Sofie Van Hoecke is an Associate Professor at Ghent University, affiliated with the Internet Technology and Data Science Lab (IDLab) - imec. She leads the PreDiCT (Predictive Diagnostics through Contextual and Trustworthy AI) team, which develops innovative machine learning solutions for predictive maintenance and healthcare diagnostics. Her educational background includes: Engineering Degree, Ghent University (2003) PhD in Computer Science Engineering, Department of Information Technology, Ghent University (thesis: 'Efficient service management in healthcare') Van Hoecke's research centers on hybrid AI systems that fuse machine learning with semantic technologies and physical modeling. Her work targets predictive maintenance (e.g., condition monitoring via thermal imaging and vibration analysis) and predictive healthcare (e.g., migraine tracking, COPD management, and infection diagnostics). Key specialties include context-aware modeling, expert-driven machine learning, and dynamic semantic dashboards for real-time decision support. She has pioneered applications in industrial IoT, medical diagnostics, and environmental monitoring. Her 2025 publications reveal a strong interdisciplinary trajectory applying knowledge graphs, conformal prediction, and explainable AI to diverse domains including network security, digital pathology, building inspection, and neurology. A unifying theme is trustworthy AI with emphasis on uncertainty quantification, out-of-distribution robustness, and seamless integration of domain knowledge into ML pipelines. Her work bridges theoretical advances with industrial deployment through imec collaborations. Scientific Awards: No scientific awards mentioned in provided sources Van Hoecke directs the PreDiCT team across multiple EU and industry-funded projects including AI-SWEEP-2 (smart wound care), HEROI2C (ICU infection management), mBrain (migraine monitoring), and SmartWaterConnect 5.0 (water grid analytics). While specific student advisees aren't listed, her leadership of a 15+ member research team and graduate teaching duties indicate active mentorship. Current grants focus on deploying context-aware AI in healthcare (S.E.P.S.I.S. Connect, ADAM) and industrial settings (RE-ENNOVATE, PACSOI). The PreDiCT team operates within IDLab's Ghent University-imec ecosystem, maintaining strong industry partnerships with Siemens, UZ Ghent Hospital, and water management authorities. Their research infrastructure includes specialized testbeds for predictive maintenance validation and clinical trial frameworks for digital health applications. Ongoing work emphasizes real-world validation of hybrid AI models in resource-constrained environments, particularly for neglected tropical diseases and chronic condition management.
