Internship - AI-Based Cardiac CT Interpretation

1w1 week ago

Philips

Hamburg, DE · Full-time · €24,000 – €36,000

About this role

Coronary artery segmentation is an essential procedure in computer-aided diagnosis of coronary artery disease. It enables further downstream tasks supporting CAD-RADS reporting. As a master student, develop and evaluate innovative approaches for coronary artery segmentation from 3D CT data.

Familiarize with the topic through state-of-the-art literature review and available tooling. Implement extensions for data curation, training environment, and analysis tools. Develop, implement, and evaluate algorithms with numerical experimentation on local cluster infrastructure.

Engage in weekly interactions with supervisor for tight feedback loops and presentation of intermediate results. Work in an office-based role requiring at least 3 days in-person per week. Collaborate within Philips' health technology environment focused on quality healthcare access.

Write thesis and prepare final presentation. Pursue potential publication to high-tier conferences like MICCAI, MIDL, SPIE, or ISBI. Contribute to meaningful work helping lives through advanced medical imaging solutions.

Requirements

Student in computer science, informatics, mathematics, or physics
Knowledge in medical image processing
Very good presentation and visualization skills
Comprehensive AI background and hands-on experience with recent NN architectures
Excellent programming skills (Python, PyTorch; C++ is a strong plus)
Proficient in English language

Responsibilities

Familiarize yourself with the topic and state-of-the-art literature review
Familiarize yourself with available tooling and implement required extensions for data curation, training environment, and analysis tool preparation
Develop, implement, and evaluate algorithm(s) for coronary artery segmentation including numerical experimentation on local cluster
Iterate training, analysis, visualization, and refinement processes
Write thesis and prepare final presentation
Engage in weekly interaction with supervisor including presentation of intermediate results
Potentially publish to high-tier international conferences (MICCAI/MIDL/SPIE/ISBI)