Brief

Researchers have proposed a new reference architecture for structured radiology reporting that links evidence directly to report content. This human-supervised system aims to extract and organize structured information, such as measurements and lesion identities, which are often lost in free text or fragmented across various systems. The framework integrates exam-specific templates, AI-assisted drafting, and interoperability standards to support reviewed reporting, longitudinal comparison, and integration with enterprise imaging workflows. AI

Researchers have developed a new multimodal framework for classifying DICOM image series, integrating both image content and acquisition metadata. This approach uses a bi-directional cross-modal attention mechanism and a metadata encoder that handles missing or inconsistent data without imputation. The system is designed to manage variable series lengths and image dimensions, demonstrating superior performance over existing methods on both in-domain and out-of-domain evaluations. AI