Frame-Based Cephalometric Image Stitching to Reduce Chatter Marks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing extraoral x-ray systems with linear scan technology suffer from chatter marks due to mechanical and motor-related deviations, leading to reduced image quality and diagnostic suitability of cephalometric images, and existing solutions either increase patient dose or are expensive.
Innovation Solution
A computer-implemented method using anatomical structures or x-ray opaque markers to adjust the position of projection images during reconstruction, reducing chatter marks by detecting markers in the images and applying a shift & add technique to align them with ideal positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If linear scan systems with secondary aperture and ceph detector are used, then x-ray dose is reduced, but mechanical deviations cause chatter marks that reduce image quality
Solution Approach 1:
The patent applies preliminary action by detecting the real position of anatomical structures or markers in projection images before final image reconstruction. The system calculates compensation values for mechanical deviations and pre-adjusts the positioning data, so that when the cephalometric image is reconstructed, the chatter marks are already corrected. This allows the use of cost-effective frame-based sensors with secondary apertures while maintaining image quality.
2Duration of action of moving object
If TDI-based linear scan systems are used, then integration time is optimized, but mechanical and motor deviations cause disturbances that lead to chatter marks
Solution Approach 1:
The patent implements feedback by continuously monitoring the real position of anatomical structures or markers in each projection image during the linear scan acquisition. The system compares the detected positions with the ideal positions and calculates compensation values that are fed back into the reconstruction process. This feedback mechanism corrects mechanical and motor deviations in real-time, maintaining acquisition stability and eliminating chatter marks while preserving optimized integration time.
3Manufacturing precision
If one-shot ceph sensors are used, then motion artifacts are prevented, but the sensors are expensive and bulky
Solution Approach 1:
The patent replaces the mechanical solution of using expensive one-shot ceph sensors with a computational approach. Instead of relying on a single expensive sensor that captures the entire image in one shot, the system uses multiple inexpensive frame-based sensors that capture sequential projection images. Software-based reconstruction and compensation algorithms then combine these images to eliminate motion artifacts, substituting mechanical complexity with computational processing.
4Productivity
If linear scan speed is increased to compensate for mechanical deviations, then acquisition time is reduced, but image quality deteriorates
Solution Approach 1:
The patent applies preliminary action by detecting and compensating for mechanical deviations in each projection image before final reconstruction. This allows the system to maintain optimal scan speeds without sacrificing image quality, as the compensation for mechanical deviations is already built into the image data during acquisition rather than requiring slower scanning to prevent deviations in the first place.
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
Improves image quality and diagnostic capability of cephalometric images by reducing chatter marks without increasing mechanical demands, allowing for cost-effective and safer imaging with frame-based sensors.
Implementation Method 1
a sequence of projection images is acquired by exposing a patient's head with x-rays from an x-ray source (2)
Data Source
Figure 1
Figure 2~3
Figure 4A~4C
AI summary
The present invention relates to a computer-implemented method of generating at least one section (A) of a cephalometric image (23) of a patient's head, comprising: a step of acquiring a sequence of two-dimensional x-ray projection images (22) by exposing the patient's head with x-rays from an x-ray source (2) while linearly moving an x-ray ceph detector (6) and an x-ray ceph collimator (8) on opposing sides along the patient's head, wherein the x-ray ceph collimator (8) has an aperture (10), and wherein the projection images (22) are sequentially read out at a predetermined rate from a predetermined frame (7) of the x-ray ceph detector (6); a step of detecting at least one anatomical structure (24) appearing in a group of consecutively acquired projection images (22) that correspond to said section of the cephalometric image (23) to be generated; a step of stitching the said group of consecutive projection images (22) such that the said anatomical structure (24) are overlapped to generate said section of the cephalometric image (23).