Geo-Registered 3D DBT Visualization for Microcalcification Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D mammography techniques, such as digital breast tomosynthesis (DBT), face challenges in reliably discerning the 3D distribution of microcalcifications within the breast, which are indicative of ductal carcinoma in situ (DCIS) or benign, due to their appearance as clusters in two-dimensional arrays, making it difficult for radiologists to distinguish suspicious branching structures.
Innovation Solution
The implementation of a geo-registered 3D DBT volume viewed through true stereoscopic imaging and augmented/virtual reality, allowing rotation, zooming, and interaction with the data using a computer system, control device, and AR/VR headset, along with geo-registration markers and a focal point pen, to enhance the understanding of microcalcifications and tumorous tissue structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 2D image slices are used to display DBT data, then the imaging process is simple and equipment requirements are low, but the ability to discern 3D distribution of microcalcifications is poor
Solution Approach 1:
The patent transitions from 2D image slice display to true stereoscopic 3D visualization of DBT data. By rendering the volumetric breast tissue data in three dimensions with proper depth perception, the system enables radiologists to discern the spatial distribution and branching structures of microcalcifications that are invisible or ambiguous in 2D representations.
2Reliability
If multiple compression configurations are used to image the breast, then diagnostic information is improved, but examination time and patient discomfort increase
Solution Approach 1:
The system performs geo-registration of anatomical landmarks and establishes a 3D coordinate framework before acquiring images from multiple compression configurations. This preliminary setup enables automated alignment and fusion of images from different views, reducing the need for manual registration and minimizing the time radiologists spend processing multiple configurations while maintaining comprehensive diagnostic coverage.
3Reliability
If the number of microcalcifications is large (single digit to 1000+), then the statistical significance improves, but the difficulty of identifying suspicious branching structures increases
Solution Approach 1:
The patent introduces automated computational algorithms as intermediaries between the raw DBT image data and radiologist interpretation. These algorithms detect, segment, and characterize microcalcifications, organizing them by spatial location, size, density, and morphological features. The system highlights suspicious branching patterns and presents processed information to radiologists, reducing the cognitive load of manually analyzing hundreds of individual calcifications while maintaining high detection confidence.
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 approach enables radiologists to gain a thorough understanding of the underlying structure of microcalcifications and tumorous tissue, reducing false positives and improving diagnostic accuracy by providing a more accurate 3D representation of breast tissue, thereby enhancing the detection of potentially malignant structures.
Implementation Method 1
images with differing gray scales depending on the density of the tissue attenuated by the X-ray beam onto the detector array
Data Source
AI summary
Digital breast tomosynthesis represents an enhanced type of mammogram for detecting breast cancer. In this disclosure, data from digital breast tomosynthesis is reconstructed into a volumetric database with each voxel having a (x, y, z) coordinate and viewed in true 3D using geo-registered head display unit and geo-registered tools for overall enhanced diagnosis. The breast is imaged under various configurations and the internal architecture of an anatomic feature three-dimensionally analyzed. Additional dataset creation and three-dimensional imaging techniques are disclosed.