Adaptive Multi-Resolution DRR Generation for Radiation Therapy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for generating digitally reconstructed radiographs (DRRs) in radiology are processing-intensive and memory-consuming, making them inefficient for manual matching in radiation therapy, especially on machines without high-powered GPUs.

Innovation Solution

A system that dynamically selects algorithms and generates adaptive multi-resolution DRRs, allowing users to interactively adjust projections for efficient rendering, reducing processing and memory requirements without compromising image effectiveness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional methods are used to generate DRRs, then image quality is maintained, but processing time and memory consumption increase significantly

Engineering Contradiction:

Improveimage qualityVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the DRR generation process into multiple resolution levels (low-resolution initial generation, mid-resolution refinement, and high-resolution final output). Each resolution level processes different portions of the image with appropriate detail, avoiding the need to process the entire image at maximum resolution from the start, thus reducing overall processing time while maintaining final image quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by generating only the necessary portions of the DRR at high resolution (regions of interest) while using lower resolution for other areas. This selective processing reduces the total computational workload and memory requirements while ensuring that critical diagnostic regions maintain high image quality.

Inventive Principle:

Principle #16Partial or excessive action

2Measurement precision

If conventional methods are used to generate DRRs, then image accuracy is maintained, but memory consumption increases

Engineering Contradiction:

Improveimage accuracyVSAvoidmemory consumption

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent implements a nested multi-resolution structure where low-resolution DRRs are generated first and serve as the base layer. Mid-resolution refinements are then nested within specific regions, and high-resolution details are nested within the most critical areas. This nested approach allows the system to maintain accurate high-resolution image data only where needed while using lower-resolution data elsewhere, significantly reducing overall memory consumption.

Inventive Principle:

Principle #7Nested doll (Nesting)

3Productivity

If high-powered GPUs are used, then rendering speed improves, but device complexity and cost increase

Engineering Contradiction:

Improverendering speedVSAvoidhardware requirements

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent implements dynamic algorithm selection that adapts to the available hardware capabilities. The system can dynamically adjust the rendering pipeline, selecting from multiple algorithms with varying computational requirements. This allows the same system to achieve efficient rendering on both high-powered GPUs and less sophisticated hardware, eliminating the mandatory requirement for complex high-powered GPU hardware while maintaining good rendering speed.

Inventive Principle:

Principle #15Dynamics

4Measurement precision

If full-resolution DRRs are generated, then diagnostic accuracy is maintained, but processing intensity increases

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidprocessing intensity

Core Design Contradiction:

Measurement precisionVSPower

Solution Approach 1:

The patent applies local quality by assigning different resolution levels to different regions of the DRR based on their diagnostic importance. Critical anatomical structures and regions of interest are rendered at high resolution to maintain diagnostic accuracy, while less critical areas are rendered at lower resolution. This regional differentiation reduces overall processing intensity while preserving diagnostic accuracy in the most important areas.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS10062186B2Method for dynamically generating an adaptive multi-resolution image from algorithms selected based on user input

Publication Date: 2018.08.28 SIEMENS HEALTHINEERS INTERNATIONAL AG

AI summary

Methods and systems are proposed herein for generative adaptive, multi-resolution images efficiently without intensive processing and/or memory consumption or hardware requirements. According to one aspect of the claimed subject matter, a system is provided that includes a computing workstation, communicatively coupled to both a data storage device and an image acquisition device. Real time images acquired by the image acquisition device are presented to the user along with one or more digitally reconstructed radiographs (DRRs)—generated using dynamically selected rendering techniques—from previously acquired image data. The user is able to verify the DRRs as a match to the verification image, and subsequently to dynamically generate additional DRRs more suitable by actuating a portion of the generated DRR. Based on the user actuation, a new DRR is generated and presented to the user for verification.