AI in Healthcare: Do the merits outweigh privacy concerns?
Life in the 21st century is being excessively driven by data. As the years have drifted past, there has been an increasing and extensive use of AI to perform tasks once performed my man. But what is AI? AI or Artificial Intelligence is a field of study where one teaches computers how to learn. AI is being applied all around us. Be it the classical music composer Musenet or the mind-boggling field of astronomy, AI has its application almost everywhere.
Today a new set of tasks is being slowly taken over by machines with the help of AI. Interestingly, data is the fuel for these machines. These machines are gradually making some of the human tasks redundant. Machines are showcasing the potential to perform tasks with absolute precision and efficiency. Will Machines in the future be able to think for themselves without any human intervention? Ultimately will machines replace humans? These questions raise some serious apprehensions in our minds. Unfortunately, there’s a lot to be perturbed about. AI is bringing about a revolution but at what cost? Are there any privacy concerns? We need to carefully weigh the pros and cons before coming to a concrete conclusion.
Integrating AI in healthcare:
Artificial intelligence (AI) in healthcare is the use of algorithms to estimate and further imitate the ability of the human brain to understand complicated medical data. This is basically done by feeding the computer a lot of data which the machine then studies over a period of time and brings about inferences by forming patterns from the data. AI can also be equipped with self-correcting and learning abilities which help the system get better accuracy over a period of time.
The healthcare industry involves two primary functions: first, screening and diagnosis, and second treatment and monitoring. Health-related AI applications mainly aim to analyse the relation between prevention or treatment techniques and their bearing on the patients. AI Algorithms scrutinize the humongous amount of data and are first able to tell you what happened in the past, then predict what will happen in the future, and finally what you should do next.
Moreover, AI is showing potential in diverse hospital administrative processes thus leading to a reduction in overall costs by around 30%. AI can automate tasks like pre-authorizing insurance, following-up on unpaid bills, and maintaining records, to ease the workload of healthcare professionals. The single greatest challenge, in the healthcare industry, is the exponential growth of data which will be doubling every 73 days by 2020 and there simply won’t be enough medical personnel to handle this new flood of information. This highlights the exigency for AI in the near future to organize and control these silos of data.
The Iron Triangle Theory in Healthcare:
‘The “Iron Triangle” in health care basically refers to the concept that accessibility, cost and quality cannot all be simultaneously improved. The premise is that an improvement in one area results in a decline in at least one of the others. Can Artificial Intelligence break the iron triangle theory? Reports suggest that the integration of AI in healthcare can possibly help in achieving the three goals of access, cost and quality simultaneously in the near future. Since bringing in AI will bring down hospital costs by almost 30%, the stakes of it being cost effective are high. Moreover, taking into consideration the recent developments in AI and the pace at which we are progressing in this field, it is highly plausible that AI in healthcare would be both effective and largely accessible.
Application of AI in Healthcare and its Merits:
AI is increasingly being applied to healthcare with the help of Internet of Medical Things (IoMT) which basically connects healthcare devices with healthcare IT systems over the internet. IoMT wearable devices are able to monitor our pulse rate and our calorie intake and are also capable of suggesting optimum fitness techniques. Moreover, with the help of IoMT, medical practitioners can keep a track of recovery of their patients simply through a mobile phone application.
AI is already being used to detect diseases, such as cancer, more accurately and in their early stages. The use of AI is enabling review of mammograms 30 times faster with 99% accuracy, reducing the need for unnecessary biopsies. Studies reveal that the number of X-Rays, CT Scan and MRI examinations is bound to increase rapidly while the growth in number of radiologists will not be proportionate. In such a case, AI will be of tremendous importance in examining these reports with great precision and efficiency. Similarly, AI has the potential to detect diseases like sepsis that are not otherwise easily detectable by a doctor thus giving a longer time window for a doctor to intervene.
The path to the patient from the research lab has always been a long and costly one. Researchers suggest that it takes an average of 12 years for a drug to travel from the research lab to the patient. Only 5 in 5,000 of the drugs that begin pre-clinical testing ever make it to human testing and just one of these five is ever approved for human usage. Furthermore, it costs a company millions of dollars to develop a new drug. AI in Drug research can go a long way in reducing time and increasing the availability of various drugs to the end consumer.
AI technology is also gaining momentum in the customer service domain. The world is likely to see healthcare bots very soon. Healthcare bots will be used to schedule appointments with the patient’s healthcare provider. These bots can help patients with their medication as well. AI also shows great potential in training future medical practitioners. AI algorithms can create real life situations through various simulation systems.
Moreover, in the healthcare industry, nearly 86% of the mistakes are preventable. Thus, integrating AI in Healthcare can substantially reduce hospital error which is one of the leading causes of patients’ death.
Privacy and Ethical Concerns:
AI definitely has proved to be a valuable tool in modern day healthcare. But one cannot overlook its privacy concerns. While doctors are bound by confidentiality and ethics, the same doesn’t hold true for AI systems. Moreover, there is no comprehensive piece of legislation to regulate the use of our data by AI systems. This puts us at the risk of our vital health data being used without our consent by third parties. Our data is already being used by mega corporations to seal billion-dollar deals. Our entire medical history is accessible at the click of a button by these third parties.
This specifically came to light when Facebook rolled out a “suicide detection algorithm” in an effort to promote suicide awareness and prevention. The system used AI to gather data from one’s posts and then predicted one’s mental state and propensity to commit suicide. This was definitely a positive use case for AI in healthcare. But the fact that Facebook is storing and using our data cannot be overlooked. Our personal health data could be sold to advertising agencies who then customize advertisements according to one’s medical condition.
AI systems are powered to carry out predictive analysis. On the one hand, this is great for customers who can use this information to take preventative action. But on the other hand, insurance companies could use this predictive genetic testing to bias selection processes and charge higher premiums. Due to higher statistical prevalence of certain diseases among certain demographics, there is a huge risk that incorrectly-trained insurance algorithms will demonstrate racist and sexist behaviour if left to their own devices.
The issue at hand is that AI is what can be called a black box technology. In simple words, a machine cannot give reasons for its conclusions just as humans can. This could lead to “toxic” biases, resulting in unfavourable outcomes.
Regulations and Compliance:
It is now well established that AI applications are as accurate or better than medical experts. As a result, they’re being adopted at a fast pace. But as AI’s application becomes more and more widespread, stringent regulations governing their usage will have to be framed. The need for comprehensive regulations is moreover necessary for deciding who will incur responsibility in case of any malfunction on the part of the machine. In case the system goes wrong, should the doctor or the developers be held responsible? This question stands relevant even today. Without a regulatory and legal framework to inculpate software companies in case of a mistake, the risk of integrating AI into healthcare is too high.
An overarching issue about the acceptance of AI into medicine is the need for regulatory approval. Approving a machine learning algorithm is different from approving a new vaccine. It takes a lot to get regulatory approval for an algorithm. Once the algorithm learns from more data and adapts, it no longer remains the same algorithm. The new learning can’t be implemented in the clinical practice without updating approvals. Thus, the regulatory cycle is not set up to address the machine learning environment.
The breadth of healthcare diagnostics and treatment methods that can be advanced through artificial intelligence has grown significantly since the 1970s but regulatory environment hasn’t. For AI to gain traction in healthcare, regulatory frameworks must be created and then companies can create the compliance structures necessary to smooth adoption into healthcare that will save lives. Tech companies need to compulsorily educate users on exactly how their data will be used and their consent must be taken. Technology companies need to be held accountable in case of any error. Governments too must implement up-to-date legal frameworks surrounding healthcare data.
In conclusion, integrating AI in Healthcare has revolutionized the way the healthcare industry functions. AI has proven to surpass the accuracy level of humans. It has also proved to be an effective cost-cutting tool in the long run. But at the same time, there are a lot of privacy issues that need to be addressed immediately. There needs to be a stable regulatory framework with which it can comply with. Once we’ve cleared these ethical and moral hurdles, AI solutions for healthcare will deliver invaluable medical breakthroughs. But ultimately it won’t be possible to completely wipe out human presence from the healthcare industry. AI will greatly reduce the burden on doctors by assisting them in a majority of tasks but will not be able to replace them, at least in the years to come. In fact, that was never the purpose of AI. In the end, we can’t forget the humans whose data those AI algorithms constantly learn from.
