base can be used to index a raw medical term (e.g., a text string from an electronic health record) to generate a computational expression of the raw medical term. The computational expression of the raw medical term can be used to generate a uniform medical term for the raw medical term. The computational expression can be compared with other computational expressions to organize the electronic health record, filter the electronic health record, and provide a uniform presentation of raw medical terms from different events in the electronic health record.

Normalization of medical terms

A computer system may generate a curated health ontology. The curated health ontology may be used by a requesting system. The curated health ontology generated from on a set of standard health codes combined with augmentation information, which supplements and/or otherwise changes the standard health codes. Once generated, the curated health ontology may be transmitted to the requesting system.

A term base can be used to index a raw term (e.g., a text string from an electronic health record) to generate a computational expression of the raw term. As part of doing so, portions of the raw term including personally identifying information can be removed and/or replaced with de-identified information. The computational expression of the raw term can be used to generate a version of the raw term that is devoid of the personally identifying information.

Index-based deidentification

Growing community of inventors

Fremont, CA, United States of America

Zhe Li