This paper presents a novel corpus in Spanish tailored for the extraction of medical image-related entities from radiological reports using Named Entity Recognition (NER) methods. The dataset was created by aggregating and refining multiple existing corpora, focusing on entities that can be visually interpreted in associated medical images. This resource aims to bridge the gap between natural language processing and computer vision in the biomedical domain. The study evaluates various NER methods, including encoder-only, encoder-decoder, and decoder-only architectures. It explores fine-tuning, zero-shot, and few-shot In-Context Learning (ICL) strategies to determine the most effective approach for entity extraction. The resulting dataset is publicly available.