Registration Point Cloud Filtering for Anatomical Survey Oversampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing minimally invasive medical procedures face inaccuracies in registration due to oversampling of data points caused by disproportionate surveying of anatomical regions, leading to misalignment between real and model-based anatomical structures.
Innovation Solution
A system that analyzes sensor parameters and data point parameters in real-time, comparing them to thresholds to determine which data points to accept or reject, using methods such as motion collection, point distance rejection, point density rejection, and survey density normalization to mitigate oversampling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the medical device surveys anatomical regions by collecting data points along its path, then the registration between real and model-based anatomical structures can be established, but oversampling of data points in certain regions occurs causing misalignment and reduced registration accuracy
Solution Approach 1:
The patent changes the parameter of data point selection by introducing filtering criteria based on motion parameters (velocity, acceleration, jerk) and spatial parameters (distance to previous points, density thresholds). This transforms the raw data stream into a filtered set of representative points that maintain registration accuracy while reducing oversampling effects.
Solution Approach 2:
The patent extracts only the necessary and representative data points from the complete set of collected points by applying multiple filtering conditions. Points that do not meet the criteria (such as points collected during stationary periods or points too close to previous points) are removed, leaving only the essential data needed for accurate registration.
2Reliability
If the medical device collects continuous data points along the entire path through the anatomic passageway, then complete coverage of the anatomical structure is achieved, but computational resources are wasted processing redundant oversampled data
Solution Approach 1:
The patent extracts only the essential data points needed for reliable registration by applying filtering criteria that identify and remove redundant points. This extraction process reduces the computational burden while maintaining the reliability of the registration by preserving only the most informative data points.
Solution Approach 2:
The patent changes the data set parameter from complete continuous sampling to a filtered subset by applying motion-based and spatial-based selection criteria. This parameter change reduces computational energy requirements while maintaining registration reliability through intelligent data point selection.
3Adaptability or versatility
If the operator manually navigates the medical device through the anatomic passageway, then flexibility in surveying different regions is maintained, but inconsistent surveying patterns cause disproportionate sampling of certain anatomical regions
Solution Approach 1:
The patent introduces feedback mechanisms that monitor motion parameters and spatial distribution of collected points in real-time. Based on this feedback, the system dynamically adjusts which points are accepted or rejected, providing corrective action to compensate for inconsistent surveying patterns while maintaining operator flexibility.
Solution Approach 2:
The patent changes the acceptance criteria parameter dynamically based on observed surveying patterns. By adjusting thresholds for motion parameters and spatial distribution, the system compensates for inconsistent manual navigation while preserving the adaptability of operator-controlled surveying.
Data Source
AI summary
Disclosed are systems and methods for mitigating oversampling of data points collected by a medical device. In some aspects, a system is configured to receive data points of a sampled survey point cloud detected by a sensor of the medical device during surveying of an anatomic structure; determine, during the surveying, at least one parameter associated with (i) the medical device and/or (ii) the received data points detected by the sensor, including a change of translational and/or rotational motion of the medical device, a distance from a data point to a nearest neighbor within the sampled survey point cloud, or a density of the data points of a sub-set of the sampled survey point cloud corresponding to sub-region of the anatomic structure; analyze the parameter(s) by comparing it to a threshold; and record individual data points in a registration point cloud when the analyzed parameter(s) satisfies the respective threshold.