Artificial intelligence (AI) is a general term for technology that allows computers or machines to “think” like humans and solve problems. Computers can process large amounts of information to detect patterns and learn to recognize those patterns and respond to them. AI programs may see patterns and understand connections that humans can’t, because they are able to process so much more information at one time. There is a lot of talk about AI in the news today, but there are already multiple examples of AI that you may have encountered in your daily life, including the digital “assistants” Siri and Alexa, GPS directions in your car, predictive text on your phone, and even self-driving cars. I should note that for this blog post, I asked Google’s Gemini chatbot to come up with an illustration, and the result is above!

 There are many ways in which AI could improve medical care, from taking on mundane but time-consuming tasks like gathering records or reconciling medications to more complex analysis of medical information to help doctors reach a rare diagnosis. The applications for AI technologies in medicine are expanding every day. For example, AI programs are being used to help radiologists read medical images, such as CT scans, and AI powered robots are being used to assist in surgeries, helping to make them less invasive with shorter recovery times. An AI medical scribe can “listen” during a clinic visit and compose a progress note to document the doctor’s findings and recommendations, and it’s even been shown that AI programs are able to pass medical certification exams!

These types of innovations have a lot of potential to aid in the diagnosis of complex conditions, such as interstitial lung disease (ILD). Getting to an accurate ILD diagnosis requires the review of lots of information from clinical history, physical exam, lab test, CT scans, and sometimes lung biopsies. ILD specialists discuss this information together in a multidisciplinary discussion with other specialty doctors, such as radiologists and pathologists, who have significant experience with ILD diagnosis. AI technologies are great at managing extensive data, learning from previous cases, and recognizing patterns. It’s not a stretch to imagine that one day an AI program will assist the multidisciplinary team with their diagnostic considerations.

Of course, this approach will have to be studied before it can be broadly adopted, but it’s certainly exciting to consider how technology might shorten the time to ILD diagnosis from two or more years to weeks or even days.

AI is already being used now in parts of the evaluation of ILD. Robotic bronchoscopies and thoracic surgeries can obtain biopsy tissue in less invasive ways with shorter recovery times. Developed by a machine-learning algorithm, the Envisia genomic classifier test is used on lung biopsy tissue from less-invasive bronchoscopies to help doctors identify a certain pattern of lung scarring called usual interstitial pneumonia (UIP) by its molecular signal, and in some patients, may help them avoid a surgical lung biopsy. And recently, the FDA approved the Fibresolve AI test to aid physicians in analysis of chest high resolution CT scans in patients with pulmonary fibrosis.

Jesse Ehrenfeld, the president of the American Medical Association, was recently quoted as saying, “It is clear to me that AI will never replace physicians—but physicians who use AI will replace those who don’t.”

AI has the potential to be transformative for medical care, including for the care of patients with ILD.