A new artificial intelligence (AI) tool is being built to help doctors better identify and treat a deadly, but under-recognized, form of lung failure. Approximately 200,000 people in the United States suffer from acute respiratory distress syndrome (ARDS), which occurs when the lungs fill with fluid, and organs are deprived of oxygen. The challenge is that up to 40 percent of the time it can be missed.

“ARDS is under-recognized because these patients are often extremely ill and have other life-threatening conditions, such as shock, pneumonia, or trauma. Since diagnosis depends on the patient meeting a number of criteria, it is easy for one of the criteria to be attributed to another acute condition, rather than to ARDS,” said Dr. Michelle Ng Gong, chief of Research, Critical Care at Montefiore Health System and professor of medicine and of epidemiology & population health at Albert Einstein College of Medicine. “By using new technology we hope to help clinicians identify ARDS as early as possible, when treatment may be most effective.”

Funded by a $1.2 million Agency for Healthcare Research and Quality grant, the new tool will screen patients throughout Montefiore and flag people at risk of developing ARDS. Once identified, clinicians will receive guidance on the best practices for treating patients with the condition.

To create the screening tool, researchers and data and computer scientists are building off of Montefiore Einstein’s existing AI platform, combing through years of de-identified patient data to determine dozens of critical, objective data points for ARDS. After developing a profile, the AI will run in the background of the electronic medical record system, flagging any patients who match the profile created for ARDS. This ARDS-specific tool is an extension of the work already conducted by the Critical Care and health data teams at Montefiore Einstein, which shows that AI can improve outcomes for people with another lung condition—severe acute respiratory failure.

“We have seen the power of AI and predictive analytics to accurately pinpoint patients at risk for other critical conditions and believe AI can be effective in helping clinicians identify patients with ARDS too,” said Parsa Mirhaji, M.D., Ph.D., director of Clinical Research Informatics, director of the Center for Health Data Innovations at Montefiore Einstein and associate professor, Systems and Computational Biology at Albert Einstein College of Medicine. “The ultimate goal is for AI to become a standard tool for clinicians, helping them provide the best care for our patients.”