News

​RapidAI has announced the receipt of U.S. Food and Drug Administration (FDA) 510(k) clearance for the latest version of its Rapid ICH product, an artificial intelligence (AI)-powered tool designed to assist in the triage and notification of suspected intracranial hemorrhage (ICH) in non-contrast head CT images. This updated version boasts a 97% sensitivity and 100% specificity, making it one of the most accurate solutions available for ICH detection. By analyzing CT scans, Rapid ICH can detect hemorrhages as small as 0.4 ml, prioritize cases within the radiology worklist, and automatically send notifications to PACS, email, and the Rapid mobile app, thereby streamlining decision-making processes for clinicians. The integration of Rapid ICH with Rapid Hyperdensity further enhances its capabilities by quantifying and characterizing hyperdense regions, providing additional context to aid in triage and transfer decisions. This advancement aims to reduce false positives and alleviate notification fatigue among healthcare providers, ultimately improving patient care in acute neurovascular situations

​Stony Brook University's Ookami supercomputer, operational since 2020, has joined the National Science Foundation's ACCESS (Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support) program as a resource provider. Ookami is the first system outside Japan to utilize the A64FX processor developed by RIKEN and Fujitsu, which was also employed in the Fugaku supercomputer, previously the world's fastest machine. This architecture is particularly suited for memory-intensive applications, such as sparse-matrix solvers, commonly used in engineering and physics simulations. Researchers nationwide can now access Ookami through the ACCESS program, facilitating advanced computational and data-driven research across various scientific disciplines

Researchers at the University of Maryland's Human-Computer Interaction Lab have developed DramatVis Personae (DVP), a web-based AI tool designed to help writers identify and address stereotypes in their creative works. DVP analyzes characters in a narrative by tracking mentions and actions, allowing authors to assign demographic information such as age, ethnicity, and gender. The tool visualizes the representation of these social identities throughout the story, enabling writers to detect and mitigate unconscious biases. User studies have shown that writers can more efficiently identify biases using DVP compared to traditional text editors, all while maintaining full control over their creative process.

​Stony Brook University's Reality Deck, located in the Center of Excellence in Wireless and Information Technology (CEWIT), offers an immersive 1.6-billion-pixel virtual reality experience. Designed by Dr. Arie Kaufman, the Reality Deck serves as a training ground for high school students interested in computer programming and visual computing. In collaboration with kidOYO®, a nonprofit organization focused on computational literacy, local high school students are mentored by Stony Brook PhD candidates. The program aims to inspire students, particularly from underrepresented groups, to pursue careers in computer science and related fields. Through hands-on experiences, students develop skills in coding, research, and immersive technologies like augmented and virtual reality. The initiative not only enhances technical abilities but also fosters a passion for innovation and problem-solving among the next generation of technologists.