Surgical Image Registration Using Search Images and Depth Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for determining a transformation rule between images acquired by a robotic visualization system and preoperative volumetric image data are time-consuming and prone to errors, often requiring manual alignment and additional hardware, such as physical markers or stereoscopic image data alignment, which is not always available.
Innovation Solution
A method that automatically and efficiently determines the transformation rule by controlling a robotic visualization system to capture search images, analyze the target region's arrangement, and acquire depth-resolution measurement data, eliminating the need for manual alignment and additional hardware, using a computer-implemented method with optional machine-learned algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical markers are attached to the patient for registration, then the transformation rule can be determined between preoperative image data and operating room images, but the process becomes error-prone and time-consuming
Solution Approach 1:
The patent extracts the registration function from physical markers and implements it through automated image processing. The system captures images of anatomical landmarks using a robotic visualization system and automatically identifies corresponding points in preoperative images through image analysis algorithms, eliminating the need for manual marker attachment while maintaining registration accuracy.
Solution Approach 2:
The system performs self-registration by automatically capturing images, identifying anatomical landmarks, and computing the transformation rule without human intervention. The automated image processing pipeline independently completes the registration task that previously required manual marker placement and manual point correspondence identification.
2Measurement precision
If physical markers are attached to the patient, then registration can be achieved, but additional steps are required both for preoperative imaging and before the actual operation
Solution Approach 1:
The patent removes physical markers from the registration process entirely. Instead, it uses automated identification of anatomical landmarks directly visible in the images. This single-step image-based approach replaces the multi-step process of attaching markers before preoperative imaging and then again before surgery.
Solution Approach 2:
The patent combines the registration functionality into the image acquisition and processing workflow itself. The same robotic visualization system that captures surgical images also performs the registration by identifying anatomical landmarks, merging what were previously separate processes (marker attachment and registration computation) into a unified automated system.
3Ease of manufacture
If markerless approach using surface geometry is used, then no additional hardware is needed, but additional hardware and time are still required for pointer instrument operation
Solution Approach 1:
The patent replaces the mechanical pointer instrument with an automated robotic visualization system. Instead of manually touching anatomical points with a pointer, the system uses automated image capture and computer vision algorithms to identify and locate anatomical landmarks, substituting mechanical interaction with optical and computational methods.
Solution Approach 2:
The system automatically performs landmark identification and registration without requiring manual surface scanning. The image processing algorithms independently detect anatomical features and compute their coordinates, eliminating the need for operator intervention in the surface scanning process.
4Quantity of substance
If stereoscopic image data from robotic visualization system is used, then comprehensive surface capture is possible, but manual alignment is required which is time-consuming and error-prone
Solution Approach 1:
The system performs automated alignment by computing the transformation rule through image-based landmark identification. The robotic visualization system automatically processes the stereoscopic images, identifies corresponding anatomical landmarks, and calculates the registration transformation without requiring manual alignment operations.
Solution Approach 2:
The patent replaces manual alignment operations with automated computational methods. Instead of manually positioning and aligning the robotic system, software algorithms automatically compute the spatial transformation between the captured images and preoperative images based on identified anatomical landmarks.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Various examples relate to techniques for recording, in the context of surgery carried out on a patient, measured data with depth resolution so that a transformation rule can be defined based on these data which mediates between images recorded by means of a robotic visualization system, a surgical microscope, for example, and pre-surgery volume image data.