System and method for prioritization and presentation of heterogeneous medical data
A system and method prioritizes and presents heterogenous medical data. The method includes retrieving medical data of a patient, the medical data including data from multiple data sources. The method includes phenotyping the medical data to generate a reasoning trail and reasoning outcome including one or more codes, the reasoning trail including a basis for which the reasoning outcome is determined. The method includes assigning a relevance score to each of the one or more codes. The method includes ranking the codes based on the relevance score of each of the one or more codes. The method includes displaying the codes in rank order.
1 . A computer-implemented method, comprising:
retrieving by a processor medical data of a patient, the medical data including data from multiple data sources;
phenotyping, using a machine learning model, by the processor the medical data of the patient and an ontology and medical concepts from the medical data to generate a reasoning trail and a reasoning outcome including one or more medical codes and concepts and converting the one or more medical codes and concepts to a predetermined medical ontology, the reasoning trail including a basis for which the reasoning outcome is determined;
assigning by the processor a relevance score to each of the one or more medical codes and concepts, wherein assigning the relevance score includes applying a weighing factor based on a presence of a second medical code within the one or more medical codes included in the reasoning outcome;
ranking by the processor the medical codes and concepts based on the relevance score assigned to each of the one or more medical codes and concepts; and
displaying the medical codes and concepts in rank order.
2 . The computer-implemented method of claim 1 , further comprising:
mapping the medical codes from one ontology to another, in response to at least one of the medical codes being in an ontology that is different from an ontology of another medical code.
3 . The computer-implemented method of claim 1 , wherein a subset of the relevance scores of the medical codes are conditionally dependent on the presence of other medical codes, and wherein the method further comprises:
filtering the ranked medical codes wherein only the medical codes having a relevance score above a predetermined threshold value are displayed.
4 . The computer-implemented method of claim 1 , wherein the multiple data sources include one of lab results, problem lists, medication lists, radiology reports, pathology reports, operation reports, and admission and discharge notes.
5 . The computer-implemented method of claim 1 , wherein the weighing factor is based on a lookup table including relevant combinations of medical codes; and
wherein the lookup table has a maximum size of N 2 , where N is the number of medical codes in the background ontology.
6 . The computer-implemented method of claim 1 , wherein the one or more medical codes are based on medical ontology comprising Systematic Nomenclature of Medicine and International Classification of Diseases.
7 . The computer-implemented method of claim 1 , further comprising receiving a user input based on the displayed list of medical codes.
8 . The computer-implemented method of claim 7 , wherein the user input selects one of the displayed medical codes to view one of the data sources associated with the medical code and the reasoning trail associated with the medical code.
9 . The computer-implemented method of claim 7 , wherein the user input filters the displayed medical codes to view medical codes from a selected data source.
10 . The computer-implemented method of claim 1 , wherein assigning the relevance score to each of the one or more medical codes includes determining a relevance of each of the one or more medical codes relative to a context of a workflow of a user.
11 . A non-transitory computer readable storage medium storing an executable program having instructions executable by a processor to:
retrieve medical data of a patient, the medical data including data from multiple data sources;
phenotype, using a machine learning model, the medical data of the patient and an ontology and medical concepts from the medical data to generate a reasoning trail and reasoning outcome including one or more medical codes and concepts and converting the one or more medical codes and concepts to a predetermined medical ontology, the reasoning trail including a basis for which the reasoning outcome is determined;
assign a relevance score to each of the one or more medical codes and concepts, wherein assigning the relevance score includes applying a weighing factor based on a presence of a second medical code within the one or more medical codes included in the reasoning outcome;
rank the medical codes and concepts based on the relevance score assigned to each of the one or more medical codes and concepts; and
display the medical codes and concepts in rank order.
12 . The non-transitory computer readable storage medium of claim 11 , wherein the processor is caused to map medical codes onto a medical vocabulary or ontology from one ontology to another, in response to at least one of the medical codes being in an ontology that is different from an ontology of another medical code.
13 . The non-transitory computer readable storage medium of claim 11 , wherein the processor executes the executable program to cause the processor to:
extract medical concepts from the medical data of the patient including free-text documents by processing natural language.
14 . The non-transitory computer readable storage medium of claim 13 , wherein the medical concepts are extracted from data sources by processing natural language with sentence boundary detection and phenotyping includes analyzing the extracted medical concepts.
15 . The non-transitory computer readable storage medium of claim 11 , wherein the processor executes the executable program to cause the processor to:
filter the ranked medical codes, by selecting only the medical codes having a relevance score above a predetermined threshold value to be displayed.
16 . The non-transitory computer readable storage medium of claim 11 , wherein the processor executes the executable program to cause the processor to:
receive a user input based on the displayed list of medical codes.
17 . The non-transitory computer readable storage medium of claim 16 , wherein the user input is one of (a) selecting one of the displayed medical codes to view one of the data sources associated with the medical code and the reasoning trail for the medical code and (b) filtering the displayed medical codes to view medical codes associated with a selected data source.
18 . The non-transitory computer readable storage medium of claim 11 , wherein assigning the relevance score to each of the one or more medical codes includes determining a relevance of each of the one or more medical codes relative to a context of a workflow of a user.