There is an increasing number of smartphone applications (app) that diagnose a patient based on their reported symptoms, one in particular is called Ada Health. The founders of Ada Health have a passion for helping those who have less access to healthcare but the app is useful for anyone looking for answers. Users answer a series of personal questions, questions related to their health and questions about symptoms they are experiencing. Using artificial intelligence this app gives a number of possible illnesses ranked by the percentage of people that had each illness and reported similar symptoms. This app takes risk factors and geolocation into account. The app offers next step guidance recommending self-care, visiting a pharmacy or visiting a health care provider. Tracking geolocation gives insight into trending heath concerns in many different populations, this information also helps with outbreak control (Azadzoy, 2019).
Personal and medical information shared on this app are stored separately, however tracking data is controversial. It poses a problem that a company has information about millions of people’s health. However, the main purpose of the app is to help people. Is it worth diagnosing an individual whose data can then be added to a larger pool of data which becomes more accurate at diagnosing the next person that has the same symptoms, even if the company collects and stores personal information? Even if information about uses is shared with the government? The same questions apply to whether it’s worth preventing an outbreak and improving trending health problems reported within a certain population.
If this app is incredibly accurate at diagnosing an individual why don’t more health professionals recommend it to patients, just as a way to confirm a diagnosis they’ve already received? It might be because a doctor would look unsure of their diagnosis and would not want a patient to lose confidence in their ability. The other way around a doctor might be offended by a patient trying to diagnose themselves. A conflict of interest is that it would take away revenue from labs if people diagnosed themselves without extensive testing.
There is a human component to any health care provider that artificial intelligence can’t replace, at least not yet. Humans have a level of reasoning a computer can’t achieve. Hopefully between AI and specialized health professionals diagnosing individuals will become even faster and more accurate. I know people who have been misdiagnosed so I want others to be aware of apps that can help lead to a diagnosis.
References
Azadzoy, H. (2019, June 26). Ada health - youtube. YouTube . Retrieved September 30, 2021,
from https://www.youtube.com/watch?v=hE6nq3vDrv0.
Medical experts. Ada. (n.d.). Retrieved September 30, 2021, from https://ada.com/team/medical/.
Hey Tera,
ReplyDeleteI think this is a really interesting topic and one I am considering myself as a career. I think AI has both advantages and disadvantages as you described, but I think in terms of helping people the benefits outweigh the downsides. With regulation, I think people's data and health information can be kept safe, and algorithms can be ethically built without assigning medical information directly to a person's identity. Interestingly, AI has received the most investment than any other economic sector in the world. I have talked to some older physicians about their contempt for modern day computers, but I think incorporating computers and AI will ultimately benefit the patient, which is what healthcare should be about. I found this article that goes along with some major points you mentioned, and I think you’ll find it interesting. Thanks for posting!
Serag, A., Ion-Margineanu, A., Qureshi, H., McMillan, R., Saint Martin, M.-J., Diamond, J., O’Reilly, P., & Hamilton, P. (2019). Translational AI and Deep Learning in Diagnostic Pathology. Frontiers in Medicine, 6, 185. https://doi-org.dml.regis.edu/10.3389/fmed.2019.00185