US 11,941,359 B2
Identifying anatomical phrases
Peter Prinsen, Eindhoven (NL); and Robbert Christiaan van Ommering, Cambridge, MA (US)
Assigned to KONINKLIJKE PHILIPS N.V., Eindhoven (NL)
Appl. No. 16/980,548
Filed by KONINKLIJKE PHILIPS N.V., Eindhoven (NL)
PCT Filed Mar. 11, 2019, PCT No. PCT/EP2019/055954
§ 371(c)(1), (2) Date Sep. 14, 2020,
PCT Pub. No. WO2019/175075, PCT Pub. Date Sep. 19, 2019.
Claims priority of provisional application 62/642,933, filed on Mar. 14, 2018.
Prior Publication US 2021/0004533 A1, Jan. 7, 2021
Int. Cl. G06F 40/289 (2020.01)
CPC G06F 40/289 (2020.01) 12 Claims
OG exemplary drawing
 
1. A method for identifying anatomical phrases in medical text, the method comprising:
accessing one or more data sources comprising medical information using a network interface, the medical information comprising medical text;
extracting, using a configured processor executing instructions stored on a memory, a set of concepts from a subset of at least one of the data sources;
identifying, using the configured processor, a plurality of terms associated with each of the concepts, wherein identifying the plurality of terms associated with each of the concepts includes identifying synonyms of each concept;
implementing a syntactic method by:
sorting, using the configured processor, the identified terms into one or more anatomy lists and one or more general lists, wherein the sorting includes initially sorting single term synonyms into anatomy-specific nouns or adjectives based on whether one of the concepts is a body structure or an anatomy-specific descriptor, then iteratively adding additional terms for further sorting into an anatomy-specific noun or adjective and a general noun or adjective, the general noun or adjective for use in describing an anatomy without denoting anything anatomical standing alone; and
defining a grammar on the one or more anatomy lists and the one or more general lists by applying, using the configured processor, at least one grammar rule to at least one term from the one or more anatomy lists and at least one term from the one or more general lists to generate at least one composite term; and
identifying, using the configured processor, the at least one composite term associated with the medical text accessed using the network interface.