Artificial Intelligence and modern medicine

BY Rhea Sebal

(The Nucleus writing competition runner up)

The practice of medicine is an art – where a physician must explore multitudinous layers of variables before making a medical decision. While reading through peer reviewed scientific literature, I came across several highly interesting – though opinionated – articles surrounding the theme of Artificial Intelligence (AI) and the extent to which it can aid the work of a physician. I would like to highlight the salient word in my last sentence –  ‘aid’ – not ‘replace’. Without a doubt, I believe that AI does not have the ability to completely replace clinicians. 

“The practice of medicine is an art”

Firstly, we must appreciate the dynamic changes AI can bring, such as the capacity to rapidly accelerate aspects of medicine in ways that were not achievable previously. At the most advanced level, robots have been equipped with technology that uses navigation and computer-assisted intervention techniques for intraoperative surgical support. These include assisting in pre-operative planning (here surgeons plan the surgical procedure on the basis of existing medical records) in accuracy, automation and speed. The objective is to delve into mapping, thus aiding the surgical team in 3D tissue tracking. Another speciality of medicine where AI appears to be particularly promising is gastroenterology, in particular the ways in which AI can improve the quality of endoscopic procedures. AI has proven useful for identifying and labelling anatomical sites of the upper digestive tract, which in turn reduce the blind spot rate of endoscopy (when these common cancers have the ability to be missed by endoscopists). The overall outcome of AI and its technology is to improve the success rate for detection of false diagnosis. It is clear and undeniable that prospects for the application of AI in medicine are vast.


“Artificial Intelligence & AI & Machine Learning” by mikemacmarketing is licensed under CC BY 2.0

Whilst it is important to fully appreciate the marvel of these technologies, we must simultaneously acknowledge the profound challenges of launching AI in medicine.The most obvious risk is that AI systems could malfunction unexpectedly, posing a grave and immediate threat. Even ‘minor glitches’ in software or hardware can have a catastrophic effect on patients. Additionally, data confidentiality and privacy concerns leave many experts questioning the safety of AI. 

Surely, if AI can facilitate diagnosis, help create new drugs, and act as virtual nurse assistants in this decade, then it doesn’t seem unreasonable to say in the coming years it will replace humans in medicine, making the job of a physician futile. Unfortunately, I believe this to be a fallacy. Proponents of AI-enabled machines one day replacing humans are simply overlooking the sheer complexity of healthcare, and frankly, the key essence of medicine in itself. Who would not want their doctor to give them or their loved one’s solace, comfort and a hand-to-hold when being told about a life-threatening diagnosis? We all need someone to listen to and empathize with us during our pain and suffering, rather than receiving information from plastic-bodied, human-shaped robots.

“The good physican treats the disease; the great physician treats the patient who has the disease.”

Although, at the end of an interaction between physician and patient, there may be medications that need to be prescribed, there are a multitude of holistic factors that must be discussed prior. Social, occupational, demographic and psychological risk factors are vital in obtaining thorough understanding of a patient’s illness, and allowing doctors to act with a much better judgment. Even if such risk assessments  were fed into machines, their communication would remain cold and impersonal. It is important to note how doctors consistently use the body language, mannerisms and subtle cues a patient gives to deduce an understanding of the presenting problem, and the effects it has on the patient. It is not uncommon for patients to withhold information from their clinician, due to fear of judgment or discomfort, when discussing sensitive topics. Thus, the human ability to infer that a patient might be struggling more than they are letting on from various indicators such as appearance, tone of voice, and personal hygiene is an intuition unable to be replaced by AI.

“medication #medicine #medical, #pharmacy” by Eyad Elbayoumi is licensed under CC BY 2.0. 

‘The good physician treats the disease; the great physician treats the patient who has the disease’. Such intricacies of medicine is something AI cannot replicate, thus displaying the exigency of physicians for now, and for many decades to come. This is an indispensable paradigm that I, as a future physician, fully advocate for. 

Leave a Reply