AI Clinical Extraction Tool for Automated Patient Data Structuring

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The manual extraction and abstraction of unstructured patient data into structured medical records for applications like cancer registries is laborious, slow, costly, and error-prone, hindering the efficiency and accuracy of healthcare data analysis and research.

Innovation Solution

An AI-assisted clinical extraction tool using machine learning and natural language processing is employed to automate the conversion of unstructured patient data from various sources into structured data records, including rule-based extraction and continuous adaptation based on new data for improved error detection and correction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual extraction and abstraction processes are used to convert unstructured patient data into structured medical records, then data can be organized and mapped to defined fields, but the process becomes laborious, slow, costly, and error-prone

Engineering Contradiction:

Improvedata accuracyVSAvoiddata processing speed

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent replaces the manual mechanical abstraction process with an automated computational system using natural language processing (NLP) and machine learning algorithms. The system automatically extracts data elements from unstructured text, classifies them into categories, and maps them to structured fields, eliminating the need for human abstractors while maintaining or improving data accuracy through consistent application of classification rules

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service by allowing the unstructured patient data to be automatically processed and transformed into structured records without human intervention. The NLP system independently performs extraction, classification, and mapping tasks that were previously requiring manual effort, making the data transformation process autonomous and scalable

Inventive Principle:

Principle #25Self-service

2Ease of manufacture

If manual abstraction processes are used to structure patient data, then data can be organized into structured records, but the process becomes laborious and costly

Engineering Contradiction:

Improvedata structuring capabilityVSAvoidprocessing efficiency

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The patent replaces manual data structuring operations with automated NLP-based extraction and classification systems. The system automatically identifies data elements in unstructured text, determines their categories, and maps them to appropriate structured fields, achieving the same structuring capability without human labor and at much higher efficiency

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system provides universal data structuring capability that can handle multiple types of unstructured medical documents (narrative notes, discharge summaries, operative reports, etc.) and transform them into standardized structured records. The same NLP pipeline and classification framework work across different document types and data sources, making the process efficient and scalable

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Loss of information

If manual extraction processes are used to create structured medical records, then data can be organized and mapped, but errors increase and timeliness decreases

Engineering Contradiction:

Improvedata completenessVSAvoidrecord generation time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The patent replaces manual extraction processes with automated NLP systems that continuously process unstructured patient data in real-time or near-real-time. The system extracts data elements, classifies them, and maps them to structured records automatically, dramatically reducing the time from data generation to structured record availability while maintaining complete data capture through comprehensive text analysis

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables continuous processing of patient data as it becomes available, rather than batch processing or delayed manual abstraction. The NLP pipeline operates continuously to extract and structure data from incoming medical documents, ensuring timely availability of structured records for clinical decision-making and research without interruption or delay

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS20220044812A1Automated generation of structured patient data record

Publication Date: 2022.02.10 ROCHE MOLECULAR SYSTEMS INC

AI summary

In one example, a method of extracting patient information for a medical application comprises: receiving patient data of a patient; processing the patient data using a learning system with Artificial Intelligence (AI)-assisted clinical extraction tool, the processing comprising: extracting, based on a trained language extraction model that reflects language semantics and a user's prior habit of entering other patient data, data elements from the patient data and data categories represented by the data elements, and mapping at least some of the extracted data elements to pre-determined data representations based on the data categories; populating fields of a data record of the patient based on the pre-determined data representations; and storing the populated data record in a database accessible by the medical application.