AI in Medical Imaging

Empowering Cardiac and Aorta Analysis with AI in Medical Imaging Analysis

In the rapidly evolving landscape of healthcare, artificial intelligence (AI) stands as a beacon of innovation, particularly in the field of medical imaging. Our initiative focuses on harnessing the power of AI to revolutionize the analysis of the heart and aorta on CT images, fostering a collaborative environment where scientists, engineers, and medical professionals converge to redefine medical imaging analysis with novel biomarkers.

The heart and aorta are central to cardiovascular health, where early detection and accurate assessment of diseases can dramatically alter outcomes. Traditional imaging analysis techniques, while effective, often require extensive time for manual analysis and are subject to observer variance as well as errors. Our project leverages AI algorithms designed to analyze complex imaging data, identify subtle anomalies, and predict cardiovascular disease.

Objectives and Clinical Impact

• To Develop Advanced AI Algorithms: Create and refine deep-learning based AI models that can accurately analyze medical images of the heart and aorta, detecting conditions in patients with cardiovascular disease and abnormal morphology.
• To Foster Interdisciplinary Collaboration: Establish a robust framework for ongoing cooperation between engineers and medical professionals, ensuring that AI developments are grounded in clinical needs and experiences.
• To Accelerate Diagnostic Processes: Reduce the time required for image analysis, enabling faster diagnosis and the initiation of treatment plans e.g. pre-procedural planning with invasive cardiologists.

The essence of our projects lies in the synergy between engineering innovation and medical expertise. We are pioneering a collaborative platform where engineers work alongside cardiologists and radiologists to tailor deep-learning based AI models specifically for heart and aorta analyses. Many more organ-specific AI algorithms are in the pipeline of our research projects. This cross-disciplinary approach ensures that our AI solutions are not only technologically advanced but also clinically relevant, enhancing diagnostic processes and facilitating accurate segmentation of organs.

In AI News

• Bjørn and Mathias from DTU have sucessfully defended their master’s thesis in “Mathematical Modelling and Computation” and are set to start their PhD programme as part of the ARTICHOKE project by Spring 2024.
• Yasmin (medical student researcher) and Teodora (engineering student researcher) have been brainstorming ways to accurately segment the left and right atrial and ventricular volumes, respectively.
• Rasmus (DTU professor) held a presentation regarding “AI Driven Medical Image Visualization” for the workshop: “Human Perception and Advanced Visualization of 3D Medical Imaging Data“.
• Rasmus has also been on the news in national Danish television, where he spoke about AI as part of education and exams for university students at DTU.