Medical Record NLP Classification Using Hierarchical Disease Terminology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing natural language processing (NLP) techniques struggle to reliably capture nuanced and varied descriptions of disease mechanisms and conditions in medical records, leading to inconsistent and incomplete classification of diseases, which hinders effective clinical decision-making.
Innovation Solution
A hierarchical rule-based language processing system using a curated dictionary and learned terminology to analyze patient data, employing NLP models that include tokenization and classification parameters to generate accurate disease classifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing NLP techniques are used to classify diseases in medical records, then the system is simple to implement, but the classification accuracy and reliability are insufficient due to inability to capture nuanced disease mechanism descriptions
Solution Approach 1:
The patent segments the disease classification task into multiple hierarchical levels: (1) tokenization of medical text into discrete units, (2) matching tokens against a structured terminology dictionary organized by disease categories, (3) calculating category-specific scores based on token matches, and (4) generating final classification recommendations. This segmentation allows the system to systematically capture nuanced disease descriptions while maintaining manageable complexity through modular processing stages.
Solution Approach 2:
The patent implements preliminary action by pre-curating a comprehensive terminology dictionary that contains disease categories, descriptors, and associated scores before the classification process begins. This pre-processed knowledge base enables the system to efficiently match and classify disease mechanisms without requiring complex real-time reasoning, thereby improving reliability while controlling system complexity.
2Loss of information
If a comprehensive disease classification system is implemented to capture all disease mechanisms, then the classification completeness improves, but the processing time and computational resources increase
Solution Approach 1:
The patent applies local quality by organizing the terminology dictionary into distinct disease categories with category-specific descriptors and scores. Instead of treating all disease information uniformly, the system selectively applies relevant category-specific criteria to each disease type. This allows comprehensive classification coverage while optimizing processing efficiency by focusing computational resources on category-relevant features rather than analyzing all possible descriptors for every case.
Solution Approach 2:
The patent utilizes parameter changes by calculating category-specific scores based on the presence and relevance of matched descriptors, then transforming these scores into classification recommendations. The system dynamically adjusts classification thresholds and weighting parameters based on category-specific characteristics, enabling efficient processing while maintaining comprehensive classification accuracy through adaptive parameter optimization.
3Measurement precision
If a detailed terminology dictionary with multiple descriptors is used, then the measurement precision of disease classification improves, but the device complexity and implementation difficulty increase
Solution Approach 1:
The patent introduces a hierarchical dimension to the terminology structure, organizing descriptors into multiple levels: general disease categories, specific disease mechanisms, and individual descriptors with associated scores. This dimensional organization allows the system to achieve high measurement precision through multi-layered matching while simplifying implementation by providing a structured framework that guides the development and maintenance of the terminology dictionary.
Data Source
AI summary
Systems and methods of the present disclosure enable improved natural language processing of patient-related medical information for clinical decision support. To do so, a processor receives patient data including a written report, and accesses a dictionary of terminology associated with a disease. The terminology includes descriptors indicative of categories of the disease. The processor inputs the written report into a tokenization function to output tokens by parsing word patterns in the written report, and generating the tokens from the word patterns. The processor determines a presence in the written report of each descriptor based on the tokens and determines a category-specific score associated with each category based on the presence of the descriptors. The processor determines a category recommendation score indicative of a particular category based on the category-specific scores and generates a category recommendation representing the particular category based on the category recommendation score.