Autonomous MR Scanner Using AI for Raw Data Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current magnetic resonance (MR) imaging devices are limited by deterministic reconstruction methods that suppress information, require costly hardware, and are not optimized for information rate, leading to subjective decision-making and high equipment demands.

Innovation Solution

The implementation of autonomous MR scanning using artificial intelligence to analyze data and adapt scanning configurations, allowing for on-the-fly adjustments and reduced hardware requirements, enabling simplified MR scanners that can operate without human intervention and provide diagnostic outputs directly from raw data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If deterministic reconstruction methods are used in MR imaging, then image quality is maintained, but information is suppressed and hardware requirements increase

Engineering Contradiction:

Improveimage qualityVSAvoidinformation suppression

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent replaces deterministic mechanical reconstruction algorithms with machine learning-based analytical models that directly map raw MR data to diagnostic outputs, eliminating the need for traditional image reconstruction and preserving all information from the raw data

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces machine learning models as intermediaries between raw MR data and diagnostic conclusions, allowing the system to work directly with raw data without suppressing information through reconstruction, while still producing clinically useful outputs

Inventive Principle:

Principle #24Intermediary (Mediator)

2Manufacturing precision

If traditional MR imaging hardware with linear gradients and homogeneous fields is used, then reconstruction accuracy is improved, but device cost and complexity increase

Engineering Contradiction:

Improvereconstruction accuracyVSAvoidhardware requirements

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent changes the operating parameters of the MR system by using non-uniform main magnetic fields, non-homogeneous RF pulses, and non-linear gradients, which are traditionally avoided but become acceptable when paired with machine learning-based analysis that can handle the resulting data complexities

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent substitutes complex hardware requirements with software-based machine learning models that can compensate for hardware imperfections, allowing simplified MR devices to achieve diagnostic accuracy without requiring precision-engineered homogeneous fields and linear gradients

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If manual operation by technologists is used, then scan quality control is maintained, but automation and efficiency are reduced

Engineering Contradiction:

Improvescan quality controlVSAvoidmanual intervention

Core Design Contradiction:

ReliabilityVSExtent of automation

Solution Approach 1:

The patent enables the MR system to perform self-diagnosis and self-optimization by using machine learning models to automatically analyze raw data, determine diagnostic quality, and guide the scanning process without requiring manual intervention from technologists

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent implements feedback loops where machine learning models continuously analyze acquired data and provide real-time guidance on scan quality, automatically adjusting scanning parameters to maintain diagnostic reliability without human input

Inventive Principle:

Principle #23Feedback

4Ease of operation

If image reconstruction is performed before analysis, then visual interpretation is enabled, but processing time and information loss increase

Engineering Contradiction:

Improvevisual interpretationVSAvoidprocessing time

Core Design Contradiction:

Ease of operationVSLoss of time

Solution Approach 1:

The patent extracts the essential diagnostic information directly from raw MR data using machine learning models, eliminating the intermediate image reconstruction step and reducing processing time while maintaining diagnostic accuracy

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent substitutes the traditional visual interpretation workflow with automated machine learning analysis that operates directly on raw data, removing the time-consuming reconstruction step and enabling faster diagnostic output

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12102423B2Autonomous magnetic resonance scanning for a given medical test

Publication Date: 2024.10.01 SIEMENS HEALTHINEERS AG

AI summary

For autonomous MR scanning for a given medical test, a simplified MR scanner may be used without or will little input or control by a technologist (e.g., by a physician, radiologist, or person trained in MR scanner operation). The MR scanner autonomously positions, scans, checks quality, analyzes, and/or outputs an answer to a diagnostic question with or without an MR image. Scan analysis, based on artificial intelligence, allows for on-going or on-the-fly alteration of the scanning configuration to acquire the data desired to answer the diagnostic question. By using a simplified MR scanner, both position of the patient relative to the MR scanner and localization of the scan by the MR scanner are jointly solved. Sensors may sense a patient in a scan position where the reduced radio frequency requirements allow for a more open bore.