In recent years, artificial intelligence has begun to develop at an accelerating pace and has also appeared in the field of healthcare. In this blog, I will point out the transformative power of AI and how it helps medical professionals work more efficiently.
Medical image diagnostics is one of the primary uses of artificial intelligence in this field. AI can analyze x-rays faster and more accurately than humans, so the patient gets a more accurate diagnosis sooner, which can be key to increasing life expectancy. For example, software from the company Lunit Insight detects breast cancer in mammograms with 96 percent accuracy (Lunit Inc., n.d.)
Predictive analytics is another potential way to use AI in healthcare. By studying the data of many people and building models from it, it is possible to determine how much risk a particular person has for a particular disease. With personalized recommendations, diseases can be prevented and this takes a big burden off the healthcare systems, which are quite overburdened almost everywhere in the world.
With the help of virtual health assistants, the patient can be more involved in the management of his own health. This technology facilitates the work of doctors and helps the spread of telemedicine. I think everyone knows the rudimentary way to do this, since during COVID you could only reach doctors by phone. Of course, no software has yet helped their work here.
The line could be continued for a long time. I think the main question is how much patients will trust the new technologies. Who will you believe in a few years when you have to consult a doctor with a problem? For your doctor or artificial intelligence? There may come a point when artificial intelligence will rather diagnose, because it will simply be more accurate and efficient, the question is how much people will trust it.
References
Lunit Inc. (n.d.). https://www.lunit.io/en/products/mmg
Hi, thank you for your interesting post!
I like how you gave a short overview of the current medical trends regarding AI and its applications. I think you are right by stating that the main issue here is if people are willing to trust AI, even though it has been proven to clearly perform better than medical professionals in certain areas, such as image diagnostics. I think one of the main issues underlying this is the question as to who assumes responsibility when the AI makes a mistake. Say an AI falsely concludes someone does not have breast cancer, and this person then is much too late with treating their cancer, who will be responsible? A possible solution would be putting a medical professional in charge of double-checking the AI, however, this would largely eliminate the efficiency improvements the AI would bring. You could argue that this potential risk is outweighed by the amount of cases where the AI will prevent misdiagnoses, but in my opinion this would be the main reason people might not trust AI. People need someone to blame in these situations, they need a scapegoat, and I think AI does not suffice as one.
The other two applications of AI in medical care you posed seem much more risk-free, and I think AI will go a long way in improving predictive analytics and virtual health assistants.
I think if the AI is open-source, anonimized, and regulated, the issue of building trust will merely be a matter of time. It will become more and more normal to go off of an AI’s judgement, until it will be the norm.