Clinical Data Inverted Index for Complex Query Retrieval

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current database systems are inadequate for handling complex queries and nested searches, particularly in clinical information retrieval, due to their design for read and write access rather than search functionality.

Innovation Solution

A system and method for storing structured and unstructured clinical information by transforming data into episode model data, generating inverted index documents, and employing specialized ranking and clustering algorithms to facilitate efficient querying and retrieval.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If clinical information is stored in relational or column store databases optimized for read and write access, then data storage and access efficiency is improved, but search functionality and complex query performance deteriorates

Engineering Contradiction:

Improvedata access efficiencyVSAvoidsearch functionality

Core Design Contradiction:

ProductivityVSEase of operation

Solution Approach 1:

The patent segments clinical information into structured data (stored in relational databases for efficient read/write access) and unstructured data (stored in inverted index documents for efficient searching). This segmentation allows each data type to be optimized for its primary access pattern, resolving the contradiction between access efficiency and search functionality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary layer that includes both a relational database and an inverted index system. The relational database handles structured data access while the inverted index handles search operations, acting as a mediator that provides both efficient access and robust search capabilities without compromising either function.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Device complexity

If data is stored in traditional database formats, then data storage simplicity is improved, but retrieval speed for complex queries deteriorates

Engineering Contradiction:

Improvedata storage simplicityVSAvoidretrieval speed

Core Design Contradiction:

Device complexityVSSpeed

Solution Approach 1:

The patent performs preliminary action by pre-processing clinical information into inverted index documents during data ingestion. This pre-processing creates optimized search structures in advance, enabling rapid retrieval of complex queries without compromising storage simplicity, as the inverted indexes are automatically generated and maintained by the system.

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If nested searches are performed using SQL queries, then query capability is improved, but query performance and execution speed deteriorates

Engineering Contradiction:

Improvequery capabilityVSAvoidquery execution speed

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent substitutes the mechanical SQL query system with an inverted index-based search system. Instead of using traditional SQL nested queries that require complex joins and table scans, the system uses inverted indexes that enable direct term-based retrieval, dramatically improving query execution speed while maintaining versatile query capabilities through the index structure.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS10445378B2Storing structured and unstructured clinical information for information retrieval

Publication Date: 2019.10.15 THE CLEVELAND CLINIC FOUND

AI summary

A computer-implemented method can include acquiring data from at least one data source, the acquired data including health data for a patient. The acquired data can be transformed into episode model data according to a context-specific data model and the episode model data can be stored in a database. The method also includes generating at least one inverted index document for at least a portion of an episode for the patient based on the episode model data.