Apparatus and Method for Generating a Perfusion Image, and Method for Training an Artificial Neural Network Therefor

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention describes an apparatus and method for generating perfusion images (which show blood or fluid flow in tissue) from non-contrast medical imaging data, such as non-contrast MRI scans. It uses an artificial neural network (ANN), typically of the U-net type, trained to predict perfusion images comparable to those obtained with contrast agents, but without the need to actually administer contrast agents to patients. The method also details how to train such a neural network using datasets where conventional contrast-enhanced imaging was used as ground truth. The invention is particularly useful for tissues like the breast or prostate, where perfusion imaging is diagnostically important but contrast injection carries risks or is otherwise undesirable.

Use CasesContent extracted from patent full text and abstract with AI.

• Medical diagnostics where perfusion imaging is needed but contrast agent administration is risky or undesired (such as patients with kidney issues, allergies, or pregnant women).
• Breast cancer screening and diagnosis using non-contrast-based MRI perfusion analysis.
• Prostate diagnostics and other tissue perfusion assessments.
• Hospitals and clinics seeking to reduce costs and complexity of MRI workflows by eliminating the need for contrast-enhanced scans in certain cases.
• Situations where rapid perfusion information is required without waiting for contrast administration.

BenefitsContent extracted from patent full text and abstract with AI.

• Enables generation of perfusion images without requiring the use of contrast agents, reducing risk and patient discomfort.
• Reduces procedural complexity, time, and costs by omitting contrast administration and associated imaging steps.
• Allows perfusion imaging for patients who cannot safely receive contrast agents.
• Minimizes the risk of imaging artifacts that arise due to patient movement between scans or errors in image subtraction processes.
• Can be integrated into existing healthcare IT and MRI infrastructure, as it mainly requires software and computational resources.
• Supports broader, safer, and more efficient access to high-value diagnostic information.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The invention provides an apparatus and a method for generating a perfusion image, as well as a method for training an artificial neural network for use therein.The method comprises at least steps of:receiving (S100) at least one non-contrast medical diagnostic image, NCMDI (1-i), acquired from organic tissue;generating (S200), using an artificial neural network, ANN (2), trained and configured to receive input data (10) based on at least one of the received at least one non-contrast medical diagnostic image, NCMDI (1-i), based on the input data (10), at least a perfusion image (3) for the organic tissue shown in the at least one non-contrast medical diagnostic image, NCMDI (1-i); andoutputting (S300) at least the generated perfusion image (3).