3D multimodal cardiac data reconstruction using angiography and computerized tomographic angiography registration.

Area of Science:

• Medical Imaging
• Cardiovascular Diagnostics
• Computational Anatomy

Background:

• Computerized tomographic angiography (CTA) and X-ray angiography (XA) are key tools for diagnosing coronary artery disease.
• Current analysis of CTA and XA data is performed separately, hindering specialists' ability to integrate information.
• A method is needed to correlate findings between CTA and XA for enhanced diagnostic capabilities.

Purpose of the Study:

• To develop a method for registering CTA and XA data to create a unified 3D model of coronary artery stenosis.
• To assist medical specialists by providing a visual relationship between coronary arteries and stenosis from both imaging modalities.

Main Methods:

• Coronary artery segmentation and labeling are performed for both CTA and XA datasets.
• Specialists mark stenosis and relevant arteries in XA images.
• Control points are automatically detected and normalized for marked arteries in both modalities.
• A geometrical transformation function is computed and applied to register XA stenosis data onto the CTA 3D model.

Main Results:

• A 3D model is generated, integrating stenosis and coronary artery information from both CTA and XA.
• The method's effectiveness in segmentation, labeling, and 3D reconstruction is validated on a dual-modality dataset.

Conclusions:

• The method visually links corresponding coronary arteries and stenosis locations between CTA and XA.
• It provides 3D positions of stenosis lesions on CTA, aiding diagnosis.
• The technique requires no changes to existing medical acquisition standards or devices, ensuring broad applicability.