TAVI-TEC: AI Tool for Automated TAVI Planning
A team of researchers has introduced TAVI-TEC, an entirely automated AI-driven system designed for preprocedural planning of transcatheter aortic valve implantation (TAVI). This innovative tool is incorporated into a web-based DICOM viewer and analyzes computed tomography angiography (CTA) scans from patients receiving SAPIEN 3 Ultra (S3U) prostheses. Utilizing deep learning, the system segments cardiovascular structures, detects calcifications, extracts centerlines, identifies landmarks, and defines the annular plane. It also measures critical annular and aortic root dimensions and produces color-coded maps indicating lumen reduction and vessel diameter for assessing vascular access. The study, published on arXiv (2607.29243), highlights the growing number of TAVI procedures and emphasizes the necessity for enhanced efficiency and standardized clinical annotations.
Key facts
- TAVI-TEC is a fully automated AI-based framework for TAVI planning.
- It is integrated into a web-based DICOM viewer.
- Processes CTA scans from patients with SAPIEN 3 Ultra (S3U) prostheses.
- Uses deep learning for segmentation, calcification detection, centerline extraction, and landmark identification.
- Quantifies annular and aortic root measurements.
- Generates color-coded maps for lumen reduction and vessel diameter.
- Study published on arXiv with ID 2607.29243.
- Aims to improve efficiency and standardize annotations in TAVI procedures.
Entities
Institutions
- arXiv