Machine Learning MR Imaging Analysis for Prostate Cancer Localization

Solution Overview

Problem

Current methods for analyzing magnetic resonance imaging (MRI) data for prostate cancer are time-consuming and suffer from significant inter-observer variability, failing to fully utilize the embedded information beyond human perception.

Innovation Solution

A computer-implemented method using a trained machine learning model to process MR imaging data, generating location data representative of cancer probability, and producing a human-readable image indicating the likelihood of cancer at specific locations in the region of interest.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional qualitative analysis by radiologists is used, then human interpretation and diagnosis can be obtained, but the analysis is time-consuming and suffers from inter-observer variability

Engineering Contradiction:
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces the mechanical system of human radiologist analysis with an automated computer-based image processing system that applies consistent algorithms to all images, eliminating inter-observer variability while maintaining diagnostic accuracy and reducing analysis time

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Loss of information

If traditional qualitative analysis is used, then radiologist expertise is utilized, but important information beyond human vision or perception is under-utilized

Engineering Contradiction:
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent transforms the analysis from human visual perception to computational analysis by converting medical images into three-dimensional volumetric data and applying radiomics feature extraction, enabling the system to detect patterns and information that are beyond human visual or perceptual capabilities

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Reliability

If multiple MRI images are analyzed qualitatively, then comprehensive cancer detection is attempted, but the process remains time-consuming and variable

Engineering Contradiction:
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent merges multiple MRI images into a single integrated three-dimensional volumetric representation, allowing simultaneous analysis of all images through a unified computational framework that improves detection reliability while eliminating the sequential analysis time required by traditional methods

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS20250292407A1Processing magnetic resonance imaging data
Publication Date: 2025.09.18
  • US20250292407A1 patent drawing
  • US20250292407A1 patent drawing
  • US20250292407A1 patent drawing

AI summary

A computer-implemented method of processing magnetic resonance (MR) imaging data comprises receiving MR imaging data for a region of interest in a body of a human or animal subject, inputting the MR imaging data to a trained machine learning model, operating the trained machine learning model to generate location data representative of a probability of cancer at a location in the region of interest, and processing the location data to generate a human-readable image of the region of interest indicative of the probability of cancer at the location.