How Can AI Help with Acute Aortic Emergencies?

Dr. Starnes is a vascular surgeon at the University of Washington. He is one of the first adopters of artificial intelligence (AI) in aortic intervention. He uses Viz.ai to help coordinate care for aortic dissections and ruptured aortic aneurysms. He began to implement this due to frustration with an outdated workflow. He serves a large patient population in Washington, Alaska, Idaho, Montana, and Wyoming. With different hospital systems and antiquated methods of communication, he realized it was very inefficient to evaluate a patient from some of these locations, and then have them transferred to Seattle for surgical repair.

Dr. Starnes overviews the outdated workflow that’s prevalent in aortic emergency care. If there is a ruptured aortic aneurysm or aortic dissection, he would first get a call from an ER physician who ordered the imaging. The transfer center wouild be contacted, and then he had to find a desktop to view images from the outside facility. If there was no way to view the images due to incompatible PACS, he had to use a screenshot of an image sent by a provider at that hospital. After reviewing the imaging, he would decide whether to accept the transfer. If a patient is transferred, he would do the procedure and then hand off the patient to the ICU team, who was rarely (never) aware of this transfer until the patient arrived in their unit.

After starting to use Viz.ai, this process has been streamlined. Dr. Starnes modeled the AI platform he uses for aortic emergencies in a similar way that AI stroke alert platforms already function. He now gets an alert on his phone, he is able to view good-quality images on his phone wherever he is, decide on the next steps, and communicate with members of the team in a HIPAA-compliant fashion all via the user-friendly interface. He uses AI software to detect ruptures and dissections and reports that it is very accurate. Dr. Starnes and colleagues at the University of Washington do over 350 aortic cases per year. The implementation of AI has helped them work more efficiently and has improved patient outcomes by cutting down the time from diagnosis to intervention. He hopes that machines can be trained to measure the aneurysm size for stent graft selection and manage elective aortas by integrating surveillance, follow-up, and elective repair. He also is very hopeful that AI will be able to identify many genetic aortopathies due to the integration of genetics and AI.