Is AI your doctor’s next best friend?

 

Imagine a world, where your healthcare is personalised to your genetic makeup, lifestyle and medical history. Well, it sounds like the standards of healthcare we’ve all been expecting! However, realistically, with the amount of patients the healthcare system is overwhelmed by, personalised medicine is not accessible to most individuals. Precision medicine is a field and movement that aims to revolutionise healthcare and raise the standards of healthcare by providing personalised medicine to every individual, using AI to fuel it. 

Stories of individuals being ignored by medical professionals are numerous. Nothing is more frustrating and embarrassing than being told your feelings and symptoms are invalid by a professional. The reality of receiving care from humans includes the risk of patients being neglected due to biases, inattention and fatigue, an unavoidable part of an overwhelmed healthcare system. Could AI help fill in the void of human errors? What about the risks of machine error and biases in AI? As of now, there is no perfect solution or answer, and research continues to show areas which both AI and humans exceed and fail at. But the synergy between AI and precision medicine demonstrates a promising future in the healthcare industry, through prevention, diagnosis and treatment.

Integration of AI into medicine

While a lot of fear surrounds the risk of professionals being replaced by AI, in the medical field it is a tool used in support of expertise and insights of professionals, augmenting their capabilities and providing tools to make more informed decisions and improve patient care. So don’t worry about your doctor being replaced, but think of it like AI is your doctor’s partner in crime, working to put you back together! After all, inpatient care can never be fully replaced. 

So how exactly is AI being used in the medical field, and what should we expect? AI technologies allow vast amounts of data to be analysed at a scale previously unimaginable. One of the main ways AI is being used is for image-based diagnostic systems that can detect abnormalities, for example, an algorithm trained using mammogram images and health records which has the potential to reduce the number of missed diagnoses of cancer by doctors. 

Apart from the incredible accuracy of diagnostic AI, possibilities for your genetic makeup and lifestyle to be analysed and taken into consideration for treatment is now more realistic than ever. AI has also been used to improve treatment for those who don’t have access to medical care due to environmental and social factors. For example, deep learning has been used to identify patients with diseases and viruses in resource-poor areas that increase the personalization of the treatment. 

Fears surrounding AI…

Now that we’ve gone through the positives of using  AI in precision medicine, it’s time to discuss the problems many of us fear with AI being used in our personal lives. Biases in healthcare are unavoidable in traditional medicine due to years of data based on western societies, overlooking women’s health and other minority groups, AI models trained on data can amplify biases towards groups of people, and can affect the quality of patient care. The only way to improve the data is through collaboration of the biomedical community in a combined effort to make healthcare better and accessible for everyone. However, fears surrounding data and privacy make many people hesitant and reluctant to trust AI. 

So how would it make you feel if your doctor used AI?Would it make you trust your doctor less or do you believe AI is far more powerful? Do the pros outweigh the potential data risks? At the end of the day, our experience as a patient is vital so while we use AI to improve medicine, let’s not forget about the importance of human touch in times of need.  

Jiya Rawat

Johnson, K.B., Wei, W.-Q., Weeraratne, D., Frisse, M.E., Misulis, K., Rhee, K., Zhao, J. and Snowdon, J.L. (2021), Precision Medicine, AI, and the Future of Personalized Health Care. Clin Transl Sci, 14: 86-93.

Leave a Reply

Your email address will not be published. Required fields are marked *