A look at what Artificial Intelligence can do in analyzing a human body. Used for diagnosing patients and coming up with treatment plans. Image Credit: https://www.engineering.com/DesignerEdge/DesignerEdgeArticles/ArticleID/17664/A-Healthy-Future-for-Artificial-Intelligence-in-Healthcare.aspx
By Samantha Das, JHBL Staff Member
Technological advancements are propelling Artificial Intelligence (A.I.) through the healthcare system, leading the quality of diagnoses for certain illnesses and diseases to improve, while also, however, increasing liability risk. In creating these new algorithms for A.I., developers may become more susceptible to liability under tort law. It is important to weigh the benefits and risks of new technology emerging in the health field when deciding whether to adapt to a new healthcare system or go back to the old ways of patient care.
A.I. is the collection of technology that uses complex algorithms and software to analyze, interpret, and comprehend complicated data. When used in the healthcare industry, A.I. can help healthcare professionals understand patient health data because A.I. develops effective medical diagnostics, treatments, and research. A common form of A.I. is “machine learning,” a statistical technique that finds patterns in massive amounts of data. When used in the health care industry, that data is being processed to learn the study of different algorithms to improve medicine. Machine learning is most applied to study the precision of medicine, such as predicting which treatment plans are likely to succeed on a patient. As machine learning and analytical algorithms are improving, A.I. is turning into a system that can outperform doctors.
Regarding diagnosing and treating diseases, A.I. has accurately identified cancer patients. For example, in an experimental study, an A.I. algorithm designed by Google Health and DeepMind to spot breast cancer had outperformed six human radiologists on average by 11.5%. Not only can A.I. improve diagnostics, but with the improvement of its system, it can perform personalized assessments, offer health care advice, assist patients with managing drug regimens, and process different languages in speech recognition. An introduction of A.I. in the current health industry can help address the lack of availability in doctors, nurses, or other clinically trained staff members. As of right now, A.I. has integrated its way into the healthcare industry; however, it is too soon to say that A.I.’s ability will completely replace the need for human healthcare providers.
In terms of medical research, A.I. has the ability to bring human health data together and store patient information into a record that is later used in clinical trials and research studies. In the healthcare industry, A.I. can be used to store patient information through Electronic Health Records (EHR). EHR is a digital version of a patient’s charts that is recorded and can be shared with other healthcare providers. Although effective, EHR, on its own, has its downfalls, including data overload and issues with exchanging information to other EHR systems as well as other healthcare providers. With the introduction of A.I., the burnout in EHR improves substantially. The ultimate goal is for A.I. to be used in EHR to collect information from all clinicians involved in patient care to identify patterns faster. This way, those who use EHR can identify important correlations and offer recommendations for treatment through a seamless process.
With constant new modifications, A.I. enhances healthcare as we know it. However, also enhanced are the risks for liability. Developers of the algorithms may become liable under tort law for flaws in the algorithms. On the one hand, this seems logical because if developers are creating new modifications and algorithms that are used in A.I., with the purpose to enhance medical care, then it should be able to do just that. If A.I. fails in its performance, then patients are the ones who suffer, and therefore someone has to take the fall if the product does more damage than good. The sensible solution is for developers to assume the risk of liability because they are the ones who built A.I. for use in the healthcare industry.
Although those responsible for making a flawed system should be found liable, courts are reluctant to extend product liability to software developers. First, A.I. is meant to help healthcare providers make final decisions. A.I. is not a complete replacement for a healthcare professional’s expertise or knowledge when making a diagnosis. Therefore, providers should know to use it with discretion and awareness of potential flaws. Suppose there is risk of a false diagnosis or a treatment plan gone wrong. In that case, the doctor administrating the decisions deserves to face liability for fully relying on A.I. without confirming through their own knowledge and expertise. Secondly, suppose developers could be found liable whenever there is a discrepancy. In that case, fears of being brought to court if something goes wrong may deter others from developing new modifications that could better improve A.I. in the healthcare industry. New innovations in A.I. is a positive development that has allowed for improvements all over the health sector. Fears of being brought to court should not stop the growth of A.I., which is why there should be a reluctance in courts when deciding whether developers should be found liable.
It is important to pay attention to the development of A.I. and its direct role in the healthcare industry. A.I. is a technology that is meant to make diagnosing and treating patients a much more seamless and accurate process. Although there are many benefits to A.I., the healthcare industry should be hesitant in implementing it. A.I. should not replace human doctors because this would simply undermine providers who spent years studying and going to school to work in their job field. Although we are training machines to be smarter than humans, there can be many different complications when robot-like machines assume predominant control in the healthcare industry.
As of now, A.I. is unlikely to replace human providers. However, as technology continues to advance, there is no way of telling what can happen. No matter how advanced technology becomes, “robots” and machines should not be at the forefront of the health field. Most often, human interaction builds rapport for hospitals, medical centers, etc. Therefore, A.I. should only be used to help with that rapport and not completely replace humans.
There should be a continuation in enhancing A.I. and making it stronger as the years go on because of the great benefits. There is nothing wrong with wanting to improve a software that can have such a positive effect on medicine across the world. The only potential danger is when there are substantial errors in its algorithms that can significantly affect a patient’s health. As times are changing, it is important to adapt to the new technologies emerging.
In conclusion, A.I. has changed the face of patient care in the health field, and as it continues to change every day with new modifications and algorithms, there are added liability risks as a result. With every new system comes risks of liability if outcomes go wrong, but that shouldn’t be a deterring factor for those developing A.I. Doctors should assume liability because they supervise A.I. performance, using their expertise and knowledge. The benefits outweigh the risks with all of its capabilities in improving patient care, which has been a well-needed asset.
Samantha Das is a second-year law student at Suffolk University Law School with an interest in health and compliance law. She is pursuing a concentration in Health and Biomedical Law.
Disclaimer: The views expressed in this blog are the views of the author alone and do not represent the views of JHBL or Suffolk University Law School.
Sources:
https://builtin.com/artificial-intelligence/artificial-intelligence-healthcare
https://ehrintelligence.com/news/how-can-artificial-intelligence-ai-improve-clinician-ehr-use
https://www.healthit.gov/faq/what-electronic-health-record-ehr
https://medicalfuturist.com/ten-ways-technology-changing-healthcare/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6324122/