Echocardiogram Workflow Pattern Analysis for Disease Prediction

Solution Overview

Problem

Current methods for disease prediction in echocardiography are computationally expensive, time-consuming, and do not account for the exam in a holistic manner, lacking diagnostic insights from automatic intelligent summarization.

Innovation Solution

A method that receives a medical workflow from a sequence of echocardiogram modalities and applies condition-indication rules to determine the likelihood of medical conditions using machine logic, without analyzing the actual content of the exam, employing a dictionary of subflows and Support Vector Machines to predict diseases with high accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If detailed analysis of various medical exams and tests is performed for disease prediction, then diagnostic accuracy is improved, but computational cost and time consumption increase

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts and analyzes only the workflow metadata (sequence of modalities, measurements taken, views examined) from the complete medical exam data, separating the diagnostic workflow information from the actual medical images and detailed test results. This extraction allows disease prediction based on workflow patterns alone, achieving 75% accuracy without analyzing the full exam content, thus reducing time and computational resources while maintaining diagnostic value

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the medical exam data into distinct workflow components (modality sequence, measurement types, viewing patterns) that can be independently analyzed. By dividing the complex exam data into these discrete workflow elements, the system can process and predict diseases from workflow patterns separately from the actual medical content, reducing overall processing time and computational burden

Inventive Principle:

Principle #1Segmentation

2Measurement precision

If detailed analysis of various medical exams and tests is performed for disease prediction, then diagnostic accuracy is improved, but computational resources required increase

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidcomputational cost

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent extracts and analyzes only the workflow metadata (sequence of modalities, measurements taken, views examined) from the complete medical exam data, separating the diagnostic workflow information from the actual medical images and detailed test results. This extraction allows disease prediction based on workflow patterns alone, achieving 75% accuracy without analyzing the full exam content, thus reducing time and computational resources while maintaining diagnostic value

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of analyzing the actual medical exam content (images, measurements, test results) to predict diseases, the patent inverts the approach by analyzing the workflow metadata (how the exam was conducted, sequence of modalities, patterns of examination) to make disease predictions. This inversion reduces computational requirements significantly while maintaining diagnostic accuracy

Inventive Principle:

Principle #13The other way round (Inversion)

3Adaptability or versatility

If conventional workflow analysis is used to identify deviation from standard workflow, then workflow standardization is achieved, but diagnostic insights are lost

Engineering Contradiction:

Improveworkflow standardizationVSAvoiddiagnostic insights

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

Instead of using workflow analysis merely to detect deviations from standard protocols, the patent inverts the approach by using workflow patterns as the primary source for disease prediction. The system learns disease-specific workflow patterns from training data and uses these patterns to predict diseases in new cases, transforming workflow analysis from a compliance tool into a diagnostic tool that provides actionable medical insights

Inventive Principle:

Principle #13The other way round (Inversion)

Solution Approach 2:

The patent changes the analytical parameters from checking workflow compliance (standard vs. actual) to analyzing workflow patterns as diagnostic features. By transforming the workflow data representation and applying machine learning models, the system extracts diagnostic information from workflow patterns that would be lost in conventional standardization approaches

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10327712B2Prediction of diseases based on analysis of medical exam and/or test workflow

Publication Date: 2019.06.25 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Use of medical workflows where a first medical workflow is obtained from a plurality of medical acts performed in sequence that related to care of a patient. A set of condition-indication rules is applied to the first medical workflow to determine first condition information. The first condition information relates to a likelihood that a first medical condition exists in the patient.