Couch-Mounted Depth Camera Collision Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current collision detection systems in radiation therapy, such as gantry mounted laser guards and room-based scanners, are inadequate as they fail to effectively detect collisions when the patient support is rotated or moved, particularly with complex dose delivery plans involving non-coplanar beams.

Innovation Solution

A method and apparatus using depth sensing cameras to generate reference and input depth images, which are processed to create a composite image, allowing for real-time detection of potential collisions between medical device parts and patients, irrespective of couch movement, by determining if image data values exceed thresholds within defined intrusion zones.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a gantry mounted laser guard system is used to detect collisions, then collision detection capability is provided, but the system fails to work when the couch is rotated

Engineering Contradiction:

Improvecollision detection reliabilityVSAvoidadaptability to couch rotation

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The collision detection system is mounted on the patient support (couch) rather than the gantry, allowing it to move dynamically with the couch during rotation and positioning. This ensures the sensor remains in a fixed position relative to the patient throughout the treatment process, maintaining detection capability during couch movements.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent replaces traditional laser-based mechanical scanning systems with depth sensing camera technology that uses optical fields and image processing. This substitution enables the system to capture three-dimensional depth information and detect collisions through computational analysis of depth image differences, rather than relying on mechanical laser scanning that fails during couch rotation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If a room-based scanning system is used, then collision profiles can be created, but the model must be updated with every couch movement and scanning may be blocked from certain angles

Engineering Contradiction:

Improvecollision profile accuracyVSAvoidmodel update complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The depth sensing camera mounted on the couch performs self-adjustment by capturing depth images from its fixed position relative to the patient. The system automatically processes depth image differences to detect collisions, eliminating the need for manual model updates or complex recalibration when the couch moves. The system serves itself by continuously monitoring the treatment area without requiring external intervention.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system creates a dynamic copy of the treatment area using depth sensing technology. By continuously capturing and comparing depth images, the system generates real-time representations of the treatment zone, replacing the need for static pre-scan models that require updating. The depth image difference method creates an instantaneous copy of spatial relationships for collision detection.

Inventive Principle:

Principle #26Copying

3Reliability

If complex dose delivery plans with non-coplanar beams are used, then treatment effectiveness is improved, but the chance of collisions increases

Engineering Contradiction:

Improvetreatment effectivenessVSAvoidcollision risk

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The system provides continuous real-time feedback during radiation delivery by monitoring depth image differences between reference and current images. When a potential collision is detected, the system can immediately alert operators or automatically adjust device positions, enabling corrective action before actual contact occurs. This closed-loop feedback mechanism allows complex non-coplanar beam delivery while maintaining safety through continuous monitoring.

Inventive Principle:

Principle #23Feedback

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This solution provides reliable and dynamic collision detection, enabling the generation of warning signals or control signals to prevent collisions, thus ensuring patient safety during radiation therapy procedures.

Implementation Method 1

the reference depth image is generated based on a time-of-flight technique

Methodology Applied

Scientific EffectTime of flight: Time of Flight

Implementation Method 2

the reference depth image is generated using infrared illumination

Methodology Applied

Scientific EffectInfrared radiation: Infrared Radiation

Data Source

PatentEP3014956B1Systems and methods for detecting a possible collision between an object and a patient in a medical procedure

Publication Date: 2024.09.25 VARIAN MEDICAL SYSTEMS INC

Figure 1A
Figure 1B~1C
Figure 2

AI summary

A method of detecting a possible collision in a medical procedure, includes: obtaining a reference depth image; obtaining an input depth image; determining a composite image using at least a part of the reference depth image and at least a part of the input depth image, wherein the act of determining the composite image is performed using a processing unit; and determining whether there is a possible collision between an object and a patient based on the composite image.