The field of medical image analysis and radiology is rapidly advancing, with a focus on developing innovative solutions to improve clinical diagnosis and treatment planning. Recent research has emphasized the importance of integrating medical visual question answering into radiology workflows, highlighting the need for more effective evaluation metrics and clinical relevance. Another key area of development is the use of radiomics features for medical image retrieval, enabling flexible querying and improved retrieval specificity. The incorporation of eye gaze data and video representations of radiologists' gaze has also shown promise in enhancing the performance of large vision-language models in chest X-ray analysis. Furthermore, there have been significant advancements in centerline tracking, radiology report generation, and vascular geometry synthesis, with a focus on developing more accurate and scalable models. Noteworthy papers in this area include:

• RadiomicsRetrieval, which proposes a customizable framework for medical image retrieval using radiomics features.
• RadEyeVideo, which enhances general-domain large vision language models for chest X-ray analysis with video representations of eye gaze.
• Trexplorer Super, which improves centerline tracking in 3D medical images.
• SISRNet, which generates clinically accurate radiology reports by focusing on salient regions with medically critical characteristics.
• HUG-VAS, which synthesizes realistic aortic geometries using a hierarchical NURBS-based generative model.