Machine Learning Predicate Device Retrieval for Medical Certification

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Manufacturers face challenges in discovering suitable or similar predicate medical devices for regulatory certification due to a disparity between professional knowledge and the current medical device classification architecture.

Innovation Solution

A method and computer device that utilize machine-learning models to extract technical items from a specific medical device's technical context, generate candidate medical devices, search databases for device information, and determine the most similar medical device based on similarity scores calculated from technical contexts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manufacturers manually search for predicate medical devices using current classification architecture, then they can identify similar devices, but the process requires excessive time and professional expertise

Engineering Contradiction:

Improveaccuracy of finding similar devicesVSAvoidtime for regulatory certification

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical search processes with an automated machine learning system. The ML model automatically extracts technical items from device contexts, generates candidate predicates, and calculates similarity scores, eliminating the need for manual professional review while improving accuracy and speed of finding similar medical devices for regulatory certification

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces a machine learning model as an intermediary between the device context and predicate database. This intermediary automatically processes technical contexts, extracts relevant items, generates candidate predicates, and ranks them by similarity, bridging the gap between device specifications and the classification architecture without requiring manual professional intervention

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If manufacturers use current medical device classification architecture to find predicate devices, then they can identify similar devices, but the process becomes overly complex

Engineering Contradiction:

Improveaccuracy of finding similar devicesVSAvoidcomplexity of search process

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential technical items from device contexts using machine learning, rather than requiring manual analysis of entire device specifications and classification architectures. This extraction process simplifies the search by focusing on key technical features that determine device similarity, reducing complexity while maintaining accuracy

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces complex manual classification and comparison processes with an automated ML system that handles technical item extraction, candidate generation, and similarity calculation, significantly reducing the complexity of the predicate device search process while improving precision

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If manufacturers rely on professional knowledge to identify predicate devices, then they can find suitable candidates, but the process is difficult to standardize and scale

Engineering Contradiction:

Improveaccuracy of finding similar devicesVSAvoidefficiency of device retrieval

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent enables the system to perform predicate device identification autonomously without requiring human professional knowledge. The ML model self-learns from device contexts, automatically extracts technical items, generates candidate predicates, and ranks them by similarity, making the process scalable and reproducible while maintaining high accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent substitutes human professional expertise with an automated ML system that processes device contexts and identifies predicate devices consistently and scalably. This replacement eliminates variability in manual analysis while improving productivity through automated high-throughput processing of multiple device candidates

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20250140392A1Method for retrieving information about similar medical devices and computer device using the same

Publication Date: 2025.05.01 NATIONAL APPLIED RESEARCH LABORATORIES

AI summary

A method for retrieving information about similar medical devices is provided, which includes the following steps: obtaining a first technical context of a specific medical device; utilizing a first machine-learning model to extract one or more technical items of the specific medical device based on technical content of the first technical context; utilizing the first machine-learning model to generate candidate medical devices using the technical items; searching a database for device information about the candidate medical devices; retrieving summary files of the candidate medical devices from the database based on the device information; utilizing the first machine-learning model to infer a second technical context of each candidate medical device; and determining a most similar medical device for the specific medical device according to a similarity score for each candidate medical device calculated from the second technical context and the first technical context using a second machine-learning model.