Fine-Tuning Foundation Models for X-Ray Image Segmentation – MT Intro Talk by Maeen Abdelbadea Nasralla Alikarrar

Join us for an introductory presentation of a Master’s thesis focused on adapting foundation models for X-ray image segmentation. By fine-tuning the Segment Anything Model (SAM) using Low-Rank Adaptation (LoRA), this study aims to enhance its performance on X-ray images, addressing the domain gap between natural and medical imaging. The research will benchmark the fine-tuned model against other medical segmentation models like MedSAM to evaluate potential benefits of fine-tuning, exploring its ability to deliver efficient and generalizable solutions for X-ray image analysis.