ECG Similarity Search Using Vector Embeddings and Subject Metadata

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing medical database systems struggle to efficiently compare and match complex data types such as ECG signals, medical images, and unstructured clinical notes, hindering the integration of diverse subject data for clinical decision-making and personalizing healthcare.

Innovation Solution

A method and system that utilize vector embeddings of ECG data, combined with subject metadata, to identify similar subjects by comparing ECG data and metadata, leveraging advanced machine learning techniques and vector databases for efficient and accurate subject matching.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional database systems are used to store and compare medical records, then data storage is simple and straightforward, but the ability to efficiently compare and match complex data types such as ECG signals and medical images is limited

Engineering Contradiction:
Improveefficiency of comparing and matching complex medical dataVSAvoidcomplexity of database system for handling complex data types
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces vector embeddings as an intermediary representation layer between complex medical data (ECG signals, medical images, clinical notes) and the database system. These vector embeddings transform complex, non-standard data types into standardized numerical vectors that can be efficiently stored, searched, and compared using traditional database operations, thereby enabling efficient matching without requiring complex specialized database infrastructure

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent transforms complex medical data into a different parameter space by converting ECG signals, images, and text into vector embeddings with specific dimensional parameters. This parameter transformation allows the system to leverage efficient vector similarity search algorithms instead of complex pattern matching operations, significantly improving comparison efficiency while using standard database systems

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If expert cardiologists interpret ECG data manually, then diagnostic accuracy is high, but the process is time-consuming and requires significant expertise

Engineering Contradiction:
Improveaccuracy of ECG interpretationVSAvoidtime required for ECG interpretation
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent implements automated diagnostic algorithms that enable ECG data to be interpreted without requiring expert cardiologist involvement for every case. The system uses vector embeddings and similarity search to automatically compare patient ECG data with historical cases, providing diagnostic insights and treatment recommendations that would traditionally require expert review, thereby reducing time loss while maintaining diagnostic quality

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent incorporates feedback mechanisms where the automated system learns from historical medical records and expert diagnoses. By continuously comparing new ECG cases with historical data and refining vector embeddings based on outcomes, the system improves its diagnostic accuracy over time, approaching expert-level performance while maintaining rapid automated processing

Inventive Principle:
Principle #23Feedback

3Productivity

If automated diagnostic algorithms are used to assist ECG interpretation, then processing speed increases, but the ability to account for individual subject characteristics and comorbidities is limited

Engineering Contradiction:
Improvespeed of ECG analysisVSAvoidaccuracy of treatment recommendations
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent merges multiple data sources including ECG signals, subject metadata (age, comorbidities, drug allergies), and historical treatment outcomes into a unified vector embedding representation. This combination allows the automated algorithm to simultaneously process ECG data and individual subject characteristics, generating treatment recommendations that are both rapid and personalized to each patient's unique profile

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent creates composite vector embeddings that integrate heterogeneous data types (physiological signals, demographic information, clinical history) into a unified representation. This composite approach enables the system to consider multiple factors influencing treatment decisions while maintaining computational efficiency, thereby improving reliability of automated recommendations without sacrificing processing speed

Inventive Principle:
Principle #40Composite materials

Data Source

PatentEP4687147A1Medical database searching system
Publication Date: 2026.02.04 KONINKLIJKE PHILIPS NV
  • EP4687147A1 patent drawingFigure 1
  • EP4687147A1 patent drawingFigure 2
  • EP4687147A1 patent drawingFigure 3A~3B

AI summary

The present disclosure provides concepts for searching a medical database comprising electrocardiogram (ECG) data. The method includes obtaining a database comprising a plurality of subject entries, each corresponding to a historic subject and comprising recorded metadata describing characteristics of the respective subject and a recorded vector embedding representing ECG data of the respective subject. A query vector embedding representing ECG data of a query subject is generated with an ECG encoder. The recorded metadata of the plurality of subject entries is compared with query metadata describing characteristics of the query subject, and the recorded vector embeddings are compared with the query vector embedding. One or more similar subject entries are identified based on a result of the comparison. Accordingly, the invention provides a searching means that is designed to take into account characteristics of subjects to identify subjects with similar cardiological conditions, thereby reducing a burden on a clinician whilst reducing a rate of misdiagnosis. For example, this may be particularly useful for triaging subjects.