Mohamad Abou Ali Defends His PhD on Smarter, More Robust AI for Medical Imaging 🧠🤖

🎓 On May 2nd, 2025, our colleague Mohamad Abou Ali successfully defended his PhD thesis at the University of the Basque Country (UPV/EHU), wrapping up years of innovative research at the intersection of artificial intelligence and medical imaging.

Supervised by Dr. Fadi Dornaika and Dr. Ignacio Arganda-Carreras, Mohamad’s thesis tackled some of the toughest challenges in applying deep learning to healthcare: limited annotated data, class imbalance, and high variability in real-world medical images.

His work introduced three major contributions:

1. 🧠 Self-aware AI — Proposing that pre-trained models (CNNs, ViTs) can “understand” how well they’re learning and adapt accordingly. This line of work opens the door to more autonomous and reliable AI systems in clinical settings.

2. 🩺 ‘Naturalize’ Data Augmentation — A novel technique that generates realistic synthetic images to boost performance on underrepresented classes. Tested on blood cell and skin cancer datasets, it achieved significant gains in sensitivity and specificity.

3. 🧬 Robust AI for Brain Cancer MRI — A deep dive into making models resilient to real-world distortions like noise, blur, and patient motion in MRI scans, improving generalization across diverse clinical scenarios.

The defense committee was composed of:

• Dr. Jinan Charafeddine (Pôle Universitaire Léonard de Vinci, France)
• Dr. Abdelmalik Moujahid (Universidad Internacional de La Rioja, Spain)
• Dr. Blanca Cases (UPV/EHU)

Congratulations, Mohamad, on this outstanding milestone and the exciting research path ahead! 🥳👏