AI in Radiology: Transforming Imaging & Diagnosis
- Academic researchers and tech firms are collaborating to create new tools leveraging artificial intelligence in radiology.
- Mutaz Shegewi, senior research director at IDC Health Insights, suggests healthcare organizations can integrate AI in radiology by leveraging existing infrastructure, such as medical imaging storage.
- Reeder of UW radiology notes that any AI technology considered for their health system must first secure FDA approval. UW Health has already implemented FDA-approved advanced image reconstruction...
AI in radiology is revolutionizing medical imaging and diagnosis,leading to enhanced patient care and streamlined workflows. Discover how FDA-approved image reconstruction, powered by artificial intelligence, is slashing scan times by up to 50% and reducing radiation exposure. Emergency rooms benefit from AI-driven triage via cloud computing, ensuring rapid data processing and swift alerts for radiologists. UW Health is at the forefront, integrating AI tools across various imaging modalities. Learn how they utilize a mix of in-house servers, cloud resources, and specialized software to analyze scans and generate critical insights for brain perfusion. News Directory 3 keeps you informed on the latest developments in healthcare technology. Discover what’s next as AI further integrates for improved diagnostic accuracy and efficiency.
AI Enhances Radiology, Improving Patient Care at UW Health
Academic researchers and tech firms are collaborating to create new tools leveraging artificial intelligence in radiology. Radiologists report early results show promise in boosting patient care and optimizing workflows.
Mutaz Shegewi, senior research director at IDC Health Insights, suggests healthcare organizations can integrate AI in radiology by leveraging existing infrastructure, such as medical imaging storage. However, Shegewi emphasizes the need to revamp enterprise systems, adding, “They need workflow integration. They need computing power, and they’re going to need governance and security.”
Faster Scans and Improved Emergency Room Care
Dr. Reeder of UW radiology notes that any AI technology considered for their health system must first secure FDA approval. UW Health has already implemented FDA-approved advanced image reconstruction in its scanning machines. This technology yields sharper images with fewer artifacts and slashes scan times by 30% to 50%, potentially reducing a patient’s radiation exposure.
“Patients like it,” Reeder said. “We like it too, becuase it means we can schedule shorter exam slots. It improves throughput and workflow. It’s a game changer.”
John Garrett, director of imaging informatics at UW, explains that UW health uses a mix of in-house servers, cloud resources, standard computers, and the imaging machines themselves to support AI tools in radiology. For example, while standard computers can run some AI models, specialized software for CT and MRI brain perfusion relies on an on-premises server to analyze brain scans and generate color-coded blood flow images.
For computationally intensive tasks, such as CT scans evaluated by 10 to 12 AI triage tools, Garrett said the university uses GPUs in the cloud. Amazon Web Services, Microsoft Azure, and Google Cloud Platform handle real-time data processing, depending on the specific algorithm.
In emergency triage, data is processed by multiple AI algorithms in the cloud. The results are then sent back to UW health’s picture archiving and communication system (PACS). Radiologists receive alerts about high-priority findings via a desktop widget on their PACS workstations, Garrett said.
UW Health radiologists also use Nuance PowerScribe,an AI-powered voice recognition software,on Dell computers to streamline report writing. Reeder, a longtime user, notes its continuous enhancement.
“It allows us to generate reports efficiently, accurately and in a standardized way,” Reeder said. “While it’s not perfect, it is accurate. You can say fancy medical words and it gets it right.”
What’s next
UW health plans to further integrate AI to improve diagnostic accuracy and efficiency, focusing on expanding its use in various imaging modalities and clinical applications.