Cases

Laser speckle contrast imaging (LSCI) is a real-time, non-invasive method for visualizing tissue blood flow, but its clinical adoption is limited by poor image interpretability and artifacts, which this research aims to overcome using AI-based enhancement techniques.

Develop novel, robust, failure-aware and resource-efficient AI methods to automatically derive quantitative biomarkers from 3d ultrasounds to monitoring growth and development of the fetus.

Developing AI-based CT-derived FFR methods for non-invasively identifying patients requiring invasive coronary artery treatment.

We are investigating an alternative radiotherapy treatment workflow on the MR-Linac for high-risk prostate cancer patients.

We investigate how the image-guided radiotherapy workflow can be improved, in terms of efficiency and safety, by processing uncertainty estimations obtained during deep learning auto-contouring.

Ultrasound can transform thoracic imaging in low-resource settings by offering a practical alternative to radiography. It can expand access to essential diagnostics, improve patient outcomes, and lower costs. With the integration of AI, clinicians can benefit from real-time image analysis and guided support, helping overcome barriers to adoption and enabling more confident, accurate assessments at the point of care.

The project aims to develop AI models to automatically analyze echocardiographic images, supporting primary care physicians in triaging patients and determining the need for cardiac referral.

This PhD aims to develop an AI system that generates structured reports from CT scans to help radiologists work more efficiently amid a growing workforce shortage and rising cancer rates.