"MedTrinity-25M" is a large-scale multimodal medical dataset released by the UCSC-VLAA team, containing 25 million medical images along with detailed annotations. This dataset features multi-granularity annotations, supports the training of multimodal large medical models, and involves a complex construction process that includes data processing, metadata extraction, region of interest localization, and the collection of medical knowledge. Detailed descriptions are generated using large-scale language models to enhance data usability. The dataset was officially released on July 21, 2024, accompanied by a pre-trained model.

MedSAM-2, based on the Segment Anything Model 2 (SAM2) framework, brings revolutionary changes to the field of medical image segmentation. Its core breakthrough treats 2D and 3D images as video sequences, enabling efficient segmentation of three-dimensional images. The 'one-click segmentation' feature allows users to specify a target in a single image, and the model can automatically recognize and accurately segment the same type of object in subsequent images without the need for continuous input. As a general model, it can handle any image.

By 2035, the number of medical imaging AI tools approved by the FDA is expected to skyrocket, with venture capital support exceeding $30 billion. Over the past seven years, the availability of medical imaging AI products has continued to grow, with medical imaging accounting for 85% of digital health venture capital. Research indicates a strong correlation between venture capital support and the development of AI products, with an estimated 350 new products expected to enter the market by 2035.

The global AI medical imaging market is projected to reach $29.8 billion by 2032. North America holds the largest market share, but the Asia-Pacific region is expected to grow the fastest. Applications of artificial intelligence in radiology and the demand for CT scans are driving market growth. Deep learning and computer vision are the primary technologies, with breast imaging and orthopedic applications being dominant.

OBIA includes 937 patients and 1.93 million medical images from the 301 Hospital, covering 9 modalities and 30 anatomical sites. OBIA realizes the de-identification, management, and quality control of imaging data, providing users with browsing, retrieval, and download services. OBIA fills the gap in the open sharing of biomedical imaging data in the country and helps promote the development of medical imaging AI.