Gene Name Normalization via Prefix Tree Tokenization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current systems for recognizing and normalizing gene names in medical text face challenges due to their complex nomenclature and variability, leading to modest performance and dependency on external sources, with recall and precision typically reaching only the high 70's.

Innovation Solution

A method and system utilizing a prefix tree structure to identify and standardize gene names by tokenizing text, comparing tokens to a database of gene identifiers, and updating an identification pointer to determine canonical gene names, while handling variations and ambiguities through tokenization and data structure traversal.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If existing systems use complex naming rules, database entries, and machine-learning systems to recognize and normalize gene names, then they can handle the variety of nomenclature, but their recall and precision only reach the high 70's and they depend on external sources

Engineering Contradiction:

Improverecall and precision of gene name recognitionVSAvoidcomplexity of mechanisms and data required

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the gene identifier matching process into character-level tokens, comparing each character sequentially against the prefix tree structure. This breaks down the complex pattern matching into simple, manageable steps that can be executed efficiently without requiring complex machine learning systems.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a prefix tree data structure as an intermediary between the input text and the gene name database. This prefix tree serves as a mediator that pre-organizes gene identifiers in a hierarchical structure, enabling efficient matching without requiring complex external tools or multiple database queries.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If existing systems use multiple specifications, standards, and databases to handle gene nomenclature, then they can cover the range of possible mentions, but they require complex mechanisms and external sources

Engineering Contradiction:

Improveability to handle various gene nomenclatureVSAvoidnumber of external sources and tools required

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent creates a universal prefix tree data structure that can handle multiple gene nomenclature standards and variations within a single system. This unified structure replaces the need for multiple separate databases and specifications, providing adaptability to various gene naming conventions without requiring external tools.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent merges multiple gene identifier patterns and nomenclature variations into a single prefix tree structure. By combining diverse gene naming patterns into one unified data structure, the system achieves versatility in handling various gene mentions without relying on multiple external databases or tools.

Inventive Principle:

Principle #5Merging (Combining)

3Measurement precision

If the system uses a prefix tree structure with character-level token comparison, then it improves accuracy and efficiency in gene name recognition, but it requires precise tokenization and traversal logic

Engineering Contradiction:

Improveaccuracy of gene identifier matchingVSAvoidcomplexity of tokenization and traversal process

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The patent segments both the input text and the gene identifiers into character-level tokens. This fine-grained segmentation enables precise matching by comparing individual characters against the prefix tree structure, achieving high measurement precision in gene identifier recognition.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary tokenization of the input text into character units before the matching process. This preliminary action prepares the data in a format suitable for efficient prefix tree traversal, reducing the complexity of the subsequent matching operation by pre-organizing the input data.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11594303B2Method and system for normalization of gene names in medical text

Publication Date: 2023.02.28 KONINKLIJKE PHILIPS NV

AI summary

A method (100) for standardizing gene nomenclature, comprising: (i) receiving (110) a source; (ii) tokenizing (120) the source; (iii) comparing (130) a first token to a prefix tree structure with a root node, edges, and leaf nodes; (iv) determining (140) which edge extending from the root node to associated first leaf nodes the first token matches; (v) updating (150) an identification pointer with the location of the first leaf node; (vi) determining (160) which of one or more edges that a second token matches; (vii) updating (170) the identification pointer with the location of the second leaf node; (viii) repeating (172) the determining (160) and updating (170) steps with subsequent tokens until a subsequent token fails to match an edge extending from a leaf node or there is no edge extending from the leaf node; and (ix) providing (180) an identification of a canonical gene name.