In the 1970s, Stanford University researchers created an AI system that asked doctors questions about a patient’s symptoms and generated a diagnosis. In the 1990s, AI algorithms began decoding X-rays, CT scans and MRI images to spot abnormalities that humans might miss. We have come a long way since then, but what if I told you that currently there is AI being developed that could predict the probability a patient would be affected by a certain ailment, or die?
The power AI would have over the healthcare industry once fully integrated is unimaginable. It could save time in many ways: refilling prescriptions, providing instant responses to patient questions after hours, scheduling appointments, filling out billing and discharge paperwork, and much more. This would in turn decrease the number of professionals needed to staff a practice, and decrease the amount of hours they would need to work, saving money in the long run. While these are all relatively low risk tasks for AI to complete, other hospitals are looking to utilize artificial intelligence to optimize patient care in hospitals. One hospital that is spending millions to develop their own AI technology for clinical use is Mount Sinai. Currently, this hospital has a program that uses AI to generate scores that predict patient death risk, a measure which has been useful so far. One such score alerted a nurse that a patient had a kink in their chest tube that was causing their blood O2 levels to drop, saving their life. Other AI measures being explored by Mount Sinai is a program used to analyze mammograms to detect cancer, predicting and flagging drugs that may cause birth defects, and even to identify people with abnormal heart rhythms.
While such developments are undoubtedly impressive, it is also important to consider the potential limitations of AI use in its current state. Many medical professionals fear using AI in healthcare will come at a great cost to patients. Some issues that have been brought up are how AI would adhere to HIPAA and patient privacy concerns, the accuracy of such programs and how they may lead to incorrect diagnoses, and overall how this might introduce distrust in medicine. Researchers have also argued that it would be difficult to measure AI performance across different racial and ethnic groups; some AI programs have already been shown to exhibit racial biases.
While integrating AI into healthcare seems like a promising idea, a lot of research needs to be done to increase its accuracy and determine a suitable role for AI within healthcare systems . Nevertheless, AI will never be able to replace human interaction between medical professionals and their patients, and should instead be used as a tool to help with diagnoses.
Edited By: Heiley Tai
Designed By: Eugene Cho
Citations
Burke, G., & O’brien, M. (2023, October 20). Health providers say AI chatbots could improve care. but research says some are perpetuating racism. AP News. https://apnews.com/article/ai-chatbots-racist-medicine-chatgpt-bard-6f2a330086acd0a1f8955ac995bdde4d
Gupta, D. (2023, August 29). 10 Ways Healthcare Chatbots are disrupting the industry. Appinventiv. https://appinventiv.com/blog/chatbots-in-healthcare-industry/
Schwartz, N. (n.d.). How Mount Sinai uses AI to predict death risk. Becker’s Hospital Review. https://www.beckershospitalreview.com/innovation/how-mount-sinai-uses-ai-to-predict-death-risk.html
Timsit, A. (2023, August 2). AI can detect breast cancer as well as radiologists, study finds. The Washington Post. https://www.washingtonpost.com/health/2023/08/02/artificial-intelligence-breast-cancer-screening/
WP Company. (2023a, August 18). Hospital bosses love ai. doctors and nurses are worried. The Washington Post. https://www.washingtonpost.com/technology/2023/08/10/ai-chatbots-hospital-technology/
WP Company. (2023b, August 18). Hospital bosses love ai. doctors and nurses are worried. The Washington Post. https://www.washingtonpost.com/technology/2023/08/10/ai-chatbots-hospital-technology/