Emory Healthcare aims to predict and manage the fall risk of hospitalized patients through artificial intelligence (AI) cameras. It is also developing a service that allows AI to instantly translate doctor explanations for foreign patients.
Alistair Erskine, chief information officer of Emory Healthcare, met with ChosunBiz before the opening ceremony of the Korea International Medical & Hospital Equipment Show (KIMES 2025) held at COEX in Seoul on the 20th. This year's Medical Korea theme is 'AI-based personalized healthcare becomes part of daily life.' Erskine served as the first keynote speaker at the opening ceremony, discussing 'the real application of AI to enhance the digital healthcare experience.'
Emory Healthcare is the largest medical system in Georgia, USA, with seven general hospitals, including Emory University Hospital. Erskine also serves as a professor of pediatrics at Emory University School of Medicine and the university's chief digital officer.
Emory Healthcare is recognized as an institution actively adopting digital technology alongside Mayo Clinic and Johns Hopkins Hospital. Emory Healthcare invested $100 million (approximately 146 billion won) to build an AI-based data consolidation system. Erskine noted that a system is also being developed where AI listens to and records conversations between patients and doctors. He explained, 'AI recognizes conversations between doctors and the medical team and immediately connects them to clinical notes for written documentation.' This aims to reduce the documentation workload for medical staff.
A service that automatically translates for foreign hospitalized patients by listening to doctors is also under development. For instance, when a Korean patient visits, an interpreter currently accompanies them, but in the future, when the doctor explains in English, the AI system plans to instantly translate it into Korean for the patient. This will allow foreign patients to communicate with doctors without an interpreter and improve the efficiency of medical treatment.
The introduction of an 'AI camera' that predicts the fall risk of hospitalized patients is also underway. The AI camera, equipped with motion detection sensors, predicts the possibility of falls based on patient movements and alerts nurses. Erskine said, 'It usually takes about 45 seconds for a patient to fall in a hospital, but the AI camera can detect falls in advance and notify nurses, allowing them to rescue the patient in just 18 seconds.' He stated, 'The pilot project showed an 85% reduction in fall rates within the hospital.'
Emory Healthcare has also developed an AI algorithm that improves the treatment of sepsis patients. Sepsis is a life-threatening disease where the blood is infected by bacteria or viruses, causing damage to organs, with a mortality rate reaching 50% in severe cases. Erskine noted, 'By analyzing patients' vital signs, we identify specific patient groups and recommend using balanced solutions instead of normal saline to them,' adding that 'more than 70% of medical staff actively utilize this system.' This AI algorithm has received approval from the U.S. Food and Drug Administration (FDA).
Managing medical data, including patient personal information and medical records, through a cloud system is also a key focus for Emory. Medical staff can check and update patient information through mobile devices. Managing data in the cloud benefits hospitals and alleviates inconveniences for patients. He cited patient diagnostic imaging data as an example.
Erskine pointed out, 'Patients have to keep their diagnostic imaging data on CDs and bring them to the hospital when needed, which is a typical medical legacy.' By storing and managing diagnostic imaging data in the cloud, patients can access it directly at the hospital without needing to bring CDs. Patients can also store their diagnostic imaging data in the MyChart cloud.
Erskine said, 'Emory Healthcare aims to utilize AI across research, clinical trials, and surgical settings,' adding that 'the AI system can quickly find suitable patients for clinical trials of specific treatments and promptly suggest the optimal treatment methods and surgical rooms for patients.'