AI for real time intravascular optical coherence tomography segmentation and characterization

Developing robust and trustworthy algorithms for real time analysis of intravascular optical coherence tomography images

ONGOING

2023

 

 | 

2027

 

RESEARCH AREAS

About project

OCT, an emerging imaging modality, stands as a beacon of innovation. Its unique capabilities enable the capture of highly detailed images of coronary arteries, delving deep into the intricate layers of blood vessels to unveil subtleties that would otherwise might remain hidden. These capabilities extend beyond mere plaque identification, providing an in-depth analysis that facilitates a comprehensive understanding of vascular health. Nevertheless, the integration of OCT into routine clinical practice presents formidable challenges. Analyzing an entire OCT pullback proves to be a time-consuming endeavor, especially when it entails characterizing lesions and extracting quantifiable data. Furthermore, the interpretation of OCT findings often lacks consensus among medical specialists, leading to uncertainties in diagnosis and treatment decisions. The core objective of this project is to harness the capabilities of deep learning to address these challenges. The envisioned outcome is the expedited analysis of OCT pullbacks through artificial intelligence. This innovation aims to streamline the diagnostic process, providing healthcare professionals with a reliable and easily interpretable solution. The integration of AI technology holds the promise of standardizing OCT procedures, making them more efficient, dependable, and ultimately more accessible to those in need. This collaborative effort, driven by a shared vision, emphasizes the transformative potential of technology in the realm of cardiac care. It is a story of collective endeavor, where the project itself, rather than any individual, takes center stage, bearing the hopes of revolutionizing medical practice and improving the lives of patients suffering from coronary artery disease.

 

Our publications & software