Will Artificial Intelligence Help Improve Healthcare?

As a student of computer science, I have always believed that most of the modern-day tasks can be performed by computers provided there is someone smart enough to design an algorithm for the said task. However, when it comes to healthcare I strongly believe we should be wary of using AI techniques that may replace a physician by autonomously diagnosing an ailment without any human intervention, since precious human lives will be at stake. Moreover, very stringent regulations will be needed to determine whether such tools should be available to untrained users, which I believe will only hinder other necessary technological developments that may benefit humanity.

With the help of online resources such as WebMD, general populace tries to self-diagnose their ailments which often leads to wrong diagnosis and use of improper medicines which further worsens the disease and in some cases the delay in getting right medication can be fatal. While AI based diagnostic tools may only be designed to assist clinicians, it wont be long before mass population start using the tools themselves for diagnosis and prescription without involvement of any medical professional which can be catastrophic. Because such tools will only be meant to augment a clinicians decision-making process and not replace one. Strict regulations may prevent the mass availability of those tools but there will then always be other illegal means to gain access to them, or due to severe regulations and restrictions development of such tools may become very expensive which in-turn might increase the overall cost of healthcare.

It is suggested in the study that with increased availability of EHR (Electronic Health Records), personalized treatments will be made possible based on numerous factors such as a patients ethnicity, genetic information, geographical information etc. While this approach of personalized treatment can be very effective, one should ponder upon the possible misuse of openly available electronic health data by corporations and blackmailers.

Presently, health insurance companies charge premium based on an individuals personal and family health history. However, with development of AI techniques and easily available health data insurance providers can pick out certain at risk individuals and then may charge very high insurance premium or even deny providing coverage. Furthermore, big pharma companies will start marketing drugs targeted towards a group without any medical relevance only to pad their bottom line.

On one hand deployment of advanced intelligent robots such as Da Vinci for assisting in complicated surgeries will improve standard of healthcare manifold but when it comes to personal care in hospitals I strongly believe that a human element is a necessary intangible component in the process of healing. No robot can match the delicate touch and compassion of a human being be it for a menial task such as bringing in food tray or more complicated task of helping a patient rehabilitate to walk again after a major invasive procedure.  Therefore, while few contributions of robotics in assisted surgery may be beneficial but a large-scale deployment of robots for patient care may not be helpful to achieve better healthcare standards.

At this point of time it is difficult to estimate whether development of artificial intelligence techniques will help improve and subsidize healthcare or make it even more expensive and complicated, shrouded in complex regulations. With appropriate regulatory oversight and human judgement in developing right tools it may turn out to be useful.
